It’s said with a kind of moral superiority, the way someone tells you they don’t eat processed sugar. And like the sugar thing, it’s mostly good for you and also slightly more complicated than the person saying it believes.

I want to pull apart what “defined risk” actually means, because there are at least three different things hiding under that one phrase, and confusing them is how people who thought they were being careful end up surprised.

A key part of “defined risk” is that you have to define the magnitude of the risk. More on this later.

The three things people mean by “risk”

When someone says a trade is defined-risk, they could mean any of three different things, and they’re usually not sure which.

Max theoretical loss. The most the position can lose at expiration, computed from the structure. A long call’s max loss is the premium. A debit spread’s max loss is the debit. An iron condor’s max loss is the width minus the credit. This number is real, it’s knowable in advance, and it’s printed on your screen. This is the thing people usually mean. The problem here, is that you can risk up to your entire account balance on defined risk trades.

Margin requirement. The cash your broker freezes to let you hold the position. For defined-risk debit structures this roughly equals the max loss, which is nice and clean (though psychologically dangerous). For undefined-risk structures (naked strangles, naked puts), margin is computed by a formula that has opinions about volatility, and those opinions change. This is the thing people usually forget, but that can be helpful in some situations.

Path risk. What happens to your account between now and expiration, not just at the end. This is the thing people don’t even know they should be thinking about, and it’s the one that actually gets them.

The strangle sleeve, my primary income engine, is undefined risk. A naked short strangle has, in theory, unlimited loss on the call side and very large loss on the put side. I say this plainly because the whole publication is built on not lying to you (or myself). The strangles are where the real money is and they are not defined-risk trades. The calendars and explosion positions are defined-risk. The portfolio is a deliberate blend.

So how do I sleep? Because “undefined” in theory and “uncontrolled” in practice are different things. Let me explain the gap, because the gap is where the whole craft lives.

The screen says $725. Your account disagrees.

Here’s the scenario that actually hurts people, and it has nothing to do with the theoretical max loss.

You sell a strangle. The broker freezes, say, $1,500 in margin. You’ve got plenty of cash, you feel responsible, life is good. Then vol spikes. Not even a catastrophe, just a real move, the kind that happens several times a year in commodities. Two things happen at once, and they conspire against you.

First, your position loses money, as expected. Second, and this is the killer, your margin requirement expands. The broker looks at the newly volatile underlying and decides this position is now riskier to hold, so it freezes more of your cash. The position that required $1,500 now requires $3,000. You haven’t done anything. You haven’t added a trade. The market moved and your broker quietly doubled the rent.

If you were running close to fully invested, you now get a margin call. And margin calls don’t care about your beautiful thesis. They force you to close positions right now, at whatever price the panicked market is offering, which is always the worst possible price. You get liquidated at the bottom, locking in losses on positions that, had you been able to hold them, would very likely have reverted as vol came back down.

This is the mechanism. This is how short-vol accounts die. Not from a single position hitting some theoretical max loss. From a vol spike that expands margin across every position simultaneously, forcing liquidation at the worst moment, turning a survivable drawdown into a terminal one. February 2018 (Volmageddon) and March 2020 are the famous examples, and the graveyard is full of accounts that were “being careful” right up until they weren’t.

Moral of the story: You CANNOT be fully invested in undefined risk strategies due to path risk and margin expansion forcing liquidation at bad prices.

Therefore, we hold at least 50% margin in reserve initially for undefined risk (so a strangle that initially requires margin of $500, we hold $500 in cash on top of that) to mute this risk, and allow for adjustment in crises.

The thing that actually defines your risk

Here’s the reframe that I think matters more than the defined/undefined distinction everyone obsesses over.

Your real risk isn’t defined by the structure of any single trade. It’s defined by whether you can survive the worst plausible path across your whole portfolio without being forced to act. And the tool that controls that isn’t the trade structure. It’s the cash reserve.

I hold roughly 50% of the strangle portion of the account in unencumbered cash at all times. Not deployed. Not earning the maximum. Just sitting there, looking lazy, doing nothing. Every fiber of your optimizing brain hates this. That cash is “wasted.” You could be selling more premium with it. Your returns would be higher.

Right up until the vol spike, at which point that lazy cash is the only thing standing between you and forced liquidation. The reserve absorbs the margin expansion. It lets you hold positions through the spike instead of being liquidated at the bottom. It converts a potential account-ending event into an uncomfortable Tuesday. The reserve is, in my view, the single most important risk control in the entire framework, and it has nothing to do with whether any individual trade is “defined-risk.”

Defined-risk structures are genuinely useful, to be clear. The calendars and explosion positions can’t expand their margin on me, can’t get me liquidated, can’t surprise me. That’s exactly why they’re in the portfolio alongside the strangles. They’re the part of the book that’s immune to the margin-expansion death spiral. But they’re the smaller engine. The strangles do the heavy lifting and the cash reserve is what makes carrying undefined risk survivable.

So is “I only trade defined-risk” wrong?

No. It’s a fine rule, especially if you’re newer, trading a small account, or know yourself well enough to know you won’t maintain a cash discipline under stress. Defined-risk structures genuinely remove the margin-expansion failure mode. If that’s the failure mode you’re worried about, and it should be, defined-risk is a legitimate and smart answer.

I just want you to choose it for the right reason, which is “I want immunity from margin expansion and forced liquidation,” not the wrong reason, which is “the screen shows me a small number so I’m safe.”

The Key Psychological Takeaway

In defined risk trades, you can risk your entire account on one trade. The psychology of not risking too much is much harder to enforce when “defined risk” is making you feel secure. The margin amount in undefined risk trades is much higher than the premium sold; there is “built in” protection at the broker level. A guardrail, built in.

The people who blow up selling premium almost never do it because they didn’t know their max loss. They do it because they were fully deployed when the margin doubled. Or, because they sized far to large with defined risk positions that were correlated, and all went poorly together. Don’t do this. It’s the whole job.

Next theory post: how to actually read a COT report without falling asleep or fooling yourself.

We are all students

It’s a reasonable fear. It’s also, often, exactly backwards. The reason a fund isn’t doing your strategy frequently isn’t that your strategy is bad. It’s that your strategy is too small for them to bother with. And that, it turns out, is the closest thing to a structural, permanent edge that an individual trader will ever have. You can’t buy it. You can’t out-research your way to it. It’s a gift that comes free with being small, and almost nobody appreciates it.

Let me explain what capacity actually is, why so many genuinely excellent strategies can’t hold institutional money, and then why I think the usual conclusion people draw from all this (”so small strategies don’t matter to big players”) is wrong in an interesting way.

What capacity means

Capacity is the amount of money a strategy can absorb before the act of deploying that money degrades the returns. Every strategy has a number. Some can hold tens of billions. Some top out at a few hundred thousand dollars before they stop working. The number is determined almost entirely by liquidity and the costs of interacting with it.

Here’s the core problem in one sentence: the market is not a price, it’s a stack of prices, and the more you buy, the worse the prices you have to accept.

When you, trading a $10,000 account, sell a corn strangle, you sell one contract. You get filled at the mid, or close to it, and your order is so small the market doesn’t even notice you exist. The price you see is the price you get. This feels normal. It is, in fact, a luxury.

When a fund managing two billion dollars wants to put on that same corn strangle in size, they don’t get to sell one contract. They need to sell thousands. And here’s what happens when you try to sell thousands of contracts into a market that only has a few hundred resting bids near the current price: you eat through the entire bid stack. The first hundred contracts fill near the mid. The next hundred fill a little worse. By the time you’ve moved your full size, you’ve pushed the price meaningfully against yourself, and your average fill is far from where you started. That gap, between the price you wanted and the price you actually got, is slippage, and for large orders in thin markets it can be enormous.

Why so many great strategies simply can’t scale

Now layer a few realities on top of raw slippage, because they compound.

Thin markets have shallow books. The niche, illiquid, neglected corners of the market (second-tier futures, obscure options chains, small-cap names, weird calendar spreads in non-headline contracts) are precisely the places where mispricings persist because nobody’s looking. But “nobody’s looking” and “thin order book” are the same fact stated two ways. The mispricing exists because the market is small. The smallness that creates the opportunity is the same smallness that prevents it from scaling. This is the central, almost cruel irony of capacity: the best inefficiencies live in exactly the places that can’t hold size.

You become your own counterparty. Once your orders are large enough to move price, you start fighting yourself. Entering pushes the price away from your entry. Exiting pushes it away from your exit. A strategy that makes 8% a year on a million dollars might make 2% on a hundred million purely because the entry and exit costs scale up faster than the edge does. At some size, you are the market, and you can’t trade against yourself profitably.

Some edges are tiny per-unit and only work via repetition on small size. A strategy might capture, say, fifteen dollars of mispricing per contract on a niche options spread. Beautiful return on a small account. But fifteen dollars per contract is invisible to a fund that needs to deploy capital in million-dollar increments; the transaction costs and price impact of getting big enough to matter would swallow the fifteen dollars whole. The edge is real and the edge is uninvestable for them, simultaneously.

Reporting, mandates, and career risk. A fund running outside money has constraints an individual doesn’t: position limits, risk committees, the need to explain to investors why 30% of the book is in some commodity nobody’s heard of, and the brutal asymmetry that a manager who tries something weird and loses gets fired, while a manager who sticks to the consensus and loses keeps their job. Even when a big player could technically access a niche edge, the institutional machinery often makes it not worth the career risk. Individuals have no risk committee. I answer to no one but my own spreadsheet.

So the standard story goes: the best small-capacity edges are the individual trader’s natural habitat, the giant funds are structurally locked out, and that’s the retail edge. All true. And this is where most people stop. I think stopping here is a mistake.

Why the usual conclusion is wrong

The conclusion people draw is: “small-capacity strategies don’t matter to big players, so they’re a permanent retail-only preserve, and big funds are stuck in their crowded large-capacity trades forever.”

I don’t buy it, and the reason is a single word: concatenation.

Here’s the thing nobody accounts for. Yes, one niche strategy might only hold $500,000 before it degrades. But there isn’t one niche strategy. There are hundreds. Hundreds of small, uncorrelated, high-return-on-capital inefficiencies scattered across thin markets, obscure contracts, neglected underlyings, and structural mispricings that each individually can’t hold size.

A sufficiently sophisticated fund doesn’t need any single one of them to scale. It needs to run fifty of them at once. Fifty strategies that each hold a million dollars is a fifty-million-dollar book. A hundred such strategies is a hundred million. And critically, because these niche edges live in genuinely different markets driven by genuinely different forces (a soybean calendar mispricing has nothing to do with a small-cap merger spread which has nothing to do with a neglected FX vol skew), they’re uncorrelated with each other. This is the entire business model of the most successful multi-strategy funds on earth. They are not finding one giant scalable edge. They are concatenating hundreds of tiny unscalable ones into something that, in aggregate, holds enormous capital and produces a smoother equity curve than any single strategy could.

Think about what that aggregation actually buys you. Each individual strategy has its bad months, its regime where it stops working, its idiosyncratic failure mode. But fifty uncorrelated strategies bleeding and printing on different schedules sum to a return stream with a fraction of the volatility of any one component. The Sharpe ratio of the portfolio of strategies can be dramatically higher than the Sharpe of the best single strategy in it, for the same reason a diversified portfolio of stocks beats a single stock: the noise partially cancels while the edge accumulates. A 0.8-Sharpe strategy is unremarkable. Fifty uncorrelated 0.8-Sharpe strategies blended together can produce a portfolio Sharpe north of 3, and that is the number that lets a fund raise and hold billions.

So the giant funds are not, in fact, locked out of the niche-edge world. The best of them live in it. They’ve just solved the capacity problem at the portfolio level instead of the strategy level. They concatenate their way to scale.

Where this leaves the individual

If that were the whole story, it’d be depressing: the funds win again, just through more sophisticated machinery. But it isn’t the whole story, and here’s the genuinely good news for someone like me, trading small.

I don’t need fifty strategies to put my capital to work, because my capital is small. I need a handful of high-ROI niche edges, run at a size where I get clean fills and pay almost nothing in slippage, and I capture those edges at full strength because I’m small enough to be invisible. The fund running my same soybean-calendar edge has to share it across a billion-dollar book and pay impact costs to get in and out at size; I take the same edge at one or two contracts and keep nearly all of it. On a percentage-return basis, I can outperform the fund on that specific edge precisely because I’m not dragging its capacity constraints behind me.

And I can borrow the funds’ best insight without their handicap. The concatenation logic, running several uncorrelated approaches at once for a smoother curve, scales down just as well as it scales up. That’s the entire architecture of what I do: three (for now!) sleeves with different dependencies, plus the separate directional book I wrote about, each one a different way of making money with a different failure mode. I’m doing in miniature exactly what the multi-strats do in size, except I get to do it in the highest-ROI niche corners that they can’t profitably reach.

That’s the edge nobody can buy. A fund can buy talent, data, technology, and leverage. It cannot buy smallness. It cannot make its billion dollars behave like my ten thousand. The capacity constraint that limits me to small size is the same constraint that lets me operate in the richest, most-neglected parts of the market and keep nearly all of what I find there. The funds concatenate hundreds of diluted edges to overcome their size. I concatenate a few undiluted ones and keep the full juice on each.

The conventional wisdom says capacity is what separates the pros from the amateurs, and that the pros’ larger capacity is the advantage. I think it’s the reverse. Capacity is a constraint that grows with success, and the individual trader’s permanent, unpurchaseable edge is having almost none of it. Stay small enough to fish where the big boats can’t follow. It’s the one advantage that gets worse the more money you manage, which means it’s the one advantage they can never take from you.

Next theory post: probably a piece on the defined risk myth, and why undefined risk positions can often be less “risky” in practice.

We are all students.

In the meantime, onward!

For about forty years, the most boring sentence in finance was also the most reliable: “put 60% in stocks and 40% in bonds.” It was the default. Your 401(k) probably does some version of it right now. The logic was elegant and, for four decades, basically true: stocks go up over time but crash sometimes, bonds are calmer and tend to rise when stocks fall (because the Fed cuts rates in a crisis, which lifts bond prices), so the bonds cushion the stock crashes and you get a smoother ride than either alone.

It worked beautifully. Until 2022, when it didn’t, and the failure is one of the most instructive events in modern markets for anyone thinking about how portfolios actually break.

The bet hiding inside the “diversified” portfolio

Here’s the thing nobody mentions when they sell you 60/40: the entire cushion depends on a single assumption. That stocks and bonds move opposite each other. The negative correlation is the whole mechanism. Take it away and you don’t have a diversified portfolio, you have two different ways to own the same risk.

And that negative correlation, the one everyone treated as a law of nature, was actually a feature of a specific regime: the roughly forty-year stretch of disinflation and falling interest rates from the early 1980s onward. In that world, bad economic news meant lower rates meant higher bond prices, reliably, while stocks suffered. The hedge held because the macro backdrop made it hold.

What people mistook for a permanent structural truth was really a long-running environmental condition. The diversification was real but conditional, and almost nobody was tracking the condition. They just saw “stocks and bonds zig and zag oppositely” and assumed it was physics.

2022: the condition flipped

Then inflation came back for the first time in forty years, and the Fed hiked rates hard and fast to fight it. Watch what that does to the machine.

Rising rates crush bond prices directly (that’s just bond math, prices fall when yields rise). Rising rates also crush stocks, because higher discount rates lower the present value of future earnings and because expensive money slows the economy. So the exact same force, aggressive rate hikes, drove both asset classes down at the same time. The negative correlation didn’t just weaken. It flipped positive at the worst possible moment.

The result: 2022 was one of the worst years in history for the 60/40 portfolio. Stocks down roughly 18%, bonds down double digits too, no cushion anywhere. The thing that was supposed to protect you in a bad year for stocks was also having its own worst year in a generation, driven by the same cause. Diversification didn’t fail at the margins. It inverted exactly when it was needed.

This is the nightmare scenario for any diversified book, and it has a name: correlations going to one. The relationships you relied on for protection all snap to the same direction precisely during stress. Your many bets reveal themselves to have been, all along, one bet wearing several costumes.

Why I bring this up (it’s not to dunk on 60/40)

I’m not here to tell you 60/40 is dead or stupid. It’s neither. It’s a reasonable default that had one genuinely terrible year after four good decades, and it has largely recovered since. If you’re saving for retirement on a long horizon, you could do far worse, and most people do.

I bring it up because 2022 is the cleanest possible illustration of the thing I worry about most in my own portfolio, and the thing my framework is explicitly designed around. Every diversified strategy has a hidden assumption about correlations, and the assumption is always weakest exactly when you’re relying on it most.

My strangle sleeve is “diversified” across twelve uncorrelated futures: grains, currencies, metals, rates, energy. In normal times those markets genuinely don’t move together, and the portfolio is smooth precisely the way 60/40 was smooth for forty years. But I have lived through the months where four positions stopped out at once, and every single time it was a broad risk-off event that moved currencies and commodities and rates together. My decorrelation, like 60/40’s, is real most of the time and partially fails during exactly the stress events where I need it. I am not immune to the 2022 lesson. I am the 2022 lesson, on a smaller scale, waiting to happen.

So what do you actually do about it

You can’t make correlations stay put. They’re a feature of the macro regime, not something a clever allocation fixes. What you can do is hold something whose payoff doesn’t depend on correlations behaving.

This is the entire reason the explosion sleeve exists, and it’s the deepest difference between my framework and a 60/40.

The overpriced sleeves we’re selling (strangles, calendars) are correlation-dependent. They rely on markets behaving normally, on diversification holding, on vol reverting. They make steady money while the regime cooperates, and they suffer together when it doesn’t. They are, structurally, the same kind of thing 60/40 is: a bet that the normal relationships persist.

The explosion sleeve is correlation-independent. It owns deep out-of-the-money options across many markets, and it’s designed to pay off precisely in the violent, everything-moves-together events that hurt everything else. When correlations go to one and the short-vol book is bleeding, that’s exactly the environment where some far-out-of-the-money option goes from worthless to enormous. The explosion sleeve isn’t diversification in the 60/40 sense (more uncorrelated bets). It’s a different species of protection: a bet that pays more the more correlated everything else becomes.

That’s the piece 60/40 never had. Its “hedge” (bonds) was just another correlation-dependent asset that happened to usually zig when stocks zagged, until it didn’t. A true tail hedge doesn’t rely on a relationship holding. It relies on the relationship breaking, and profits from the break.

The honest caveat

The explosion sleeve usually costs money to carry. Most months it bleeds premium while the short sleeves make it. That drag is real and it’s the price of the protection, the same way insurance premiums are a real cost in every year your house doesn’t burn down. I’ve watched most explosion positions expire worthless and I felt every dollar of it. The temptation to cut the “wasteful” hedge is strongest right before you need it, which is, of course, the whole trap. 2022 caught a generation of investors who had quietly decided the cushion was unnecessary because it had been so long since they needed one.

The 60/40 lesson isn’t “bonds bad.” It’s “know which of your bets are secretly the same bet, and hold at least one thing that gets better when they all go wrong together.” That second thing is uncomfortable to own because it loses money in the calm times that make up most of your life. But the calm times aren’t when portfolios die. The correlated-stress times are. And those are the only times a real tail hedge is trying to be useful.

Next theory post: A piece on why capacity (the amount of money a strategy can hold before it stops working) is the individual trader’s single biggest and most underrated edge over the giant funds.

We are all students.

I owe you this one. Life and trading got ahead of the posting schedule, so rather than pretend April and May happened in separate universes, you’re getting both months in one post. From here on out, these reviews will be monthly and on time. Probably.

If you missed the March review, the format is simple: what happened, what it means, and what I learned. No retroactive edits, no conveniently forgetting the trades that didn’t work. The framework posts cover the why. These reviews cover the what.

One formatting note going forward: trades are tracked in the month they were opened, not the month they closed. Some trades span multiple months. You’ll see their entry in one review and their resolution updated in a later one. This keeps the narrative clean and lets you follow each trade from the moment it enters the portfolio.

The Account

Starting value (April 1): ~$10,185 Ending value (May 31): ~$9,769 Net P&L (April + May, closed trades): -$416.00 Two-month return: ~-4.1% Cumulative return (since March 6): -2.31%

Two down months after a positive March. The account is underwater. Let’s talk about why.

Sleeve Reminder

Quick refresher for anyone jumping in (covered in depth in the framework posts):

Sleeve 1 (Short Strangles) sells the overpriced center. 20 delta, 45 DTE, high IVR futures. Mechanical management: 50% profit target, 2x stop, 21 DTE time stop.

Sleeve 2 (Double Calendars) captures theta on low IVR underlyings with positive vega exposure. 25% profit target, 50% circuit breaker, 3 DTE time stop on short legs.

Sleeve 3 (Explosion Positions) buys the underpriced tails. Deep OTM options where the Convexity Score identifies cheap tail convexity. Direction agnostic. Most expire worthless. The winners cover the losers.

Same thesis across all three: financial returns are leptokurtic. The center is overpriced. The tails are underpriced. We trade both sides.

April Trades (Opened in April)

Canadian Dollar Strangle (/6CM6) — CLOSED ✗

Sleeve: Short Strangles

Structure: Short 0.710/0.735 strangle

Entry: April 1 | 65 DTE

Exit: May 1 Credit received: $360.00

Gross P&L: -$270.00

Net P&L: -$282.68

Days in trade: 30

The first strangle loss in the live account.

The CAD was caught between oil support (bullish for the loonie) and economic weakness (bearish), which was supposed to be the perfect rangebound setup. It didn’t stay rangebound long enough. The position hit exit criteria and we took the loss.

This was also the trade where I deviated from the standard 45 DTE parameter, entering at 65 DTE for liquidity and margin reasons (documented in the trade recap). Whether the longer duration contributed to the loss is debatable. What isn’t debatable: -$270 on $360 in credit is a loss of 0.75x the premium, below our 2x stop level. The mechanical management got us out before the worst case.

Grade: A-. Loss within expected parameters. The system took a hit and moved on.

/MES Double Calendar #2 — CLOSED ✗

Sleeve: Double Calendars

Structure: 6450/6770 double calendar, 2 contracts

Entry: April 6

Exit: April 22

Total debit: $318.00

Gross P&L: -$247.50

Net P&L: -$266.86

Days in trade: 16

This is the trade I want to spend time on, because it illustrates something important.

The market ripped. /MES rallied over 100 points through our upper strike during the life of this trade. The 6770 call calendar got blown through. The underlying moved decisively outside the profit zone and kept going.

A -77.95% loss sounds catastrophic. And the dollar amount (-$266.86) was the largest single loss in the account at the time. But here’s the detail that matters: the maximum possible loss on this trade was $318 (the full debit). We lost $247.50 of it. We did not lose the maximum.

This is the resilience of the calendar structure doing its job, even in failure. When the underlying blows through your strikes on an iron condor, the loss hits max width quickly and completely. On a calendar, the back month long options retain value even when the short options are deep in the money. The vega exposure provides a partial offset. The result is that even a “bad” calendar loss tends to be less than the theoretical max.

We lost 78% of the debit, not 100%. On a trade where the market moved 100+ points through our range, that’s the structure protecting us. Not exactly a celebration. But meaningful in the portfolio math over dozens of trades.

The broader lesson: calendars are more vulnerable to strong directional moves than strangles with wide strikes. The tradeoff is defined risk and positive vega exposure. April showed me the downside of that tradeoff in real dollars. The lesson about re-entry timing is also real. The first calendar closed on April 2 for a profit. I entered this one four days later on April 6, without giving enough time to assess whether the environment had shifted. The market was beginning a directional leg that eventually destroyed the trade.

Grade: B. The structure protected us from max loss. The entry timing did not protect us from the wrong environment.

Corn Strangle #2 (/ZCN6) — CLOSED ✓

Sleeve: Short Strangles

Structure: Short 440/490 strangle

Entry: April 16 | 36 DTE

Exit: May 1 | 21 DTE (time stop)

Credit received: $362.50

Gross P&L: +$63.00

Net P&L: +$48.08

Days in trade: 15

The time stop fired at exactly 21 DTE with the position at +17% of credit, well below the 50% profit target. We closed anyway. The rules don’t negotiate.

The 36 DTE entry (shorter than the standard 45 DTE) compressed the window. With only 15 days between entry and time stop, the trade didn’t have room to reach the full target. But 0.32% ROC per day was actually above our historical average across 130+ strangle trades. The trade was contributing well per unit of time even if it didn’t hit the headline number.

Not every trade needs to be a home run. Consistent base hits compound.

Grade: B+. Profitable, mechanically managed. Shorter DTE entry compressed the window and is tracked as a deviation.

Natural Gas Explosion Position (/NGN6) — OPEN

Sleeve: Explosion Positions

Structure: Long 2.30 put / 5.50 call, 5 delta

Entry: April 7 | 79 DTE

Total debit: $390.00

Current status: Slowly decaying.

The Iran conflict hasn’t transmitted to Henry Hub pricing. US natural gas remains domestically insulated. ~25 DTE remaining as of this writing.

This one is likely heading toward expiration unless a supply shock finally breaks through the domestic insulation. The thesis (global LNG disruption eventually tightening US supply) hasn’t been invalidated, but the clock is running out. No adjustment. We hold or it expires.

May Trades (Opened in May)

British Pound Strangle (/6BU6) — CLOSED ✓

Sleeve: Short Strangles

Structure: Short 1.320/1.385 strangle

Entry: May 4 Exit: May 27 (profit target)

Credit received: $562.50

Gross P&L: +$281.25

Net P&L: +$268.32

Days in trade: 23

The best strangle trade of the account so far by dollar amount. Sterling cooperated perfectly: the BoE held at 3.75%, the pound stayed rangebound, and theta decayed the premium to the 50% target in 23 days. IVR at entry was 76.8. The credit was fat. The management was mechanical.

This is what Sleeve 1 looks like when it works. High IVR, rangebound underlying, profit target hit, move on. The boring trade that makes money.

Grade: A.

Wheat Explosion Position (/ZWU6) — OPEN (and interesting)

Sleeve: Explosion Positions

Structure: Long 525 put / 920 call, 5 delta

Entry: May 4 | 73 DTE (entered for $250 while IVR was 15.4)

Total debit: $250.00

Current status: -$162.50 (-65%) after the May 12 WASDE report sent wheat up ~7% overnight, but the move did not follow through.

The May WASDE dropped new crop wheat ending stocks at 762 million bushels versus expectations of 845 million. An 83 million bushel miss. KC HRW futures hit the daily limit up. Winter wheat conditions are at 28% good to excellent, the worst for this time of year since 2023, after declining for 10 consecutive weeks.

We haven’t reached the first management target (3x, or $750 total value) or expiration yet. We’re not taking anything off. The rules say wait. The fundamental story (deteriorating crop, tight global stocks, elevated input costs) doesn’t resolve overnight. 52 DTE remaining. Full analysis in the trade recap post.

If this one reaches 3x, a single winner covers the /ZB loss, funds the next cycle of tail convexity, and turns the Explosion Position sleeve profitable. That’s the math of Sleeve 3. You don’t need them all to work. You need one.

/MES Double Calendar #3 — OPEN

Sleeve: Double Calendars

Structure: 7300/7540 calendar (reflecting the market’s rally to new highs)

Entry: May 20

Total debit: $285.00 Current status: Early.

The strikes are set at the new range after the rally that destroyed calendar #2. Lessons from April applied: higher strikes, patience on entry timing.

Australian Dollar Strangle (/6AU6) — OPEN

Sleeve: Short Strangles Structure: Short 0.690/0.735 strangle

Entry: May 27 Credit received: $570.00

Current status: Just opened. The fifth strangle entry, diversifying further across the currency complex. AUD shares some of the CAD dynamics (commodity currency, rate differential, China exposure) but with different specific catalysts.

March Trades: Final Updates

Three trades were opened in March and covered in the March review. Here’s how they resolved:

Trade #1, Corn Strangle (/ZCK6): Already closed in March at profit target. +$185.16 net. No update needed.

Trade #2, Treasury Bond Explosion Position (/ZBM6): Expired worthless on May 22. -$301.92 net. The thesis (bonds sell off during inflationary conflict rather than rallying as safe haven) was directionally correct but insufficient in magnitude. 65 days, $302 gone. This is the expected outcome for most Explosion Positions. The strategy has a low win rate by design. You buy cheap tail convexity, most of it decays to zero, and the occasional 5x or 10x winner covers the accumulated losses. If you can’t stomach watching positions expire worthless on a regular basis, Sleeve 3 is not for you.

Trade #3, /MES Double Calendar #1: Closed April 2 at profit target. +$119.07 net in 9 days. The fastest trade resolution in the account. The IVR entry deviation flagged in March didn’t hurt this particular trade. One data point doesn’t validate a deviation, but it’s logged.

The Numbers

April + May Combined (Closed Trades)

Metric April May Combined Trades closed 2 4 6 Wins 1 2 3 Losses 1 2 3 Gross P&L -$118.75 -$222.63 -$341.38 Net P&L -$147.79 -$268.20 -$416.00 Commissions $29.04 $45.32 $74.36

Sleeve Performance (All Closed Trades, Inception Through May 31)

Sleeve Trades Wins Losses Win Rate Net P&L Strangles 4 3 1 75% +$218.88 Calendars 2 1 1 50% -$147.79 Explosions 1 0 1 0% -$301.92 Total 7 4 3 57% -$230.83

The strangles are carrying the portfolio. The calendars are net negative, punished by a trending market. The explosion sleeve is rolling its positions and is waiting for the payoff. That’s the story in three lines.

What I Learned

The calendar sleeve needs more discipline on re-entry timing. Calendar #1 closed April 2 for a profit. Calendar #2 opened April 6 and lost 78% of its debit when the market kept trending. Four days between a winning exit and a losing entry isn’t enough time to assess whether conditions have changed. It is important to understand when the market is trending, and be careful not to enter too many range bound trades under these conditions.

Explosion Positions expiring worthless is information, not failure. The /ZB position decayed to zero over 65 days. Watching real money evaporate on a position you chose to hold to expiration is harder than I anticipated, even with full intellectual understanding that this is the expected outcome. The emotional recalibration is useful. If $302 is uncomfortable, how will $3,000 feel at a larger account size? The answer is: about the same, which is why you size the sleeve as a fixed percentage and think in terms of the system, not individual positions.

Structural protection on calendars is real, even when it doesn’t feel like it. The market blew through the second calendar by 100+ points. On a binary outcome trade, that’s max loss. On a calendar, it’s 78% of max loss. That 22% structural recovery ($70.50 saved) doesn’t feel like much on one trade. Over 20 losing calendars, it’s $1,400 in avoided losses. The structure earns its place in the portfolio even when individual trades fail.

Strangles across uncorrelated underlyings smooth returns, as designed. The CAD strangle lost. The GBP strangle won. Same framework, same sleeve, different underlying, different outcome. Combined net: roughly flat. If I’d traded only one currency, the result would look like either a failure or a success depending on which one. Trading both gives you something closer to the expected value of the system. This is the diversification thesis working in practice, not theory.

Commissions remain a meaningful drag at this scale. $89.20 in cumulative commissions across 7 closed trades. If commissions were zero, the account would be $89 higher. At scale, this is irrelevant. At $10K, it’s shaving real basis points off every trade.

Looking Ahead

The wheat Explosion Position is the most important open trade in the account. Deteriorating crop conditions, tight global balance sheet, and the WASDE shock provide a fundamental catalyst with 52 DTE of runway. If it reaches 3x, one winner reshapes the entire sleeve performance. If it fades, we absorb the loss and move on.

The natural gas explosion is running out of time. The Australian dollar strangle and the third /MES calendar are both in their early days.

For June, the framework continues. New strangle entries as IVR signals appear. New Convexity Score rankings for the next Explosion Position cycle. The same mechanical management, the same trade by trade documentation, the same refusal to pretend the losses didn’t happen.

The account is down 2.31% after three months. The question is not whether the system produces losses (it does, by design) but whether the wins, over time, are large enough and frequent enough to make the math work. Seven trades is not enough to answer that question. But it’s a start. And it’s all on the record.

Three months in. Every trade documented, every loss explained, every deviation tracked. As always, not financial advice.

We are all students.

The Trade

Underlying: /ZW (Chicago SRW Wheat Futures, September 2026)

Structure: Explosion Position (long deep OTM put + long deep OTM call)

Strikes: 525’0 put / 920’0 call

Expiration: July 24, 2026 (73 DTE at entry)

Entry date: May 4, 2026

Premium paid:

• Put: $118.75

• Call: $131.25

• Total debit: $250.00

Management rules (Explosion Position standard):

• Losers: let them expire. No salvage attempts.

• Winners: manage in thirds (1/3 at 3x, 1/3 at 5-8x, final 1/3 rides). In the small account, we may take the middle road and manage at 5-8x.

• No delta adjustments. No spreading against winners. No “locking in.”

Why Wheat

If you’ve read the framework posts, you know the Convexity Score ranks 18 futures markets each month on a combination of tail frequency (how often extreme moves actually happen) and option cheapness (how little it costs to own those tails). Wheat consistently ranks near the top of that screen, alongside natural gas, crude oil, and the yen.

The structural reasons are similar to natural gas, just with different catalysts. Wheat is a weather dependent commodity with inelastic demand (people don’t eat less bread when prices go up, at least not in the short term). Supply is geographically concentrated and can be disrupted by drought, flooding, frost, or geopolitical conflict (as the world learned in 2022 when Russia’s invasion of Ukraine sent wheat to $13/bushel). The distribution of returns has dramatically fatter tails than the options market prices in.

But there’s a broader point here that I want to make, because it goes to the heart of why Sleeve 3 exists.

Tail Events Don’t Care About the S&P

When most people think about market risk, they think about stocks going down. When they think about hedging, they think about buying SPY puts or VIX calls. When they think about tail events, they think about 2008 or March 2020.

This is a remarkably narrow view of what “tail risk” actually looks like in the real world.

The S&P 500 is one market. It’s the most watched, most liquid, most analyzed market on the planet, which means it’s also one of the most efficiently priced for tail risk. SPY put skew is steep. VIX is always pricing in some level of crash premium. The cost of owning equity tail convexity is high precisely because everyone wants it.

Commodities are a different world entirely. The participants are different (farmers, processors, utilities, airlines, not hedge funds and pension managers). The catalysts are different (weather, disease, infrastructure failure, trade policy, not earnings misses and Fed speeches). The correlations to equities are low and unstable. And the options markets are thinner, less efficiently priced, and dominated by hedgers with specific directional needs rather than speculators trying to price every contingency.

This means tail options in commodities are systematically cheaper than tail options in equities, per unit of actual tail probability. Our Convexity Score quantifies this, but you can feel it intuitively: a 5 delta wheat call costs a fraction of what a 5 delta SPX put costs in notional terms, and wheat actually makes 4 sigma moves more frequently than the S&P does.

The implication for portfolio construction is significant. If you want convexity exposure to genuine “the world is different now” events (wars, supply shocks, climate events, infrastructure failures), the commodity complex is where you get the most bang for your buck. Not because commodities are riskier in some abstract sense, but because the options market misprices their tails more aggressively than it misprices equity tails.

This wheat position exists because of that structural mispricing. Not because I predicted a USDA report would send wheat limit up. I didn’t predict that. Nobody predicts that. You own the tails consistently, across multiple markets, every month, and eventually the tails show up.

The Setup: What the Market Was Pricing (and Missing)

When I entered this position on May 4, wheat was trading around $6.10/bushel. IVR was 15.4, which is low (low IVR is a key input to our selection). The options market was pricing in relatively muted expectations for wheat vol over the next 80 days.

This was interesting because the fundamental picture was anything but muted.

US winter wheat conditions were deteriorating rapidly. The USDA’s weekly crop conditions had been declining for ten straight weeks, from 58% good to excellent down to about 28% at the time of entry. That’s the worst rating for this time of year since 2023. Drought across the Southern Plains was worsening, and recent rainfall had missed the driest areas.

The Iran conflict was elevating input costs across the grain complex. Higher energy prices mean higher fertilizer costs, higher chemical costs, higher diesel for irrigation and transport. Wheat is one of the most input intensive crops. Farmers were already signaling acreage shifts away from input heavy crops. Argentina, Australia, and Canada were all projected to see production declines of 10 to 25% from the prior year.

Global trade flows were disrupted. Russia’s IKAR consultancy had just cut its wheat export forecast to 44.5 million tons from 46 million. The Strait of Hormuz situation was creating shipping complications for grain moving through adjacent waterways. Algeria had been actively tendering for large wheat purchases, signaling tight supplies in North Africa.

And the May WASDE report was 8 days away. The May report is the single most important USDA release of the year for grains, because it contains the first official new-crop balance sheet projections. The market knew this was coming. What the market didn’t know (or wasn’t pricing) was how large the revision would be.

All of this said “wheat could move a lot in the next 73 days” while the options market was saying “probably not.” IVR at 15.4 meant we could own both tails for $250 total. That’s less than half the cost of the natural gas Explosion Position. For a crop facing a deteriorating domestic crop, supply chain disruptions from the Middle East, and the most important data release of the year on the calendar.

The Convexity Score was screaming.

And Then the WASDE Dropped

Today, May 12, the USDA released the May WASDE report at noon Eastern. The wheat numbers were not close to consensus.

New crop US wheat ending stocks came in at 762 million bushels. The market was expecting 845 million bushels. That’s an 83 million bushel miss to the downside, or roughly 10% tighter than anyone was pricing. For context, the historical average miss on USDA’s initial wheat forecast is about 51 million bushels. This was well outside that range.

The HRW production number confirmed what the crop conditions data had been screaming: the USDA’s prior estimate of 629 million bushels for hard red winter wheat was disconnected from reality. The market has been sitting 30 to 40 million bushels below that figure for weeks, and the report validated the bears.

Winter wheat conditions, reported just yesterday, fell another 3 points to 28% good to excellent. Heat is expected to arrive into the crop this week. The annual wheat tour is underway, and the commentary from the field is grim.

The market’s response was immediate. KC HRW futures hit the daily limit up of 45 cents. Chicago SRW rallied 32 to 42 cents. /ZW (our contract) surged roughly 7% in the session, now trading at 691’0 at time of writing.

Here’s what’s important: we’re up 33% on the position and the call side hasn’t even scratched the surface of its potential. The 920 call is still deep out of the money (current price is 691, the strike is 920). What’s happening is that the call is gaining value from two sources simultaneously: delta (the directional move toward the strike) and vega (implied volatility expanding on the WASDE shock). Both forces are pushing the option price higher even though the underlying is still 229 below the strike.

Are We Taking Profit?

Not yet. And this is where the discipline gets tested.

A 33% return in 8 days is excellent by any normal standard. If this were a strangle, we’d be well past the exit criteria. But Explosion Positions are not normal trades. The entire framework depends on letting winners run to multiples of the initial debit, because the losers (which are more frequent) expire worthless.

Our management rule says: first partial at 3x ($750 total value). We’re at $332.78 right now. We need the position to roughly triple from here before we start taking any off.

That sounds like a lot, but consider the dynamics. If wheat continues to rally on the deteriorating crop conditions and tight global balance sheet, the 920 call accelerates in value as it moves from deep OTM toward less deep OTM. Gamma increases. Delta increases. Each subsequent dollar of wheat rally produces more call value than the previous dollar did. This is the convexity working for us.

If instead the rally stalls here and wheat settles back into a range, the vol expansion fades, the call loses value, and we go back to waiting. The $250 debit is the maximum loss regardless of what happens. We already knew that going in.

73 DTE remaining. Plenty of time. The WASDE was the catalyst, but the fundamental story (deteriorating US crop, tight global stocks, rising input costs) doesn’t resolve overnight. More data will come. More crop tours. More weather events. The options have time to develop.

We wait. We follow the rules. And if this one keeps running, it becomes the first monetized Explosion Position in the live account, which would be a significant validation of the framework.

The Bigger Picture

Let me zoom out for a moment, because this trade illustrates something fundamental about the portfolio.

While wheat is surging on a USDA crop report, the S&P 500 is basically flat today. Bonds are doing their own thing. The British pound strangle is quietly decaying in our favor. None of these things are correlated in any meaningful way.

A traditional “60/40” portfolio or even a sophisticated equity-centric hedge fund wouldn’t be positioned for a wheat supply shock. They don’t own grain tail convexity. They don’t think about WASDE reports. The catalyst that’s driving this position has nothing to do with the Fed, nothing to do with earnings, nothing to do with the VIX. It’s about rainfall in Kansas and acres planted in Argentina and storage estimates at the USDA.

This is what decorrelated tail exposure looks like in practice. The explosion events in commodities and currencies are often completely disconnected from what’s happening in equities. By owning convexity across different futures markets (grains, energy, metals, rates, currencies), we’re positioned for the shocks that equity-focused investors aren’t even watching.

Not every Explosion Position works. Most don’t. That’s the expected outcome for any individual tail trade. But when one hits (and they do hit, more often than the models predict, because leptokurtosis), the payoff is designed to cover many months of premium decay on the ones that didn’t.

We’re not there yet. 33% is nice but it’s not 3-5x. A lot can happen in 73 days. But tonight, with wheat limit up and the fundamental story intact, I want to document this moment. This is what it feels like when the framework meets reality in real time. Not in a backtest. Not in a theory post. In the live account.

What I’m Watching

Follow through in the overnight and tomorrow’s session. Limit up moves can reverse just as fast as they appear, especially if the initial reaction was amplified by short covering and fund repositioning. The COT data showed funds adding nearly 62,000 long contracts across the grain complex last week. If that positioning unwinds, we could give back today’s move.

Crop conditions over the next 2 to 4 weeks. The USDA releases weekly updates every Monday afternoon. If conditions continue to deteriorate from 28% good to excellent, the market reprices further. Heat arriving this week could accelerate the decline. The annual wheat tour will provide field level confirmation (or contradiction) of the satellite and model data.

Weather across the Southern Plains. Rain in the right places at the right time could stabilize the crop and take some premium out of the move. Continued drought and heat would do the opposite. We own both tails, so the direction of the weather resolution matters less than the magnitude of the market’s response.

The global balance sheet. The WASDE showed tight new crop stocks globally. Argentina, Australia, and Canada are all producing less. Russia’s export pace is slowing. If additional supply revisions come through over the next few months, the bullish case compounds.

Theta working against us. This is the part that’s easy to forget when the position is green. We’re paying $10.15/day in time decay across both legs. That’s the cost of being positioned. If wheat settles into a range and vol compresses, that theta grinds the position back toward zero over the remaining 73 days. The move needs to continue (or a new catalyst needs to emerge on the put side) for this trade to reach the management targets.

Subscribe for every trade, every fill, every outcome. The Explosion Positions are the most volatile part of the portfolio and the most interesting to follow in real time. Tonight might be the night we find out what a monetized tail event looks like in a live account.

We are all students.

Quick portfolio status: Trade #1 (first corn strangle) closed at profit target back in March. Trade #4 (Canadian dollar strangle) closed for a loss on May 1, which we’ll cover in a separate post. This one, the second corn strangle, also closed on May 1 at the 21 DTE time stop. Two exits on the same day, one winner and one loser. That’s trading.

Let’s get into it.

The Trade

Underlying: /ZC (Corn Futures, July 2026 contract) Structure: Short strangle Strikes: 440 put / 490 call Expiration: May 22, 2026 (36 DTE at entry) Entry date: April 16, 2026 Exit date: May 1, 2026 (21 DTE, time stop)

Premium received: $362.50 Close cost: $300.00 Gross profit: $62.50 Fees: ~$14.18 Net profit: ~$48.32 Return on risked capital: 4.77% ROC/day: 0.32% Days in trade: 15

Management rules (same as always):

• Profit target: 50% of credit received → close at ~$181.25 profit

• Stop loss: 2x credit received → close at ~$725 loss

• Time stop: close at 21 DTE regardless of P&L ← this one triggered

Why Corn Again

The first corn strangle (Trade #1, /ZCK6 425/495) was one of the cleanest trades we’ve done. Entered in early March during the initial Iran panic when commodity IV spiked across the board, closed at the 50% profit target 21 days later for +$185.16 net. Textbook execution.

So when corn showed up on the screener again in mid April with elevated IVR, the question was: do we go back to the same underlying?

The answer is yes, and the reasoning is simple. The framework doesn’t care about narrative. It cares about the variance risk premium. If implied volatility on corn is overpriced relative to what corn is likely to actually do, we sell it. Whether we sold it last month or last year or never before is irrelevant to the math. Each trade is an independent bet on the distributional mispricing.

The macro backdrop in mid April was similar to (but not identical to) the March setup. The Iran conflict was still elevating implied vol across the commodity complex, but the initial shock had subsided and markets had entered a more grinding, uncertain phase. Corn specifically was caught between competing forces: rising input costs from elevated energy prices (fertilizer, diesel, chemicals) were squeezing margins, but a faster than average US planting pace was suggesting adequate supply. The USDA had planting at 25% complete as of late April, six percentage points ahead of the five year average.

Those opposing forces meant corn was chopping sideways while the options market continued pricing in the possibility of a breakout. That’s the exact setup we want. Tall middle, fat implied tails. Sell the center.

The Expiration Choice (A Deviation, Documented)

The eagle eyed among you will notice this trade was entered at 36 DTE, not our standard 45 DTE. The /ZCN6 May 22 expiration was the nearest liquid option chain, and the next one out was significantly further dated. Given the choice, I took the 36 DTE entry.

This is a deviation from the stated parameters and I want to be transparent about it. At 36 DTE, you have less time for the trade to work and you’re closer to the gamma acceleration zone than I’d ideally like. The tradeoff is that the position resolves faster (good for capital recycling) and the theta decay per day is slightly higher (good for daily P&L accrual).

The practical consequence: with only 36 DTE at entry and a 21 DTE time stop, this trade had a maximum of 15 days to hit the 50% profit target. That’s a tight window. As it turned out, the trade was profitable but hadn’t reached the target when the time stop triggered. Whether the shorter DTE entry caused that or whether it would have played out the same at 45 DTE is unknowable on a single trade. But the deviation is logged.

How It Played Out

The position spent 15 days doing mostly nothing exciting, which is what you want from a short strangle. Corn bounced around in the 450s to 470s range. Neither strike was seriously threatened. The 440 put was far enough below the market that it decayed steadily. The 490 call had some pressure from the directional drift higher but never got close to being tested.

By May 1, at exactly 21 DTE, the position was showing $62.50 in gross profit. That’s 17.2% of the credit received, well below our 50% target but solidly in the green. The time stop rule is unambiguous: close at 21 DTE regardless. So we closed.

This is one of those trades that tests your discipline. The position was profitable and trending in the right direction. Theta was accelerating. IV was stable. Every instinct says “just give it another week, it’ll get to 50%.” And maybe it would have. But the 21 DTE time stop exists because gamma risk increases nonlinearly as you approach expiration, and the risk/reward of holding a short strangle inside 21 DTE deteriorates even when the position looks healthy. One surprise USDA report, one overnight gap on an Iran headline, and a comfortable winner can turn into a stop loss in hours rather than days.

The rule saved us from having to make that judgment call. We took the $48.32 net profit and freed up the margin for the next trade.

The Math on Time Stops

This is worth dwelling on for a moment, because it illustrates something important about the framework.

A $48.32 net profit on $1,309 in margin (doubled to $2,618 for our safety holdout) over 15 days works out to 0.32% return per day on risked capital. That’s actually better than our historical average of 0.19% per day across all strangle trades (including losers). So even though this trade didn’t hit the 50% profit target, it contributed positively to the portfolio at an above average daily rate.

The 50% target is the ideal outcome. The time stop is the “good enough” outcome. The 2x stop is the “take the loss and move on” outcome. Not every trade needs to be a home run. Consistent base hits compound.

Across 130+ historical strangle trades, roughly 25% are closed at the time stop rather than at the profit target or stop loss. The average P&L on time stop exits is positive but small (typically 10-25% of credit received). These trades don’t make the highlight reel, but they contribute to the overall positive expected value of the system by cycling capital efficiently and avoiding the gamma trap of holding too close to expiration.

Grade: B+

Profitable. Mechanically managed. No deviations on the exit side. The entry was slightly shorter DTE than preferred (36 vs 45), which compressed the window and likely contributed to not reaching the profit target. But the trade did what it was supposed to do: captured premium in a rangebound environment and exited before gamma risk became a problem.

The one criticism is the DTE choice. If a 45 DTE expiration had been available, we’d have had 24 days to reach the target instead of 15. That extra week and a half might have been the difference between a B+ and an A. Something to track going forward: does taking shorter DTE entries systematically result in more time stop exits? We’ll see as the sample builds.

Account Update

This is trade #7 overall and the second entry in the short strangle sleeve. Cumulative strangle sleeve performance on closed trades: 2 wins, 1 losses (though one was a full target hit and this one was a time stop exit). We’ll go into more detail in the April recap post.

The Canadian dollar strangle (trade #4) also closed on May 1, for a loss. We’ll cover that in the next recap. That one is going to change the running numbers. But that’s why we document everything.

Subscribe to Leptokurtic Capital for every trade recap as it happens, plus the framework posts that explain why the strategy works. The math, the fills, and the philosophy. No black boxes.

We are all students.

Let’s get into it.

The Trade

Underlying: /6B (British Pound Futures, September 2026)

Structure: Short strangle Strikes: 1.320 put / 1.385 call

Expiration: July 2, 2026 (59 DTE at entry)

Entry date: May 4, 2026

Premium received:

• Put: $293.75 (0.0047 × 62,500)

• Call: $268.75 (0.0043 × 62,500)

• Total credit: $562.50

Management rules (same as always):

• Profit target: 50% of credit received → close at ~$281.25 profit

• Stop loss: 2x credit received → close at ~$1,125 loss

• Time stop: close at 21 DTE regardless of P&L

Quick Return Math

Margin on /6B from the broker: ~$1,251.41 (note that the margin req in the screenshot does not include the credit, so it’s $1,813.91 - 562.50). Doubled for our safety margin holdout: ~$2,500.

$281.25 / $3,628 = ~7.8% on a successful position over ~25 days average hold (our historical average), or roughly 0.3% return per day on risked capital for a winner.

The credit on this one is larger than our typical strangle entry (the corn trade was $393.75, the Canadian dollar was $360). That’s partly because /6B options have a 62,500 unit multiplier and partly because sterling implied vol is running hot right now. More on that in a moment.

Why This Trade, Why Now

The British pound is in an interesting spot. It has actually been one of the stronger G10 currencies against the dollar over the past few months, which feels counterintuitive given that the UK economy is, to put it charitably, not great. But there are structural reasons for this, and those reasons are exactly what create the setup we want.

The Bank of England is stuck, and everyone knows it. The BoE held rates at 3.75% in April (8-1 vote, with one member actually wanting to hike). This is the same rate they’ve been at since February. The problem is familiar: CPI inflation has risen to 3.3% and is expected to go higher as energy costs from the Iran conflict feed through to households. The BoE’s own scenarios have inflation peaking anywhere from 3.5% in the benign case to 6.2% in the worst case. Meanwhile, GDP growth is running at maybe 0.1-0.2% quarterly and the labor market is softening.

Sound familiar? It should. This is the same stagflationary dynamic we talked about with the Canadian dollar. Central bank can’t cut because of inflation, can’t hike because the economy is already weak, so they sit on their hands and communicate in increasingly creative euphemisms for “we don’t know what’s going to happen either.”

The result is a currency that’s rangebound for macro reasons, not because nothing is happening. The pound is being supported by relatively high UK rates (3.75% is higher than the ECB’s 2.0%, so GBP gets carry inflows from European investors). It’s being capped by the growth picture and by the fact that the dollar is bid as a safe haven during the Middle East conflict. These forces roughly offset, which means cable (GBP/USD) has been chopping around rather than trending.

IVR at entry: 76.8. This is well above our preferred threshold for strangle entries. The options market is pricing in a level of movement that the underlying is unlikely to deliver if it stays in this macro equilibrium. That’s the variance risk premium doing its thing. Hedgers on both sides of sterling (exporters worried about GBP weakness, importers worried about GBP strength, banks hedging rate exposure in both directions) are bidding up implied vol simultaneously, and we’re selling it.

The 1.320/1.385 strikes give us a range of 6.5 cents on either side of the current 1.3617 price. That’s roughly a 4.5% move in either direction before we’re in trouble. For context, the pound hasn’t made a 4.5% monthly move against the dollar since the initial panic after the Iran conflict started. In a rangebound environment, 52 days of sitting between those strikes is the base case.

What the Framework Says

The leptokurtic thesis is clean here. Sterling is spending most of its time doing very little (tall middle of the distribution) while the options market prices in the possibility of a breakout driven by BoE policy surprise, further energy shock escalation, or a shift in the safe haven trade (fat implied tails). We’re selling the overpriced center.

FX strangles, particularly on major pairs like GBP/USD, are among the most consistent variance risk premium harvesters in the futures universe. This is one of the Tier 1 underlyings from our historical analysis. EUR/USD, GBP/USD, and JPY/USD have all shown strong profit factors in the strangle framework over many trades, specifically because they attract hedging flow from both sides in uncertain macro environments.

Adding British pound diversifies the strangle portfolio further. We now have (or have had) positions across grains (/ZC), currencies (/6C, /6B), and bonds (/ZB). The return correlation between a corn strangle and a sterling strangle is essentially zero in normal environments. That’s the decorrelation benefit working as designed.

A Note on Theta

One thing worth pointing out for newer readers: look at the theta on this position. The strangle is decaying at $10.80 per day at current levels ($4.92 from the put, $5.88 from the call). That means every day the pound sits between our strikes and vol doesn’t expand, we’re making roughly $10.80. Over 52 days, that’s $561.60 of potential theta decay, which is (not coincidentally) almost exactly equal to the premium we collected.

Of course, theta isn’t linear. It accelerates as expiration approaches, and it can be overwhelmed by delta or vega moves in the meantime. But the daily accrual gives you a sense of why these trades work when the underlying cooperates. Time is literally on our side.

What I’m Watching

The June 18 Bank of England decision. This is the next scheduled MPC meeting. We entered on May 4, so our 21 DTE time stop hits around June 11. That means if we haven’t already hit the 50% profit target by then, we’re out a full week before the BoE announcement. The timing works in our favor here. We don’t have to make a discretionary call about whether to hold through a central bank event because the management rules get us out before it matters.

UK CPI data. The next inflation print will be closely watched given the BoE’s explicit concern about second round effects from energy prices. A hot number could shift rate expectations toward a hike, which would strengthen GBP. A miss could reignite cut expectations. Either way, a large surprise in either direction creates the kind of directional move that could test our strikes. The management rules handle this mechanically.

Middle East de-escalation or escalation. Any shift in the Iran conflict affects sterling through two channels: the energy cost channel (UK is a net energy importer, so higher oil hurts the economy and weakens GBP) and the safe haven channel (conflict resolution weakens USD, which strengthens GBP). These two channels partially offset each other, which is actually what we want. It means even if the macro picture shifts, both sides of the strangle don’t get hit simultaneously.

Dollar strength broadly. If the DXY breaks higher on further safe haven flows, that’s directionally bearish for GBP/USD and could pressure the put side. Our 1.320 put strike is a long way from current levels, but worth watching. The call side at 1.385 gets tested if there’s a sudden “risk on” move that takes the dollar down.

Current Status

The position is 7 days old and showing +$118.75 in unrealized P&L, up $37.50 on today’s session alone. The put side is doing the heavy lifting at +$131.25 (the 1.320 strike is far enough below the current 1.3617 that it’s decaying nicely), while the call side is slightly underwater at -$12.50 (1.385 is closer to the money, so that leg hasn’t decayed as much and any upward drift in GBP pressures it a bit).

That puts us at 21.1% of the way to the 50% profit target after 7 of the roughly 25 days we’d expect to hold this. The theta is running at $10.80/day and the position is behaving exactly as the framework predicts: time passing, vol not expanding, premium decaying. Nothing to manage yet. We wait.

Account Update

This is trade #8 overall and the fourth entry in the short strangle sleeve (following corn(s) and Canadian dollar).

More trade update posts coming this week, as well as the April monthly recaps.

Subscribe to Leptokurtic Capital for every trade recap as it happens, plus the framework posts that explain why the strategy works. The math, the fills, and the philosophy. No black boxes.

We are all students.

Hello readers! Apologies for my first delay in posting. Spring pollen allergies and a cold hit me at the same time! Back at it though, and there will be some more trade recaps coming up soon.

Most of what we do on Leptokurtic Capital assumes you already know what a short strangle is, why calendar spreads have theta differentials, and what a 20 delta option looks like on a risk graph. That’s fine for the readers who come in with options experience, but I’ve gotten enough messages from newer subscribers asking “wait, what exactly is a put?” that it’s worth slowing down and starting from the beginning.

This is the first in a short series of introductory posts. It will not teach you everything. It will teach you the four fundamental trades every options trader needs to understand before anything else makes sense, and hopefully give you enough scaffolding to read the rest of the publication without constantly hitting terms you don’t recognize.

One caveat up front: I’m going to simplify things in places. Options pricing is actually governed by a partial differential equation and the behavior of real options deviates from the textbook picture in ways that matter a lot once you’re trading them. But you can’t learn the exceptions before you learn the rule, so let’s start with the rule.

What an Option Actually Is

An option is a contract. That’s it. It’s an agreement between two people (or more accurately, two brokerage accounts) where one side pays the other side some money today in exchange for the right to do something later.

There are two kinds of options.

A call option gives the buyer the right to BUY 100 shares of the underlying at a specific price (the “strike”) on or before a specific date (the “expiration”).

A put option gives the buyer the right to SELL 100 shares of the underlying at the strike price on or before expiration.

That’s really all there is to the definitions. Everything else flows from these two sentences.

The price the buyer pays for this right is called the premium. The seller of the option collects the premium upfront and in exchange takes on the obligation to honor the contract if the buyer chooses to exercise it.

A quick note on the 100 shares thing: one option contract typically represents 100 shares of the underlying stock. If a call has a premium of $2.50, you’re actually paying $250 to own it ($2.50 × 100). This trips up a lot of beginners. The quoted price is per share but the contract covers 100 shares. Futures options work similarly but the multiplier depends on the contract.

The Four Trades

Because there are two types of options (call and put) and two things you can do with them (buy or sell), there are exactly four fundamental positions. Every complex options strategy in existence is just combinations of these four.

1. Buy a Call (Long Call)

You pay a premium. In exchange, you have the right to buy 100 shares at the strike price until expiration.

You make money when: The underlying goes up enough that the right to buy at the strike is worth more than what you paid for it.

You lose money when: The underlying stays flat or goes down, or doesn’t go up enough to overcome the premium you paid.

Maximum loss: The premium you paid. This is capped. If you spend $250 on a call and the underlying crashes, the worst that happens is your call expires worthless and you lose $250.

Maximum gain: Theoretically unlimited. If the underlying triples, the call can be worth many multiples of what you paid.

This is the trade most beginners start with because it’s the easiest to understand intuitively. You think something is going up, you buy a call, you profit if you’re right. The psychology matches owning shares but with leverage.

2. Buy a Put (Long Put)

You pay a premium. In exchange, you have the right to sell 100 shares at the strike price until expiration.

You make money when: The underlying goes down enough that the right to sell at the strike is worth more than what you paid.

You lose money when: The underlying stays flat or goes up.

Maximum loss: The premium you paid.

Maximum gain: Very large but not technically unlimited (the underlying can only go to zero).

This is how you bet on something going down without shorting shares. It’s also how you buy insurance on a long stock position (often called a “protective put”). And it’s one of the two main structures we use in Sleeve 3 Explosion Positions, where we buy deep out of the money puts on futures that our Convexity Score ranks as having mispriced tail risk.

3. Sell a Call (Short Call)

You collect the premium upfront. In exchange, you take on the obligation to sell 100 shares at the strike price if the buyer chooses to exercise.

You make money when: The underlying stays below the strike price until expiration, and the call expires worthless. You keep the premium.

You lose money when: The underlying rises above the strike price. Your losses grow as the underlying keeps rising.

Maximum gain: The premium you collected. Also capped.

Maximum loss: Theoretically unlimited. If you sell a naked call and the underlying gaps up 50% overnight, you can lose dramatically more than what you collected.

This is where it starts getting dangerous for beginners. Selling naked calls (without owning the underlying shares) has unlimited loss potential, and because the maximum gain is the premium received, the risk/reward looks terrible on paper. The reason people do it anyway is that the probability of winning is typically much higher than 50%, which changes the math.

4. Sell a Put (Short Put)

You collect the premium upfront. In exchange, you take on the obligation to BUY 100 shares at the strike price if the buyer chooses to exercise.

You make money when: The underlying stays above the strike price until expiration, and the put expires worthless.

You lose money when: The underlying falls below the strike price.

Maximum gain: The premium collected.

Maximum loss: Very large but capped at the strike price (the underlying can only go to zero, at which point you’d be assigned 100 shares worth nothing when you agreed to pay the strike).

Selling puts is the most common way people first interact with the short side of options, because the math feels more manageable than short calls and because the assignment outcome (being forced to buy shares you might want to own anyway) is psychologically palatable.

Why Trade Options Instead of Shares or Futures Directly?

This is the real question, and honestly the answer is nuanced enough that a lot of people shouldn’t bother.

Leverage without margin calls (for buyers). A long call gives you exposure to upside without tying up the capital required to buy 100 shares. If a stock is at $100 and you think it’s going to $120, you could spend $10,000 on 100 shares to make $2,000 on a $20 move. Or you could spend $500 on an at the money call and potentially make $1,500-2,000 on the same move if it happens quickly enough. The leverage is real, but the tradeoff is that if the stock stays flat, you lose the $500 while the shareholder loses nothing.

Defined risk. When you buy an option, you know your maximum loss the moment you enter the trade. It’s the premium you paid. This is genuinely useful for position sizing and risk management, especially compared to futures where adverse moves can blow through stops and cause losses larger than you planned for.

Probability based income (for sellers). This is the less obvious one, and it’s what most of Leptokurtic Capital is about. When you sell an option, you’re not betting on direction. You’re betting that the underlying will stay within a certain range. Because of the variance risk premium (implied volatility is systematically higher than realized volatility), selling options at high probability strikes has a structural edge over time. Not on every trade, but over hundreds of trades.

Expressing views that shares can’t express. “I think the stock will stay between $95 and $105 for the next month” is not a view you can trade with shares. You can trade it with a short strangle. “I think volatility will increase but I don’t know which direction” is another view that has no cash equity equivalent. Long straddles and calendars handle this.

Tail exposure. Deep out of the money options are the only way to get cheap exposure to rare, extreme outcomes. If you think the market is underpricing the probability of a 20% monthly move, you can’t express that with shares. You express it with 5 delta options that cost almost nothing and pay off enormously if the tail event happens. This is Sleeve 3 in a nutshell.

The downsides are real, though. Options have time decay working against buyers. Bid-ask spreads are wider than equities, especially on less liquid names, which eats into edge. The Greeks (delta, gamma, theta, vega) add complexity that most people underestimate until they’ve been hurt by it. Assignment risk is a real thing that catches people off guard. And the learning curve is steep enough that many people lose money for years before figuring out what they’re actually doing.

If you have a simple directional view on a stock and don’t need leverage, just buy the stock. Options are a tool for specific jobs. Using them when a simpler instrument would work is how people overcomplicate their way into losses.

Some Things I Wish I’d Known Earlier

Liquidity matters more than anything else when you’re starting out. A strategy that works beautifully on paper can be destroyed by wide bid-ask spreads. Before you trade any options, look at the open interest and volume on the specific strikes and expirations you’re considering. If the bid-ask is more than 5-10% of the option’s price, you’re already at a significant disadvantage before the trade even starts. This is why the big names (SPY, QQQ, AAPL, major futures contracts) are where most people should start. The pennies you save on spreads compound faster than you’d think.

Implied volatility is the entire game, and most beginners don’t know it exists. When you buy an option, you’re not just betting on direction. You’re also betting on what volatility does between now and expiration. If you buy a call because you think a stock is going up and it goes up slowly while IV collapses, you can be directionally correct and still lose money. Learning to read IV and IV rank should be one of the first things you focus on after understanding the four basic trades.

The probability of profit on a trade is not the same as the expected value. A short option trade with an 85% probability of profit sounds great until you realize the 15% loss is five times larger than the 85% win. High win rate strategies need careful loss management because one big loss eats several wins. Conversely, low win rate strategies (like buying tail options) can be extremely profitable if the rare wins are large enough. Evaluate trades on expected value, not on win rate alone.

Mechanical management beats discretionary management for most people. The hardest thing about options trading is not the analysis. It’s managing your own psychology when a position moves against you. Pre-committing to rules like “close at 50% profit, exit at 2x stop loss, time stop at 21 DTE” removes most of the emotional decision making. It sounds boring because it is boring. Boring is how you survive long enough to get good.

Position sizing will save or kill you long before your analysis does. A perfect trade system with terrible position sizing will blow up. A mediocre trade system with conservative position sizing will survive to compound. When you’re starting out, size smaller than feels right. Then size smaller than that. The math of compound returns rewards survival far more than it rewards aggression.

Don’t trade earnings if you don’t know what you’re doing. Earnings releases compress years of directional movement into a single overnight event, and the options market knows this. IV gets bid up beforehand and collapses afterward (the “volatility crush”). You can be right about direction and still lose money because the IV crush was larger than the price move. Until you understand why, stay away from earnings plays. When you do understand why, you’ll probably still stay away from most of them.

Where to Go From Here

The four trades above are the atoms. Everything else is molecules. A covered call is a long stock plus a short call. A credit spread is a short option plus a long option at a different strike. A strangle is a short call plus a short put. A calendar is a short option in one expiration plus a long option in another. Once you understand the four base trades, every “complicated” strategy is just figuring out which base trades are stacked together and why.

The framework posts on this publication (the Sleeve 1, 2, and 3 posts in particular) assume you understand these basics. If you’re reading this and some of it still feels fuzzy, that’s fine. Read it again. Trade tiny sizes to feel how the pieces actually move. Paper trade if you must, but know that paper trading teaches you the mechanics without teaching you the psychology, and the psychology is 80% of the game.

There will be more introductory posts in this series. Next up is probably implied volatility and the Greeks, because those are the concepts that most beginners skip over and then spend years being confused about. After that, probably the actual mechanics of opening, closing, and getting assigned on real trades. If there are specific beginner topics you want me to cover, hit reply and let me know.

One last thing. Options are genuinely useful tools, but they are also the single most effective way to lose money quickly if you don’t know what you’re doing. Take your time. Trade small. Read everything twice. The market will still be here in six months when you actually understand what you’re looking at.

If this post was useful, the first few posts walk through the three sleeve portfolio approach that everything on this publication is built around. If it was too basic, never fear, futures posts will get back to the usual depth. We are all students.

This is the second double calendar in the account and the sixth trade overall. Trade #3 was our first calendar (also on /MES), which gave us a feel for how the structure behaves in real time. This one is a cleaner setup in some ways and a messier one in others.

Let’s get into it.

The Trade

Underlying: /MES (Micro E-mini S&P 500 Futures) Structure: Double calendar spread Entry date: April 6, 2026

Put side:

• Sold: April 24 /MES 6450 put

• Bought: April 30 /MES 6450 put

Call side:

• Sold: April 24 /MES 6770 call

• Bought: April 30 /MES 6770 call

Total debit paid: $317.50

Same strikes across both expirations. This is a true double calendar (no diagonal adjustments like the slight strike offset on Trade #3’s put side). Six calendar days between the short and long legs.

Management rules (same as always):

• Profit target: 25% return on debit → close at ~$397 total value (~$79 profit)

• Stop loss: 50% loss on debit → close at ~$159 total value (~$159 loss)

• Time stop: close entire structure when short leg reaches 3 DTE (April 21)

Why /MES Again

I’ll be honest about this: /MES is not the ideal underlying for the calendar sleeve from a diversification standpoint.

The framework calls for calendars on low IVR underlyings across different sectors. You’d want equity indices, maybe a low vol currency, maybe gold during a quiet stretch. Spreading the calendar sleeve across uncorrelated names, just like we spread the strangle sleeve across different futures.

In practice, the liquidity just isn’t there for many futures options at the expiration dates I need. Calendar spreads require matching specific expiration dates on both legs, and the tighter the structure (short gap between expirations), the more dependent you are on having liquid weekly or semi-monthly expirations available. Most futures options outside of /ES and /MES don’t offer that kind of granularity.

So most calendars going forward will probably be on /MES. That’s a concentration I’d rather not have, and I want to be transparent about it. The system would function better with more underlying diversification in this sleeve. But running a well-structured calendar on a liquid product beats forcing an awkward calendar on an illiquid one where the bid/ask spread eats your edge.

This is one of those small account realities. At $100K you have more flexibility. At $10K, you work with what gives you clean fills.

The Setup

IVR on the S&P is still elevated. We’ve had the Strait of Hormuz whipsaw playing out in real time over the past week: escalation fears spike vol, diplomatic rumors compress it, then another headline spikes it again. The VIX has been oscillating rather than trending in one direction.

For a calendar, this is actually a reasonable environment. Here’s why.

The calendar doesn’t need vol to collapse or to spike. It needs time to pass. The theta differential between the short and long legs generates profit as long as the underlying stays roughly in range and vol doesn’t crater.

And I don’t see a strong IV contraction coming soon. The geopolitical situation is unresolved. The Iran conflict is the kind of event that keeps a floor under implied vol even on days where nothing new happens, because traders know something new could happen overnight. That baseline level of uncertainty is actually helpful for the calendar: it keeps the long-dated options we own from losing vega value, while the short-dated options we sold still decay on schedule.

If direction stays neutral (the S&P continues to chop rather than trend), we win via theta. If vol expands on an escalation, we win via vega on the long legs. The scenario that hurts is a sharp directional move that blows through one side of the calendar, or a sudden vol collapse from a peace deal. Both are possible. That’s why we have the 50% stop.

Current Status (as of posting)

The position is down approximately $50 from entry, which puts us at about a 16% unrealized loss on the debit. This is within normal noise for a calendar that’s been open for a couple of days in a volatile market.

What happened: the Strait of Hormuz headlines produced a sharp intraday move on /MES followed by a reversal. The kind of back and forth whipsaw that makes calendars uncomfortable in the short term (large moves disrupt the structure temporarily) but doesn’t necessarily kill the trade. As long as /MES settles back into the range by expiration, the theta differential reasserts itself.

I’m watching this heading into the weekend. A gap open on Monday driven by overnight geopolitical news could push us toward the 50% stop. That’s a real possibility and I want to name it rather than pretend it isn’t there.

And that’s ok. Not every trade will be a winner. If this one stops out, we take the loss, record it, and move to the next cycle. The system has a tested edge over many trades. Any individual trade can lose. The discipline is in following the rules regardless of what this particular entry does, and tracking whether the losses are within the expected parameters of the strategy.

If I close this for a 50% loss, that’s $159 gone. Unpleasant, but it’s defined risk. That’s the whole point of the calendar structure. I know my worst case walking in.

What I’m Watching

Weekend risk. This is the immediate concern. Geopolitical headlines don’t respect market hours. A Saturday escalation in the Middle East could produce a gap open on Sunday evening (futures open at 6pm ET) that puts immediate pressure on the position. I’ll be checking the overnight session and making a management decision early Monday if needed.

The 6450 to 6770 range. This is approximately a 320 point window, or about 5% of the underlying price. /MES needs to stay within this zone for both calendars to maintain their favorable structure. The S&P moving 5% in 18 days (our time to the short leg close) is unusual but not impossible given current conditions. For reference, the S&P has moved more than 5% in a two week period several times since the Iran conflict began.

The April 21 time stop. The short legs expire April 24. We close when they reach 3 DTE, which is April 21. That gives us 15 days for the trade to work. Either we hit the 25% profit target before then, hit the 50% stop, or we close at the time stop and take whatever the P&L is at that point.

Theta acceleration. With only 6 calendar days between the short and long expirations, the theta differential is modest per day but accelerates as the short leg approaches expiration. The best theta capture should come in the final week (April 14 to 21) when the short leg’s decay curve steepens. We just need to survive until then.

How This Fits the Portfolio

Open positions as of this trade:

• /ZB explosion puts (Trade #2, Sleeve 3)

• /6C short strangle (Trade #4, Sleeve 1)

• /NG explosion position (Trade #5, Sleeve 3)

• /MES double calendar (Trade #6, Sleeve 2)

Four active trades across three sleeves. The strangle is short vega, the calendar is long vega, the explosion positions are long convexity. A vol spike from here hurts the strangle, helps the calendar, and potentially activates the explosion positions. A vol collapse helps the strangle, hurts the calendar, and kills the explosion positions. A quiet range bound market is best for the strangle and calendar while the explosion positions bleed.

No scenario wipes all four. That’s the portfolio construction working as designed, even at this early stage with only a handful of positions.

A Note on Losses

I said above that this trade might stop out over the weekend. I want to expand on that for a moment because I think it’s important for what this publication is about.

There will be losing trades. Many of them, if we’re doing this right. The strangle sleeve historically wins about 84% of the time, which means roughly 1 in 6 trades loses. The calendar sleeve will likely have a lower win rate than that given the narrower profit zone. The explosion positions lose most of the time by design.

If every trade recap I publish is a winner, something is wrong. Either I’m cherry picking what I show you (I won’t) or the sample is too small to be meaningful (which it still is at N=6).

The value of a system isn’t that it never loses. It’s that the losses are defined, expected, and small enough relative to the wins that the trades are +EV and the portfolio compounds over time. Trade #6 might be the first public loss. If it is, it’ll be documented with the same detail as the wins. That’s the commitment.

Subscribe to follow every trade in real time. The wins, the losses, and the boring ones in between. That’s where the real education lives.

We are all students.

This is the second Explosion Position in the live account. Trade #2 (the /ZB treasury bond puts) was our first foray into Sleeve 3. This one targets a different asset class entirely, and it’s the one our screening framework was practically designed for.

If you’ve read the framework posts, you know natural gas holds a special place in the leptokurtosis data. If you haven’t, this trade will show you why.

Let’s get into it.

The Trade

Underlying: /NG (Natural Gas Futures)

Structure: Explosion Position (deep OTM call + deep OTM put, 5 delta each on entry)

Strikes: 2.30 put / 5.50 call Expiration: 79 DTE at entry Entry date: April 7, 2026

Premium paid:

• Total debit: $390.00

Management rules (Explosion Position standard):

• Losers: let them expire. No salvage attempts.

• Winners: Ideally we would take profits in thirds (1/3 at 3x, 1/3 at 5-8x, final 1/3 rides with no target), but in the small account, we’ll take the middle road and manage at 5-8x until we scale up.

• No delta adjustments. No spreading against winners. No “locking in.”

Why Natural Gas, Why Now

Let me start with the structural case, because it’s the stronger one. Then I’ll layer in the timing.

Natural gas is the single most leptokurtic asset in our 18 futures universe.

When I built the Tail Richness Ratio screen, natural gas wasn’t just at the top of the list. It was in a different category. Over the past 15 years, /NG has produced 3-sigma monthly moves roughly 20 times more frequently than a normal distribution predicts. At the 4-sigma level, where a Gaussian model says you should see essentially zero occurrences in 15 years, natural gas had 5.

Five impossible events in 15 years. That’s natural gas.

The reasons are structural. Natural gas demand is weather dependent and inelastic (you can’t decide to heat your house less when it’s negative 20). Supply is pipeline constrained and can’t ramp quickly. Storage is the buffer, and when storage gets tight, the market reprices violently. Add in the growing LNG export infrastructure creating a link between US domestic prices and global energy markets, and you have an asset that is fundamentally prone to explosive moves in both directions.

This is also why we DON’T sell strangles on natural gas. The same tail frequency that makes it a terrible strangle underlying makes it an excellent Explosion Position candidate. The tails are fat. The options market underprices that fatness (relative to empirical frequency). We buy instead of sell.

That last point is worth sitting with. The same distributional observation (leptokurtosis) tells you what to sell and what to buy. On assets where the center is overpriced and the tails are manageable, you sell strangles. On assets where the tails are dramatically fatter than priced, you buy Explosion Positions. Same thesis, opposite expression. Natural gas sits firmly on the “buy the tails” side.

Now the timing.

IVR on /NG at entry was approximately 25. That’s low. Implied vol has been compressing through late March as US production has been strong (lower 48 output around 107 bcfd), weather has turned seasonal, and the domestic supply picture looks comfortable on the surface.

Low IVR on an Explosion Position is what we want. It means the options are cheap. Our debit is small relative to the potential payoff. The market is not pricing in a spike, which is exactly when the spike tends to happen (natural gas has a nasty habit of going from “everything is fine” to “everything is on fire” in about 72 hours, usually on a storage report or a weather model flip).

The Iran catalyst, and why I chose 79 DTE.

The standard Explosion Position duration is 30 to 45 DTE. This one is at 79 DTE, which is a deliberate deviation. Here’s the reasoning.

The available expirations near 45 DTE didn’t give me the coverage window I wanted. The Iran conflict has shut down approximately 20% of global LNG supply through the Strait of Hormuz closure. Qatar’s Ras Laffan complex (one of the world’s largest LNG hubs) has been damaged. QatarEnergy declared force majeure on its contracts. This is the largest supply disruption in global energy history, per the IEA.

But (and this is the interesting part for our purposes) the disruption hasn’t fully transmitted to US Henry Hub prices yet. US natural gas is somewhat insulated because the US is largely self sufficient in production and has pipeline supply that doesn’t transit the Strait. Henry Hub has been trading around $3/MMBtu while global LNG benchmarks have spiked.

The question is whether that insulation holds. If the conflict extends through summer, US LNG export facilities will face a tug of war: Asian and European buyers desperately bidding for any available cargo versus domestic demand. If export flows increase to fill the global gap, domestic supply tightens. If the conflict disrupts additional infrastructure or spreads, the energy complex reprices more broadly. Either scenario could produce the kind of violent move that defines natural gas.

79 DTE gives me coverage through the next 2 to 3 months of this conflict’s evolution. It captures the summer cooling season (air conditioning demand peaks in June/July, which overlaps with our expiration window). And it covers the April 29 Bank of Canada meeting, multiple OPEC decisions, and whatever geopolitical escalation or de-escalation happens in between.

I’m paying for the extra time. 79 DTE options cost more than 45 DTE options. But the theta decay on deep OTM options at 79 DTE is slow (the decay curve is flat that far out), so I’m not bleeding aggressively. And the coverage window captures the period where I believe the probability of an extreme move is highest.

The Strikes

2.30 put. Natural gas is currently around $3.00. A move to $2.30 would be a roughly 23% decline. This captures a scenario where the Iran conflict de-escalates faster than expected, global LNG flows normalize, and the domestic glut that was building pre-conflict reasserts itself. It also captures a broader economic slowdown (remember, Canada and parts of Europe are flirting with recession) that craters industrial demand for gas.

5.50 call. A move to $5.50 would be roughly an 83% increase from current levels. Extreme? Not for natural gas. /NG traded above $7.80 in the past year alone (January 2026 polar vortex spike). In 2022, it hit $10. The 52 week range on this contract spans from $2.62 to $7.83. A move to $5.50 is well within the historical envelope.

Both strikes are deep OTM but within the range of moves that natural gas has actually made, multiple times, in the recent past. This is the difference between buying convexity on natural gas versus, say, gold. A 5-delta gold call is buying a move that gold almost never makes. A 5-delta natural gas call is buying a move that gas makes every couple of years.

What the Framework Says

The Convexity Score on /NG ranked near the top of the universe this month, as it usually does. The combination of high Tail Richness Ratio (fatter tails than almost any other liquid futures contract), low IVR (cheap options relative to own history), and a live geopolitical catalyst that hasn’t fully priced into the domestic market makes this about as clean an Explosion Position setup as the framework produces.

The direction agnostic structure (owning both calls and puts) is doing real work here. I have a mild directional lean toward the call side given the global supply disruption, but I have been wrong about direction before and I will be wrong again. The put protects against the scenario I haven’t thought of. Maybe a massive SPR release craters prices. Maybe warm weather plus de-escalation creates a perfect storm to the downside. I don’t need to predict. I need to own the tails.

How This Fits the Portfolio

The portfolio is developing its cross-asset character. We have a currency strangle, a bond explosion position, an equity calendar, and now an energy explosion position. The correlations between these in normal markets are low. In a crisis, the explosion positions are designed to produce outsized returns that offset strangle and calendar volatility and directional exposures.

Natural gas adds something specific: sensitivity to supply shocks. The /ZB explosion position is a rates/macro play. The /NG explosion position is an energy/supply play. Different catalysts, different markets, different tail triggers. That’s the diversification working as designed.

What I’m Watching

Weekly EIA storage reports (every Thursday). This is the single biggest short-term catalyst for /NG. A surprise draw can move the market 5 to 10% in a session. A series of draws can compound into the kind of sustained move that takes Explosion Positions from “slightly interesting” to “very profitable.” Conversely, a surprise build would confirm the domestic glut thesis and probably send us toward the put side.

LNG export flow data. If exports ramp significantly as global buyers scramble for non-Qatari supply, that’s bullish for Henry Hub and call-side positive. The current record is about 18.5 bcfd. A move above 20 bcfd would signal meaningful tightening.

Iran conflict trajectory. The obvious one. Escalation or geographic spread is energy bullish. A diplomatic breakthrough craters the energy complex. We own both sides, so either extreme benefits us. What hurts is the middle: conflict simmers but doesn’t escalate, LNG flows stay disrupted but US stays insulated, nothing dramatic happens. That’s the scenario where the options decay to zero. Which, to be clear, is the most likely outcome for any individual Explosion Position. These positions lose money more often than they make it. The edge is in the magnitude of the wins when they hit.

Weather models. It’s early April, so the market’s attention is shifting from heating demand to cooling demand projections. A hotter than expected summer forecast for the eastern US would be bullish for gas. An early start to hurricane season (which can disrupt Gulf of Mexico production) would add to tail risk. Again, not predicting. Just mapping the catalyst landscape.

Account Update

This is Explosion Position #2 and Trade #5 overall. $390 at risk, which is the expected monthly bleed from this sleeve. If both explosion positions expire worthless, that’s roughly $800 in tail convexity purchased this cycle. That’s the cost of being positioned for the move that doesn’t usually come, but comes more often than people think.

Cumulative by sleeve: Strangles 1/1 closed, +$185.16 net. Calendars and Explosion Positions still developing.

Subscribe for every trade, every fill, every outcome. The explosion positions are the most volatile part of the portfolio and the most interesting to follow in real time.

We are all students.

I recently read a piece by Antoine Guillon (Tale of Two Tails on Substack) about what a convexity strategy actually looks like in practice. It’s one of the better things I’ve read on the subject, and I want to engage with it directly because I think his framework and ours are solving the same problem from opposite directions. Where those directions converge, and where they diverge, is where the interesting stuff lives.

If you’ve been following Leptokurticapital, you know we buy tail convexity as one of three portfolio sleeves. But I haven’t spent enough time on a question that Guillon’s piece forced me to think about more carefully: is there a better way to structure the convexity than simply buying 5-delta options and waiting?

Almost certainly. Let me walk through it.

The Sharpe Ratio Trap (We’re in It)

Guillon’s opening argument is something we should take personally. He points out that a large part of the investment industry optimizes for Sharpe ratio, and that this optimization often incentivizes selling convexity. Strategies that sell premium produce smooth, consistent returns with low measured volatility. They look great on a risk-adjusted basis. Until they don’t.

Sound familiar? Our strangle sleeve is exactly this trade. We sell the overpriced center of the distribution and collect steady premium. The Sharpe ratio on 130+ strangle trades looks good. The win rate is 83%. The equity curve, in normal times, is smooth and ascending.

We know this. We’ve talked about it extensively. The whole reason the Explosion Position sleeve exists is to own the other side of this payoff. We are simultaneously the seller of convexity (strangles) and the buyer of convexity (Explosion Positions). The question Guillon raises implicitly is whether we’re buying it intelligently or just buying it.

The Institutional Approach: Structured Convexity

Guillon’s preferred vehicle is the broken wing put butterfly. Without getting deep into the mechanics (his post does a good job of that), the basic idea is: instead of buying a naked put and bleeding theta every day, you construct a multi-leg position that uses the vol surface itself to finance part of the convexity.

You sell options at the richer part of the surface (typically around the short middle strike where IV is highest relative to realized probability) and use that premium to subsidize the wings where your actual convexity lives. The result is a position that can have positive theta at inception, meaning it actually makes money from time passing, while still maintaining convex exposure to large downside moves.

This is elegant. And it addresses the single biggest criticism of our Explosion Position approach: cost.

Our 5-delta options expire worthless most months. That’s by design. But the bleed is real. If we’re spending $X per month on Explosion Positions across a few underlyings, we need a tail event roughly every 6-12 months with a 10x+ payoff to break even on the sleeve. The Tail Richness Ratio data suggests this is positive EV over time. But the path, as Guillon correctly emphasizes, is months of small losses punctuated by occasional large gains. That path is psychologically difficult and capital-intensive.

A structured approach that self-funds some of the convexity cost through vol surface arbitrage is, in theory, a better mousetrap. You get the tail exposure with less bleed. Your breakeven frequency is lower. You can maintain the positions longer without the cost dragging on performance.

Where I Think the Structured Approach Has a Blind Spot

Guillon’s framework is built entirely around equity index options (he doesn’t specify the underlying, but the context, liquidity requirements, and put-focused structure all point to SPX or similar). This is the institutional default. It’s where the liquidity is deepest, the vol surface is most developed, and the structured products are most tradeable.

But this is also where tail convexity is most expensive.

We’ve covered this in previous posts. The institutional hedging ecosystem creates persistent, price-insensitive demand for equity downside protection. Every pension fund, every insurance company, every risk-managed portfolio is buying SPX puts. This flow keeps the equity put skew steep and the wings expensive. Even a cleverly structured butterfly on SPX is still buying convexity in the most expensive marketplace.

Our framework takes a different approach. Instead of designing more efficient structures on the same (expensive) underlying, we go shopping in less crowded markets. The Convexity Score screens 18 futures across 6 asset classes for where the raw convexity per dollar is cheapest. And the data consistently shows that commodity and currency tail options are dramatically cheaper than equity tail options, relative to the actual frequency of tail events.

Wheat doesn’t have a pension fund complex buying 5-delta puts every quarter. The yen doesn’t have an institutional hedging mandate inflating the wing IV. The options in these markets are priced primarily by market makers using models, not by price-insensitive hedgers competing for protection. This structural difference in the options ecosystem is, I believe, a larger source of edge than any structural optimization you can achieve within a single underlying.

Put differently: Guillon is optimizing the structure. We’re optimizing the selection. Both matter. But if forced to choose between a brilliantly structured butterfly on SPX and a simple 5-delta strangle on the asset with the highest Convexity Score, I’d take the strangle. The selection edge, buying where the mispricing is largest, dominates the structural edge of a more efficient payoff on an expensive underlying.

The ideal, of course, is both: the right structure on the right underlying. That’s where this research leads eventually.

Path Dependence Is Real and We Need to Talk About It

The most valuable insight in Guillon’s piece, for our purposes, is his discussion of path dependence. He describes how during the August 2024 VIX spike, his convexity structure actually lost money initially because the rapid vol expansion shifted the underlying into the “valley of death” (the vulnerable zone around the short strike of his butterfly) before the tail payoff could materialize.

This is important. A convex position can lose money during a tail event if the path of the move and the vol dynamics conspire to put you in the worst part of your payoff profile at the worst time.

Our framework is less exposed to this specific problem because we’re using simpler structures (naked long options rather than multi-leg butterflies). A 5-delta call on wheat doesn’t have a “valley of death.” If wheat goes up, the call makes money. If wheat goes down, the call loses its premium and that’s it. The payoff is monotonic: more movement in your direction always means more profit. There’s no intermediate zone where the position works against you.

But we have our own path dependence problem, and it’s duration. If we buy a 30 DTE option and the tail event happens on day 35, we missed it entirely. I wrote about this in the oil post: a 7 DTE option would have caught the first day of the Iran war and little else. A 45 DTE option captured the whole move. The timing of entry relative to the event is our version of Guillon’s path dependence.

This is an argument for both longer duration (more coverage window) and rolling positions (always having live exposure). We currently buy monthly and let them expire or pay off. A rolling structure that maintains continuous exposure, perhaps with staggered entries at different DTEs, would reduce the timing risk. This is something I’m actively thinking about and haven’t implemented yet.

Convexity as a Design Problem

Guillon closes with a line I find genuinely compelling: “Convexity is not something you buy. It is something you design.”

I agree with the spirit of this. And I think our framework is a design, just a different kind of design than what he’s describing.

His design optimizes the structure of the payoff (using vol surface relationships to self-fund the wings). Ours optimizes the selection of the underlying (using cross-asset tail frequency data to find where convexity is cheapest). His approach is elegant within a single, liquid market. Ours is broader but cruder in structure.

The next evolution of the framework, I think, involves combining both insights. What if we screened for the cheapest convexity across 18 futures (our current approach) AND used structured positions rather than naked options to implement the exposure on the top-ranked assets?

The practical challenge is liquidity. Broken wing butterflies require multiple strikes with reasonable bid-ask spreads. SPX has this. Wheat options at the 5-delta level might not. The very markets where the convexity is cheapest are often the markets where the options chains are thinnest. This is probably not a coincidence. The mispricing exists partly because the market microstructure doesn’t support efficient arbitrage.

So for now, I think the right approach is: simple structures on underpriced underlyings, with the structural optimization as a future research direction for the more liquid corners of the universe. Crude oil and the yen, where both the Convexity Score and the options liquidity are favorable, might be the first places to experiment with more structured approaches.

The Takeaway

Reading Guillon’s piece clarified something I’d been feeling but hadn’t articulated: our Explosion Position framework is a selection strategy disguised as a convexity strategy. The edge isn’t primarily in the structure of the trade. It’s in the choice of where to trade.

That’s a genuine differentiator. Most of the convexity literature (Taleb, Spitznagel, and now Guillon) focuses on equities. The cross-asset dimension, the idea that you should be buying convexity wherever it’s cheapest rather than wherever it’s most liquid, is underexplored. That’s where I think Leptokurticapital adds something to the conversation.

But the structural insights matter too. We’re leaving money on the table by using the simplest possible structure (naked long options) on every underlying. As the framework matures and the data accumulates, designing more efficient convexity structures on our top-ranked assets is the natural next step.

For now: screen first, structure second. Find where the tails are fattest and cheapest. Then worry about the most efficient way to own them.

If you haven’t read Guillon’s piece, I’d recommend it. It’s a different angle on the same problem, and the places where we disagree are where I think the most productive thinking happens.

Subscribe to Leptokurticapital. We are all students.

Trade #1 (corn strangle) closed at our profit target. Trade #2 (/ZB explosion position) and Trade #3 (/MES calendar) are still open. This is the second entry in the short strangle sleeve and the fourth trade overall.

Let’s get into it.

The Trade

Underlying: /6C (Canadian Dollar Futures, June 2026)

Structure: Short strangle (~20 delta both sides)

Strikes: 0.710 put / 0.735 call

Expiration: 65 DTE at entry Entry date: April 1, 2026

Premium received: $360

Management rules (same as always):

• Profit target: 50% of credit received → close at ~$180 profit

• Stop loss: 2x credit received → close at ~$720 loss

• Time stop: close at 21 DTE regardless of P&L

Why This Trade, Why Now

The Canadian dollar right now is caught between two forces that are roughly canceling each other out. And that’s exactly the kind of environment where short strangles can thrive, due to elevated fear (IV) but canceling directionality.

Force 1: Oil should be good for the loonie. Canada is the world’s fourth largest oil producer and a massive exporter. Oil above $90 (WTI has been pushing past $92 lately) historically supports the Canadian dollar because it improves Canada’s terms of trade and drives USD into the country to buy energy exports. If you just looked at the oil chart, you’d expect CAD to be strengthening.

Force 2: Everything else is pushing it the other direction. The US dollar is bid as a safe haven because of the same geopolitical crisis that’s boosting oil. Meanwhile, Canada’s domestic economy looks soft. The Bank of Canada held rates at 2.25% on March 18 and basically said they’re stuck: growth is weak (GDP contracted 0.6% in Q4 2025, labor market lost 84,000 jobs in February), but they can’t cut because oil driven inflation is about to push CPI higher. Governor Macklem explicitly called out the dilemma of rising inflation during economic weakness, which is a polite way of describing the early stages of stagflation.

The net result? USD/CAD has been chopping around in the 1.36 to 1.39 range. The commodity tailwind and the risk off headwind are roughly offsetting. The loonie has weakened to about 1.38 (a two month low), but it hasn’t broken out of the broader range. Analysts have fair value estimates around 1.34 to 1.35 but don’t expect the pair to get there until the geopolitical picture clarifies.

This is a range bound thesis. Not because I have a strong view on the direction of CAD (I don’t), but because the opposing forces are strong enough to contain each other. Oil supports the floor, USD strength caps the ceiling. The volatility should be implied more than realized.

And that’s the trade. IVR at entry is approximately 60, which comfortably meets our threshold for elevated implied volatility in the strangle sleeve. We’re getting paid for a level of expected movement that the underlying is unlikely to deliver.

A Note on the Expiration Choice

Regular readers will notice this trade is at 65 DTE rather than our standard 45 DTE. Here’s why.

The available /6C expirations at entry were 37 DTE and 65 DTE. No expiration fell near the ideal 45 day window. Given the choice between too short and too long, I took the longer duration. Three reasons.

First, the liquidity profile was meaningfully better at 65 DTE. Tighter bid/ask spreads, more open interest, less slippage on the fill. At this account size, slippage is not theoretical; it’s a real cost that shows up in your net P&L. The difference between getting filled at mid and getting filled two ticks wide on both legs adds up.

Second, the margin requirement was lower at 65 DTE. This matters for capital efficiency when you’re running a $10K account across multiple sleeves.

Third (and this is the real reason I’m comfortable with it): the IV pickup for going shorter was negligible. We’re capturing 6% implied vol at 65 DTE versus 6.1% at 37 DTE. That’s almost the entire variance risk premium with double the time for the trade to work and less margin consumed. The textbook says 45 DTE is optimal for theta/gamma balance, and that’s true on average across thousands of trades. But when the available expirations bracket your target and one side offers better liquidity, better margin, and essentially the same IV, the practical choice is clear.

I’m documenting this because it’s a deviation from the stated parameters, and deviations get tracked. If this becomes a pattern (taking longer dated expirations for liquidity/margin reasons), the data will show whether it helps or hurts over time.

What the Framework Says

The leptokurtic thesis applies cleanly here. Financial returns are fat tailed and tall in the middle. The Canadian dollar, caught between opposing macro forces, is spending most of its time doing very little (tall middle) while the options market prices in the possibility of a breakout driven by oil, trade policy, or central bank surprises (fat implied tails). We’re selling the overpriced center.

Our Tail Richness Ratio work on currencies shows that FX markets tend to have efficiently priced tails in normal environments but develop pockets of overpriced implied vol during macro uncertainty, exactly because market participants hedge both directions simultaneously. When everyone is worried about CAD strengthening on oil AND weakening on risk off flows, we get hedging bids from both sides. That’s the variance risk premium widening, and that’s what we harvest.

Canadian dollar also adds diversification to the strangle portfolio. Our corn strangle (trade #1, closed for profit) was a grain play. This is a currency play. The correlation between /ZC and /6C returns is low in normal environments. That’s the decorrelation benefit of trading strangles across the full futures universe rather than concentrating in one asset class.

What I’m Watching

The April 29 Bank of Canada decision. This is the next scheduled rate announcement and it comes with a full Monetary Policy Report. We’ll be at approximately 36 DTE at that point, which is past the midpoint of the trade. If the BoC signals a surprise direction (cut due to growth fears, or hawkish tilt due to oil inflation), that could generate a directional move. The management rules handle this: if we’ve hit 50% profit before the announcement, we’re out. If we haven’t, the 21 DTE time stop will be approaching shortly after.

Oil price trajectory. A further spike above $100 to $110 could break the loonie out of its range to the upside (CAD strengthening, USD/CAD falling). Conversely, a de escalation in the Middle East that craters oil would remove the floor under CAD. Both scenarios threaten the put or call side respectively, but our strikes are wide: 0.710 to 0.735 gives us a pretty wide range on either side of the current price given CAD’s historical movement.

CUSMA/USMCA review chatter. The US Canada Mexico trade agreement review is starting to get attention for later this year. Any early signals about trade policy shifts could introduce an idiosyncratic CAD catalyst that has nothing to do with oil. This is the type of risk you can’t quantify in advance, which is why the 2x stop exists.

Correlation with other positions. The /ZB explosion position (Trade #2) has a theoretical correlation to this trade through interest rate expectations. If bonds sell off hard (our explosion thesis), that implies higher rates, which could strengthen USD and weaken CAD. Not a direct relationship but worth monitoring at the portfolio level.

Account Update

This is position #2 in the strangle sleeve (following the corn trade which closed at profit) and position #4 overall.

Cumulative strangle sleeve performance: 1 for 1 on closed trades, +$185.16 net. This trade adds $360 in premium at risk.

Subscribe to Leptokurtic Capital for every trade recap as it happens, plus the framework posts that explain why the strategy works. The math, the fills, and the philosophy. No black boxes.

We are all students.

The Account

Starting account value (March 6): $10,000

Ending account value (March 30, closed trades only): $10,185.16

Net P&L for March: $185.16

Total Return: ~1.8%

Active Sleeves This Month

A quick reminder of the three sleeve structure (covered in depth in the framework posts):

Sleeve 1 (Short Strangles) harvests the overpriced center of the distribution. We sell 20 delta strangles on high IVR futures at 45 DTE and manage them mechanically: 50% profit target, 2x premium stop loss, 21 DTE time stop. The center of the leptokurtic distribution is overpriced by implied volatility. We sell it.

Sleeve 2 (Double Calendars) captures theta on low IVR underlyings while maintaining long vega exposure. 30 delta, 2 week by 3 week structure. 25% profit target, 50% circuit breaker, close at 3 DTE on the short legs.

Sleeve 3 (Explosion Positions) buys the underpriced tails. Deep OTM options on futures where our Convexity Score screening identifies the cheapest tail convexity. Direction agnostic. These are independent positive EV trades, not hedges.

All three sleeves exploit the same observation: financial returns are leptokurtic. The center is overpriced. The tails are underpriced. We trade both sides.

Trade by Trade

Trade #1: Corn Short Strangle (/ZC) — CLOSED ✓

Sleeve: Short Strangles Underlying: /ZC (Corn Futures)

Structure: 425p / 495c strangle Entry: March 9 | 52 DTE Exit: March 30

Profit target hit Credit received: $393.75 Gross P&L: +$200.00

Net P&L (after commissions): +$185.16 Return on credit: 47.0%

This one did exactly what the system is designed to do. Corn sat in a range, time passed, theta decayed, vol contracted slightly, and we closed at our mechanical target. No adjustments, no drama, no staring at the screen hoping.

The gross profit of $200 on $393.75 in credit is right at our ~50% profit target. The net profit of $185.16 accounts for the round trip commissions, which is worth tracking (commissions on futures options are not trivial at this account size, and pretending they don’t exist is how people fool themselves about their actual edge).

Duration: 21 days. Our historical average hold across 130+ strangle trades is 27 days, so this one resolved a bit quicker than usual. That’s fine. Faster resolution means faster capital recycling.

What the system predicted: High IVR entry, uncorrelated underlying, mechanical management. The Tail Richness Ratio for corn showed the center of its distribution was sufficiently overpriced to justify the trade. Corn delivered exactly the boring, range bound month that the distribution says happens more often than implied vol suggests.

Grade: A. Textbook execution. Entered on criteria, even during a chaotic macro environment, managed mechanically, exited at target.

Trade #2: Treasury Bond Explosion Position (/ZB) — OPEN

Sleeve: Explosion Positions

Underlying: /ZB (30 Year Treasury Bond Futures)

Structure: Deep OTM puts (5 delta)

Entry: ~March 18 | ~65 DTE at entry

Current DTE: 53

Current status: Slightly in profit (directional move in our favor)

The thesis here is a potential tail mispricing in bonds. The conventional wisdom during geopolitical conflict is that treasuries rally as a safe haven. Our Convexity Score screening flagged /ZB as offering unusually cheap tail convexity because the market was pricing in that safe haven assumption.

The contrarian case: in an active, escalating conflict (particularly one with inflationary supply shock characteristics like the current situation in the Middle East), bonds might actually sell off rather than rally. Rising energy prices feed inflation expectations, which feed rate expectations, which pressure the long end. The “safe haven” narrative assumes the crisis is deflationary. This one isn’t.

/ZB has moved down since entry, which puts our puts slightly in the money relative to where we bought them. But “slightly in profit” on an Explosion Position is almost meaningless. These positions are sized for asymmetric payoffs. We’re looking for 5x, 10x, 20x multiples of our premium paid. A small directional drift in our favor is nice but nowhere near exit criteria.

Management plan: Continue to hold. We are 12 days into the trade with 53 DTE remaining. Plenty of time for the thesis to play out (or not, which happens way more often. Remember, these are tail hedge/convex positions).

Grade: TBD Too early to evaluate. The thesis is working directionally. The question is whether the move is large enough, fast enough, to overcome time decay on deep OTM options. That’s always the question with Explosion Positions. However, we have to note that most of the juice/work is done on entry here; we identified some cheaper than usual convexity, bought it, and now we wait.

Trade #3: /MES Double Calendar — OPEN

Sleeve: Double Calendars

Underlying: /MES (Micro E-mini S&P 500)

Structure: 30 delta double calendar, 17d/37d at entry

Entry: March 24

Total debit: $436.25

Current P&L: ~+6% ROI Short legs DTE: 10

This trade is working but not yet at target. We’re shooting for our 25% profit target on the debit paid, or we close at 3 DTE on the short legs, whichever comes first. With 10 DTE on the short legs, we have about a week for the remaining theta to kick in.

Transparency note: I flagged this in the trade recap post, but it’s worth restating here. The IVR at entry was approximately 37, which is above my preferred threshold of sub 25-30 for calendar entries. I took the trade because the structure was otherwise attractive and I wanted to get the calendar sleeve active in the portfolio for documentation purposes.

Management plan: Hold for 25% profit or 3 DTE time stop. The 50% loss circuit breaker has not been threatened, even though the put leg is being tested (6386 market price at time of writing vs our 6390 calendar put side). This is mostly due to vol expansion on the downward directional move, and shows the power of the calendar structure. If the short legs reach 3 DTE without hitting the profit target, we close.

Grade: Incomplete. On track. The IVR entry deviation is noted and will factor into the process review.

Month in Numbers

One closed trade in month one is a small sample. I’m not going to pretend that a 100% win rate on N=1 tells us anything about the strategy. What it tells us is that the system executed correctly on its first public trade. The mechanics worked. The fills were real.

What I Learned

Commissions matter at this account size. The $14.84 round trip on the corn strangle represents 3.8% of the gross profit and 7.4% of the credit received. At a $10K account trading futures options, commissions are a real drag. This is why I track net P&L obsessively.

The calendar IVR threshold exists for a reason. The /MES calendar was entered above my preferred IVR range. It’s currently profitable, which proves nothing about whether the deviation was correct. The system has parameters for a reason, and bending them in month one (even for documentation purposes) sets a precedent I need to be careful about. Future months will track how many trades deviate from entry criteria and whether those deviations correlate with outcomes.

Patience on Explosion Positions is the hardest part. The /ZB trade is “working” in the sense that it’s slightly profitable. Every instinct says to celebrate. But the entire point of this sleeve is asymmetric payoff. Closing an Explosion Position at a small profit because it feels good is the single fastest way to destroy the positive expected value of the strategy. The management rules exist to override the instinct.

Looking Ahead: April 2026

The strangle sleeve will be looking for the next high IVR entry across the futures universe. With the Iran/oil situation elevating implied vol across the commodity complex, there should be no shortage of candidates. The question is selectivity. Elevated IV means higher premium but also wider potential moves. The screening criteria don’t change just because the macro environment is exciting.

The /ZB Explosion Position continues to develop. 53 DTE is plenty of runway.

The /MES calendar should resolve one way or another in the next 7-10 days.

A Note: There will be a future post where I explain my convexity ranking system in depth.

I’ll publish convexity score rankings for paid subscribers when the paid tier launches (targeting ~500 free subscribers). Until then, the screening methodology are covered in the framework posts.

This is month one of a public trading journal. The account is real. The trades are real. If you want to follow along from the beginning, the full framework starts HERE. As always, not financial advice.

We are all students.

This is the first live calendar trade in the portfolio. If you’ve been following from the beginning, you’ve seen me sell the center with strangles (sleeve 1, the corn trade is still live). Now we’re adding the second sleeve. Selling the center differently.

For anyone who hasn’t read the full theory post on calendars (it’s in the archive, titled “Selling the Center, Differently”), here’s the 30-second version before we get into the trade.

The Structure in Plain English

A double calendar spread is four legs. You sell a shorter-dated option and buy a longer-dated option at the same strike, on both the put side and the call side. You pay a net debit to put it on (the longer-dated options cost more than the shorter-dated ones).

The trade makes money two ways. First, the short-dated options you sold decay faster than the long-dated options you bought (theta positive). Every quiet day, the spread widens in your favor. Second, if implied volatility increases, the long-dated options gain more than the short-dated options lose (vega positive). This is the structural difference from strangles, which are short vega. Calendars are long vega.

Being theta positive AND vega positive at the same time is the whole point. You make money from time passing quietly, but you also benefit if things get loud. The strangle sleeve hates vol expansion. The calendar sleeve likes it. They offset each other at the portfolio level.

Delta is approximately neutral. We’re not betting on direction. We’re betting on time decay and, implicitly, on vol not collapsing further from here.

The Trade

Underlying: /MES (Micro E-mini S&P 500 Futures) Structure: Double calendar spread Entry date: March 24, 2026

Put side:

• Sold: April 10 /MES 6390 put @ 367.50 credit ($367.50)

• Bought: April 30 /MES 6375 put @ 587.50 debit ($587.50)

Call side:

• Sold: April 10 /MES 6750 call @ 282.50 credit ($282.50)

• Bought: April 30 /MES 6750 call @ 498.75 debit ($498.75)

Total DEBIT paid: $436.25

This is the max I can lose on this trade (excluding some edge cases around early assignment and execution, but for practical purposes, $436.25 is the risk).

Management rules:

• Profit target: 25% return on debit → close at ~$545 total value (profit of ~$109)

• Stop loss: 50% loss on debit → close at ~$218 total value (loss of ~$218)

• Time stop: close entire structure when short leg reaches 3 DTE (April 7)

Why /MES, Why Now

A few reasons.

First, the account size. /MES (Micro E-mini) is the right instrument for a $10K account trying to run calendars on equity indices. The full-size /ES contract would eat too much margin. Micros let me express the same thesis with appropriate sizing.

Second, IVR on /ES (and by extension /MES) is in an interesting spot right now. We’ve had the Iran-driven vol spike, and implied volatility is elevated but starting to normalize. This is a decent environment for a calendar: we’re entering with enough IV that the long-dated options have meaningful vega exposure (if vol re-expands, we benefit), but the front-month IV is also rich enough that the short-dated options we sold carry good premium. We’re getting paid on the theta side while keeping the vega optionality.

Third, I wanted the first calendar to be on something most readers can relate to. Everyone watches the S&P. If I put the first calendar on the Australian dollar, half of you would tune out before I finished explaining the contract specs (fair enough, honestly).

A Note on the Duration Mismatch

Sharp-eyed readers will notice that the short and long legs are 17 days and 37 days apart, not the 2-week/3-week structure I described in the theory post. The difference is about liquidity.

The target structure would have expirations exactly 14 and 21 days out. In practice, /MES options don’t have weeklies at every expiration the way SPY or SPX equity options do. The available expirations closest to what I wanted were April 10 and April 30. That gives me 17 days on the short leg and 37 days on the long leg, with a 20-day gap between them instead of the ideal 7.

The wider gap changes the character slightly. More time between expirations means more vega exposure on the long leg (good if vol expands) but a slower theta differential (the front option doesn’t decay as much faster relative to the back option as it would in a tighter structure). It’s a tradeoff. I’d rather have the slightly wider structure on a liquid expiration than force the perfect structure on an illiquid one where the bid-ask spread eats the edge.

This is the kind of real-world compromise that theory posts don’t cover but matters a lot in practice. The textbook says one thing. The options chain says another. You adapt.

The Put Strike Difference

You might also notice the put strikes are slightly different: 6390 on the short leg, 6375 on the long leg. In a perfect double calendar, both legs are at the same strike. I offset by 15 points here to slightly improve the fill. The practical impact is minimal (15 points on /MES is a tiny fraction of the underlying price) but it technically makes this a diagonal rather than a pure calendar on the put side. The Greek profile is almost identical. I’m noting it for full transparency, not because it materially changes the thesis.

What I’m Watching

Theta working in our favor. The April 10 short options should decay faster than the April 30 longs over the next two weeks. Every day that /MES stays roughly in its current range, the spread widens. The 25% profit target ($109 on $436.25) is achievable if we get 10-12 days of relatively quiet price action.

Vol direction. If the Iran situation drags and IV spikes again, the calendar benefits from the vega exposure. The long April 30 options gain more from an IV increase than the short April 10 options lose. This is the built-in hedge that makes calendars complement strangles in the portfolio.

Conversely, if IV collapses quickly (a sudden peace deal, say), the calendar could lose value even if the underlying hasn’t moved. The long-dated options we own would lose more vega value than the short-dated options we sold. That’s the risk scenario for this structure.

The 6390-6750 range. This is roughly our comfort zone. If /MES stays between these strikes, both the put calendar and call calendar maintain their favorable structure. A large move in either direction pushes one side deep in the money, which disrupts the theta differential and can trigger the 50% stop. The S&P would need to move roughly 3%+ in either direction to seriously threaten the strikes, which is meaningful but not impossible in the current environment.

Time stop at 3 DTE. The April 10 short options expire on April 10. I’ll close the entire structure by April 7 regardless of P&L. Just like with strangles, I don’t want gamma risk from expiring options. The last few days before expiration are where calendars can get unpredictable as the short leg’s gamma accelerates.

How This Fits the Portfolio

We now have two active sleeves:

Sleeve 1: Corn strangle (still live, approaching profit target, IV continuing to normalize)

Sleeve 2: /MES double calendar (just entered, $436.25 at risk)

Sleeve 3: /ZB bond explosion position,(entered 6 days ago, $296.87 at risk)

The three active positions have complementary risk profiles. The corn strangle is short vega. The /MES calendar is long vega. If there’s a broad vol spike, the corn strangle gets pressured but the /MES calendar benefits. If vol bleeds lower across the board, the corn strangle wins easily but the calendar’s vega exposure works against it (partially offset by the theta). If there is a large black swan tail event, the strangle and calendar likely stop out, but the explosion position should compensate.

They don’t perfectly hedge each other (different underlyings, different magnitudes), but the portfolio as a whole is less vulnerable to a vol regime change than any position alone.

This is the portfolio construction principle in practice, not just in theory. We’ll see how it plays out.

Subscribe to Leptokurticapital. The framework is becoming a portfolio.

I recently shared some thoughts on Reddit about selling strangles on futures and the response was pretty overwhelming (in a good way). One of the most common follow-up questions, both in comments and in DMs, was some version of: “OK, the strategy makes sense, but how do you decide how much to put on each trade?”

This is, I think, the most underrated question in all of trading. Not which trade. Not when to enter. How much.

I’ve seen plenty of traders with good strategies blow up. I’ve never seen a trader with good sizing blow up. The inverse is also true and much more common: mediocre strategies with excellent sizing can grind out surprisingly good returns. Because sizing determines whether you survive long enough for the edge to play out. Can’t make returns if you go to zero, after all.

So let’s talk about the math of sizing. And more specifically, about a formula developed by a Bell Labs scientist in the 1950s that is simultaneously one of the most elegant results in probability theory and one of the most dangerous tools a trader can misuse.

John Kelly and the Information Problem

John Larry Kelly Jr. was a physicist at Bell Labs in the 1950s. He wasn’t a trader. He was working on information theory, specifically the problem of how to transmit signals over noisy telephone lines without losing data.

In 1956 he published a paper called “A New Interpretation of Information Rate” in the Bell System Technical Journal. The core insight was deceptively simple: if you have an edge in a repeated game (a bet where the odds are in your favor), there’s a mathematically optimal fraction of your bankroll to wager on each round that maximizes the long-run growth rate of your wealth.

Bet too little and you’re leaving money on the table. Your wealth grows, but slowly. Bet too much and you risk ruin. Your wealth fluctuates wildly and, with high probability over enough rounds, you go broke despite having an edge.

The optimal fraction, the one that maximizes long-run compound growth, is now called the Kelly Criterion.

Kelly himself apparently didn’t think much about the other implications of his formualation. But others did. Ed Thorp, a mathematics professor who famously beat blackjack (and wrote “Beat the Dealer” about it), was one of the first to apply Kelly’s formula to both gambling and financial markets. Thorp went on to run one of the most successful quantitative hedge funds in history (Princeton Newport Partners, returning ~20% annually for nearly two decades with only three losing months). He credited Kelly sizing as a foundational part of his approach.

The connection between information theory and optimal risk sizing is, philosophically, one of my favorite things in all of math. Kelly showed that the growth rate of your wealth is literally equivalent to the information rate of the channel between you and the market. How much you know determines how much you should bet. There’s something deeply satisfying about that, at least to me.

The Formula (Here Comes Math, Don’t Panic)

For a simple binary outcome (you either win or lose a fixed amount), the Kelly fraction is:

f = (p × b - q) / b*

Where:

• f* = fraction of bankroll to risk

• p = probability of winning

• b = ratio of amount won to amount lost (the “odds”)

• q = probability of losing (which is just 1 - p)

Let’s plug in my strangle numbers as an example. Win rate is 83.65%, average winner is 0.47x of risk, average loser is 1x of risk.

So: p = 0.8365, q = 0.1635, b = 0.47

f* = (0.8365 × 0.47 - 0.1635) / 0.47

f* = (0.3932 - 0.1635) / 0.47

f* = 0.2297 / 0.47

f ≈ 0.489*

Kelly says bet about 49% of the bankroll on each trade. Nearly half.

If you’ve been following this publication, your immediate reaction should be: “That seems way too high.” And you’d be right. Our goal is to risk 2% per trade. That’s roughly 4% of Kelly. Which might sound absurdly conservative until you understand why.

Why Everyone Uses a Fraction of Kelly

Full Kelly is mathematically optimal in a very specific sense: it maximizes the expected logarithmic growth rate of wealth over an infinite number of trials, assuming you know the true probabilities with certainty.

That last part is doing a LOT of heavy lifting.

In practice, nobody uses full Kelly. Not Thorp (he reportedly used half Kelly). Not Renaissance Technologies. Not any serious fund I’m aware of. The reasons are instructive.

You don’t know the true probabilities. The win rate of 83.65% is estimated from 130ish strangle trades I’ve done, and is backed by others trading similar strategies with similar results. That’s a sample, not the population though. The true win rate could be 80%. It could be 87%. It could shift over time as market conditions change. Kelly assumes you know p and b with certainty. You don’t. You have estimates with confidence intervals. And the consequences of overestimating your edge are much worse than the consequences of underestimating it. Bet half Kelly when full Kelly is optimal and you get about 75% of the optimal growth rate (not bad at all, honestly). Bet double Kelly because you overestimated your edge and you get a negative growth rate. You go broke. The downside of over-betting is ruin. The downside of under-betting is slower growth. These are not symmetric risks.

Kelly assumes independent bets. Each trial is assumed to be independent of every other trial. In reality, my strangles are partially correlated (though not entirely; and especially when we’re diversified across asset classes). Still, a broad vol spike hits multiple positions simultaneously to some degree. The effective “bet” in a correlated environment is larger than the nominal bet, because several positions can lose at the same time. Kelly doesn’t account for this. If you size each position at full Kelly as if it were independent, your actual portfolio-level risk is well above Kelly.

Kelly maximizes growth but the path is brutal. Even at full Kelly with a genuine edge, the drawdowns are enormous. The standard deviation of returns under full Kelly is roughly equal to the expected return. Translation: you should expect to be down 50% at some point. Maybe more. The math says you’ll eventually recover and compound past it. Your psychology (and investors, if you have them) might not survive long enough for “eventually” to arrive. Half Kelly cuts the expected growth rate by only 25% but cuts the variance roughly in half. Quarter Kelly cuts growth more but makes the ride something a human can actually tolerate. There’s a reason Thorp, one of the most mathematically rigorous traders in history, used half Kelly. He understood that the theoretical optimum is only useful if you can actually execute it over hundreds of trials without panicking and quitting. (See a previous post on psychology!)

Utility isn’t logarithmic for everyone. Kelly maximizes log utility of wealth, which implies that losing half your money is exactly offset by doubling your money. This is a specific assumption about risk preferences. Many people (and most institutions) are more loss-averse than this. For them, fractional Kelly isn’t just practically wise. It’s actually closer to the theoretical optimum for their personal utility function.

Now Add Leptokurtosis and Things Get Really Interesting

Here’s where this connects to the core thesis of this publication, and where I think I have something to add beyond the standard Kelly discussion.

The Kelly inputs I used above (83.65% win rate, 0.47x winner, 1x loser) come from 130 trades. That’s a decent sample for estimating the center of the return distribution (the win rate, the average winner). It is a terrible sample for estimating the tails (the worst-case losses, the probability of extreme simultaneous drawdowns).

This is the leptokurtosis problem applied to position sizing. And I don’t think it gets talked about enough.

If you’ve read my previous posts, you know that across every futures market I’ve tested, extreme moves happen 10-20x more frequently than a normal distribution predicts. Three-sigma months occur when they “shouldn’t.” Four-sigma months occur when they “really shouldn’t.” My 130 trades probably contain a handful of tail events, but the sample is almost certainly too small to accurately represent their true frequency.

What does this mean for Kelly?

It means the inputs are biased in a specific direction. The win rate and average winner are probably approximately correct (these are central tendency measures, and 130 trials is adequate for that). But the average loser is probably understated, because the sample hasn’t fully captured the fat tails. The true loss distribution has more extreme events than my 130 trades suggest.

Kelly with understated loss estimates will recommend a bet size that’s too large. Not because Kelly is wrong as a framework. Because the data you’re feeding it is incomplete. You’re telling Kelly “my average loser is 1x of risk” when the truth is more like “my average loser is 1x of risk in 95% of months, but in 5% of months things go badly wrong and the loser is 2-3x of risk because of gaps, correlation spikes, and liquidity failures.”

This is exactly why I size at roughly 4% of Kelly rather than 50% or even 25%.

The 2% per position sizing isn’t arbitrary. It’s informed by asking: “What’s the worst realistic number of simultaneous stops I could face in a month where everything goes wrong, and what would the portfolio drawdown be?” I walked through this in detail in a previous post. Six simultaneous stops at 2% each is 12% (or potentially worse if there’s a blow-by stop, which I also modeled). That’s a bad month. But combined with the margin reserve and the tail convexity sleeve, the portfolio survives it.

If I were sizing at 25% of Kelly (roughly 12% per position, something like what I will need to do in my 10k account), that same six-stop scenario would be a 72% drawdown. Ouch. Now, we have other strategies that can help offset this, and large drawdowns in smaller accounts like this are less painful, but the point is that sizing incorrectly can be lethal.

The Meta-Insight

There’s something almost philosophically recursive about this that I find hard to stop thinking about.

The same distributional mispricing that creates our trading edge (leptokurtosis makes the center overpriced and the tails underpriced) also makes our risk management inputs unreliable (backtests undercount tail events). We’re using the existence of fat tails to justify the trades, while simultaneously needing to account for fat tails in how much we bet on those trades.

The resolution is fractional Kelly taken to an extreme. Size for the true (leptokurtic) distribution of outcomes, not the Gaussian approximation and not the backtest sample. In practice, this means:

Smaller positions than Kelly suggests (because tails are fatter than your sample shows).

A dedicated margin reserve (because margin requirements expand during the exact tail events your sample underrepresents).

And (this is the part I think is novel) a dedicated sleeve of the portfolio that explicitly buys the tails. The Explosion Positions aren’t just a return-generating strategy. They’re also a sizing tool. By owning tail convexity, you partially offset the tail risk that makes aggressive sizing dangerous. You can think of it as buying the right to be slightly less conservative on sizing, though in practice I keep the conservative sizing AND the tail positions, because belt AND suspenders is the only responsible approach when the risk you’re managing is, by definition, larger than your models suggest.

What the Funds Actually Do

It’s worth noting that serious quantitative funds have converged on similar conclusions through different reasoning. I can’t claim to know this for sure, since I’ve not worked on Wall Street, but this is what I hear from friends who do.

Most systematic hedge funds use something in the range of 10-25% of Kelly, which they typically derive not from the simple binary Kelly formula but from more sophisticated portfolio optimization models that try (key word, try) to account for correlation, fat tails, and parameter uncertainty. The fact that extremely well-resourced funds with PhDs and decades of data still size at fractions of Kelly should tell you something about how seriously the smart money takes the gap between theoretical optimality and practical survival.

Thorp, as mentioned, used half Kelly and he was operating in environments (like convertible arbitrage) where he had better edge estimates than most traders ever will. The less certain you are about your edge, the smaller your Kelly fraction should be.

A practical heuristic I’ve found useful: your Kelly fraction should be roughly inversely proportional to how fat the tails are in your strategy’s return distribution. If you’re trading something with thin tails and well-estimated parameters (like a statistical arbitrage with thousands of trades per month), you can afford to run at 25-50% Kelly. If you’re trading something with fat tails and limited sample sizes (like short strangles on futures with 130 trades), 5-10% of Kelly is more appropriate.

My 4% of Kelly might seem extreme. But I’d rather be the trader who compounds at 80% of the optimal rate for 20 years than the one who compounds at the optimal rate for 3 years and then blows up because a 4-sigma month happened when the model said it wouldn’t.

The Takeaway

Sizing is not the boring administrative step you do after finding a good trade. Sizing keeps you alive. A brilliant strategy with bad sizing is worse than a mediocre strategy with great sizing, because the brilliant strategy will eventually encounter a tail event that the sizing can’t survive, and then both the strategy and the capital are gone.

The Kelly Criterion gives us the theoretical framework. Leptokurtosis gives us the reason to be much more conservative than Kelly’s raw output suggests. And the multi-sleeve portfolio structure (strangles + calendars + Explosion Positions) gives us a way to manage the tail risk that Kelly can’t account for, by owning convexity that pays off in exactly the scenarios where aggressive sizing would be fatal.

In other words: sell the center (smaller than Kelly says you should), and buy the tails (with the capital that conservative sizing frees up). The sizing strategy and the trading strategy are expressions of the same thesis. It’s all leptokurtosis, all the way down.

Next time: the vol surface as a three-dimensional map of distributional assumptions, where it’s wrong, and how all three sleeves exploit different parts of the same mispriced surface.

Subscribe to Leptokurticapital. We just did math on a Tuesday. Or whatever day you’re reading this. The point is, we did math, and I think it was worth it.

This is the second trade recap post, and the first from Sleeve 3: Explosion Positions. If you haven’t read the theory post on these, the short version is: we systematically buy deep out-of-the-money options on futures where the market underprices tail risk. The thesis isn’t that any one of these hits. It’s that across a portfolio of them, the ones that do hit pay enough to cover the ones that don’t, and then some.

Trade #1 was a short strangle on corn (Sleeve 1, selling the center). This one is the opposite side of the framework. We’re buying, not selling. And we’re buying because the options market thinks 30-year Treasury bonds are going to sit still.

I think that’s a bad bet right now.

The Trade

Underlying: /ZBM6 (30-Year Treasury Bond Futures, June 2026) Structure: Long 5-delta strangle (Explosion Position) Strikes: 104 put / 127 call Expiration: 66 DTE at entry Entry date: March 17, 2026

Premium paid:

• Combined cost: 0”19 ($296.88 total)

• This is the max loss. We can’t lose more than we paid.

Previous year Implied Volatility Rank (IVR) at entry: 19

As a reminder, this means it is in the 19th percentile of its own implied volatility relative to the last year.

Why 66 Days Instead of 38

This is worth explaining because it’s a framework decision, not arbitrary.

We had two expiration choices: 38 DTE and 66 DTE. The 38-day option had an implied volatility of about 11.2%. The 66-day option was running at 11.8%.

On the surface, 66 days costs more in absolute terms. But we’re getting a higher IV per dollar, which means better convexity exposure for the premium spent. The extra 28 days gives us more time for a tail event to materialize, and we’re buying that time at a relatively cheap rate because the vol curve is barely upward sloping here.

Think of it this way: we’re buying lottery tickets. One ticket expires in 5 weeks, the other in 9 weeks. If the 9-week ticket costs only modestly more but gives you almost twice the time for something dramatic to happen, that’s the one you want. Especially when the implied volatility baked into the 9-week ticket is actually higher, meaning the market is giving us slightly more convexity per unit of premium.

More time + more vol per dollar = better explosion position.

Why Treasury Bonds, Why Now

Here’s the macro context, and it’s wild.

The U.S. is currently at war with Iran. We say we’re not, but I don’t believe it. That started on February 28th. Oil spiked. The Strait of Hormuz is under threat. The standard playbook says: war happens, investors flee to safe haven bonds, yields drop, bond prices rise. Textbook.

Except that’s not what happened. Treasury yields actually rose after the strikes began. The 10-year went from around 3.93% to as high as 4.26% in the first week. Long bonds sold off. The “safe haven” didn’t behave like a safe haven.

Why? Because the market is stuck in a genuinely weird spot. Oil prices are spiking, which means inflation risk is rising, which means bond investors are demanding higher yields to compensate for the loss of purchasing power. Normally war is deflationary (fear, risk-off). But this war comes with an oil supply shock that’s inflationary.

This is the kind of environment where tail events happen in rates. The market is pricing in a narrow range because the push and pull between inflation fears and recession fears are roughly balanced right now. But that balance is fragile. If either side of the argument wins decisively (oil stays above $100 and inflation rips, or the economy buckles and we’re suddenly in recession territory), /ZB is going to move a lot.

At an IVR of 19, the options market is saying: we don’t expect much. With the world on fire? That’s a price I’m willing to pay.

The Tail Richness Argument

This is where our framework earns its keep.

If you’ve followed the Tail Richness Ratio work (still a work in progress, I’ll keep saying that until the sample sizes make me more confident), /ZB is one of the futures where empirical tail events happen more frequently than a normal distribution would predict. That’s the whole thesis of leptokurtic distributions applied to markets: the tails are fatter than options pricing assumes.

Treasury bonds in particular are vulnerable to regime shifts. When rates move, they can move violently. We saw it in 2022 when the long bond had its worst year in history. We saw it during COVID when bonds ripped higher in a matter of days. We saw it during the SVB crisis in 2023. These aren’t once-in-a-generation events. They keep happening, and they keep happening more often than the implied vol surface says they should.

At a 5-delta, we need a large move for this to pay off. The 104 put is roughly 10 full points below current levels, and the 127 call is about 13 points above. Those are big moves. But we’re paying under $300 for the right to participate if either of those moves happens in the next 66 days.

And right now, the catalysts for a big move are staring us in the face. An active war that’s disrupting global energy markets. A Fed that’s stuck between cutting (economy weakening) and holding (inflation sticky). A bond market where the safe haven bid has broken down. Any one of these resolving decisively could send /ZB to levels that make this position worth multiples of what we paid.

Sizing and Honesty

I need to be upfront here: this is a much larger position relative to the account than I’d normally take. In a full-sized account, an Explosion Position would be a small allocation, maybe 0.5% to 1% of capital, because the expected value comes from running many of these across different underlyings over time. You need the portfolio effect. Any single one is likely to expire worthless.

At our current account size, we don’t have the luxury of running 10 or 15 of these simultaneously at the right sizing. So this one is oversized relative to the ideal. That’s a real limitation.

Similarly, management on Explosion Positions is normally more nuanced than what we can execute here. In an ideal world, if the position doubles in value on a vol spike but hasn’t hit the target, you might take half off to lock in some return and let the rest ride. With a single contract, that granularity isn’t available. It’s hold or close, no in-between.

I may also adjust the duration approach on these. The framework says hold to expiration or until a defined exit target. But at this account size, if I get an early windfall from a vol spike, I might take the win rather than gambling on further movement. I’ll document whatever I do and explain why.

The $296.88 max loss is defined and manageable. But I want to be clear that this isn’t the ideal sizing or management structure. It’s what we can do with the account we have.

Management

Exit targets:

• If the position reaches 200%+ of premium paid, evaluate closing for profit

• If vol spikes and one side doubles in value, consider closing that side

• Hold to expiration if neither target is met (max loss = premium paid)

What I’m watching:

• Oil prices and the Hormuz situation (escalation = vol spike in rates)

• Fed meeting language and rate cut expectations

• Any credit market stress that could trigger a flight to quality

• Inflation data (CPI, PPI) that could shift the narrative

Why This Matters for the Framework

This is the trade that most people get wrong about what we do here.

When I tell people I sell short strangles on futures, they hear “picking up pennies in front of a steamroller.” Fair enough, that’s the criticism, and there’s truth to it if that’s all you do.

But the Explosion Positions are the other side. We’re also buying the steamroller. Not in the same underlying at the same time (that would just be a calendar spread), but across the portfolio. Sleeve 1 sells the center. Sleeve 3 buys the tails. The combined portfolio has a return distribution that’s different from either sleeve alone.

Most premium sellers never buy options. Most options buyers never sell premium. We do both, systematically, because the leptokurtic structure of returns means both sides of that trade have edge if you know where to look.

Quick Portfolio Update

Trade #1 (Corn Strangle): Still open, roughly breakeven to slightly positive. IV has continued to normalize. Watching the USDA Prospective Plantings report on March 31.

Trade #2 (Treasury Bond Explosion): Just entered. Max risk defined at $296.88.

More soon.

Subscribe to Leptokurticapital for weekly trade recaps alongside the theory posts. This is where the framework meets reality.

By the end, either you’ll be convinced the framework is resilient, or you’ll think I’m insane. Both are valid conclusions, honestly. Let’s see what happens.

The Setup: What We’re Holding

Let me paint a picture of a specific month. It’s October. We’ve been running the strategy for a while, and we’ve added capital. We can split positions well now, which is cool. Here’s what the portfolio looks like going into the month (using approximate allocations on a scaled-up version of the portfolio, say $100K, since the $10K account won’t have this many positions simultaneously, but this is what we’re building toward):

Strangle sleeve (12 positions, ~24% deployed, 25% margin reserve): Short strangles on /6E (euro), /6J (yen), /6B (pound), /ZS (soybeans), /ZW (wheat), /ZC (corn), /GC (gold), /SI (silver), /HG (copper), /CL (crude oil), /NG (natural gas), /ZN (10-year note). All 20-delta, entered around 45 DTE, currently around 25-30 DTE. Some are profitable, some are near breakeven. Normal month so far.

Calendar sleeve (8 positions, ~26% deployed): Double calendars on /ES (S&P 500), /NQ (Nasdaq), /ZB (30-year bond), /6A (Aussie dollar), and a few others. Low IVR at entry, 30-delta, short legs around 10-12 DTE.

Explosion Positions (3 assets, both directions): This month’s Convexity Score top 3 were natural gas, wheat, and the yen. We own 5-delta calls and 5-delta puts on each, 30-35 DTE remaining. We got all of them at low IVRs, which is nice, though we do have some unwanted cross correlation by owning both explosion positions and strangles in the yen and NG. Luckily, we entered them in different vol environments in this case, and their sizes are quite different as we’ll see.

A diversified, multi-sleeve, multi-asset portfolio. Exactly as designed. Everything looks normal, including some inevitable unwanted incidental correlations.

Then the world does what the world does.

Week 1: The Shock

A major geopolitical event. Let’s not get too specific (I don’t want to accidentally predict something and have to deal with that), but imagine something on the scale of Russia invading Ukraine, or a major escalation in the Middle East (Ahem. Iran), or a surprise sovereign debt crisis in a G7 country. The kind of thing that makes the front page of every newspaper simultaneously and triggers the phrase “risk-off” approximately ten thousand times on financial television. Whatever that means…

Here’s what happens to our portfolio in the first 48 hours:

Equities crater. /ES drops 7% in two days. /NQ drops 9%. This is a significant but not unprecedented move (March 2020 did worse in a shorter window). VIX spikes from 16 to 35.

Currencies go haywire. The yen strengthens sharply (flight to safety). The euro weakens. The pound weakens harder (it always does in crises, the pound is dramatically fragile in risk-off environments). The Aussie dollar collapses (commodity currency, China-linked, risk-off = down).

Commodities whipsaw. Crude oil spikes on supply disruption fears. Gold spikes on safe-haven demand. Grains initially sell off on broad deleveraging but then spike on supply chain concerns. Natural gas is volatile in both directions (because when isn’t it, when the world is ending).

Bonds rally hard. Flight to quality. 10-year yields drop 30 basis points in a week. 30-year bonds rip higher.

This is a correlated risk-off event. The kind of thing that makes “diversification” feel like a joke, because everything is moving, and a lot of it is moving against you.

Let’s count the damage.

The Strangle Sleeve: Taking Hits

Of our 12 strangles, here’s what’s happening:

/6J (yen) strangle: stopped out. The yen strengthened past our call strike. The 2x stop triggers. Loss: 2% of portfolio (by design).

/6B (pound) strangle: stopped out. Pound weakness blew through the put strike, and liquidity dried up too! We blew through the stop, for an outsized loss. Loss: 6%.

/6E (euro) strangle: stopped out. Same dynamic, weaker euro. Loss: 2%.

/CL (crude oil) strangle: stopped out. Oil spiked past the call strike on supply fears. Loss: 2%.

/GC (gold) strangle: close to stop. Gold’s rally is testing the call side. Not stopped yet but uncomfortable. We’re watching.

/SI (silver) strangle: stopped out. Silver followed gold but more violently (as it does). Loss: 2%.

/ZN (10-year note) strangle: stopped out. Bond rally broke through the call strike. Loss: 2%.

That’s six stops hit in the first week, including a 3x blow by. Five positions, each losing roughly 2% of portfolio, and one losing 6%. Total strangle damage: approximately 16% of portfolio.

The remaining six strangles (/ZS, /ZW, /ZC, /HG, /NG, and /GC if it doesn’t stop) are under pressure but haven’t triggered stops. Grains are whipping around but haven’t made a sustained directional move large enough to breach 20-delta strikes. Copper is weak but holding. Natural gas is being natural gas (chaotic, but our strikes are wide enough to absorb it so far).

Important: this is the scenario the 2% position sizing is designed for. Six simultaneous stops is bad. It’s a bad month. But 18% is not an existential drawdown, especially if other sleeves can help to offset this somewhat. The portfolio is still functional. The margin reserve (25% of capital, remember) has absorbed the margin expansion that always accompanies a vol spike, so we haven’t been force-liquidated. We’re hurt but standing.

The Calendar Sleeve: Mixed

The calendars are having a complicated week.

Vega is helping. Implied volatility has exploded higher across the board. Our long-dated options have gained more than our short-dated options have lost (on the vol axis). The calendar spreads are wider. This is the structural offset working as designed.

But delta is hurting. The large directional moves in equities and bonds have pushed the underlying prices away from our calendar strikes. The /ES and /NQ calendars are losing their favorable structure as the index falls well below the put calendar strikes. The /ZB calendar is gaining on bonds rallying toward the call calendar strike, but overshooting it.

Net effect on the calendar sleeve: roughly flat, maybe slightly negative. The vega gain and the delta loss are partially offsetting. Some calendars are up, some are down. A few hit the 50% circuit breaker and we close them. Total calendar impact: maybe -2% to -4% of portfolio. Not great, not terrible (I really need to stop quoting Chernobyl in a post about worst-case scenarios).

This is actually the calendar sleeve doing its job. In the strangle post, I said the worst-case for strangles is a broad vol spike that moves everything directionally. That’s exactly what’s happening. And the calendar sleeve’s long vega exposure is absorbing some of the blow. Not all of it (the delta damage is real), but enough to matter at the portfolio level. Without the calendars, the portfolio would be down 18% total from strangles alone. With them, we’re down maybe 10-12% total across both sleeves on average combined. The calendars are a partial shock absorber. Not perfect (nothing is), but the portfolio is better with them than without them. Calendars likely lose less here, in other words.

The Explosion Positions: Waking Up

Now here’s where the framework earns its keep. Or doesn’t. This is the moment of truth for the thesis.

Remember our Explosion Positions: 5-delta calls and puts on natural gas, wheat, and the yen.

Yen calls: firing. The yen has strengthened dramatically. Our 5-delta yen calls, which cost almost nothing at entry, are now deep in the money, even though we stopped out of our Yen strangle at -2%, we’re super happy here. Our calls are up 8x. If the move continues (and in a genuine crisis, the yen often keeps strengthening as carry trades continue to unwind for weeks), they could go to 15-20x.

Following the management rules: we sell one-third at this level. That one-third alone returns roughly 2.5x the total cost of all six Explosion Positions this month (both calls and puts across all three assets). We’ve covered our basis. The remaining two-thirds ride for free.

Natural gas puts: active but not yet explosive. Natural gas initially spiked on the crisis but has now reversed as recession fears grow (less economic activity = less energy demand). Our NG puts are up maybe 2-3x. Not yet at the first profit-taking threshold. We hold.

Wheat calls: interesting. Wheat has spiked on supply chain disruption fears (the geopolitical event involves a major grain-exporting region, let’s say). Our wheat calls are up 4x. We sell one-third per the rules.

The other three Explosion Positions (yen puts, NG calls, wheat puts): Worthless or near-worthless. The moves went against them. That’s expected. In any given tail event, roughly half your Explosion Positions are on the wrong side. That’s fine. You bought both sides precisely because you can’t predict direction. The ones that hit more than pay for the ones that don’t.

Total Explosion Position P&L this month: after selling the first thirds on yen calls and wheat calls, and accounting for the worthless positions, we’re up roughly 8-15% of portfolio from this sleeve alone, depending on how far the moves go.

The Net: Adding It Up

Let’s tally the month:

Strangle sleeve: -16% (5 planned stops, 1 outsized blow-by stop stop) Calendar sleeve: -3% (rough estimate, partially offset by vega) Explosion Positions: +10% (conservative estimate, could be much higher if moves continue)

Net portfolio: approximately -5%.

That’s a bad month. Let’s not sugarcoat it. Down 5% is not fun. Your account is smaller than it was. The experience of the month was terrible (you watched six strangles blow up, felt the anxiety of the remaining six being under pressure, sweated the calendar circuit breakers, and even though the Explosion Positions are printing, loss aversion means the strangle pain is dominating your emotional state).

But let’s put it in context.

The S&P 500 is staring down -12-15% in this same scenario. A traditional 60/40 portfolio is down 8-10% (bonds helped but not enough). A pure short-premium portfolio without the calendar or tail sleeves is down 18% or more (all the strangle losses, none of the Explosion Position gains or calendar mitigation).

And our -5% includes multiple winning Explosion Positions that still have two-thirds of their size running with zero cost basis. If the crisis deepens over the following weeks (as 2008 did, as COVID did), those remaining positions could turn the month from -5% to flat, or even positive. The asymmetry of the Explosion Positions means the final P&L of a crisis month often isn’t known until weeks after the initial shock.

This is, I want to stress, a realistic scenario. Not a fantasy. Not an optimistic projection. This is roughly what would have happened to this portfolio structure in the first two weeks of COVID if you’d been running it in February 2020. Some version of this will happen again. Maybe this year, maybe in three years. The point is that the framework is designed to survive it with smaller percentage losses while the rest of the world is in significantly more pain.

What We Do Next (This Is the Hard Part)

So it’s week 2 of the crisis. The portfolio is down 5%. Six strangles have been stopped out. The market is in chaos. CNBC is using words like “unprecedented” (it’s not, something else unexpected literally happened like 5 years ago, but sure).

Here’s what we do:

We re-enter strangles. Yes, really. The six positions that stopped out freed up capital and margin. And here’s the thing: IVR across the board is now extremely elevated. The VRP (variance risk premium) is wide. The premium available for selling is the richest it’s been in months or years. This is exactly when the strangle sleeve has its highest expected value. The rules say sell premium when IVR is high. IVR is very high. We sell.

This feels insane, by the way. You just got punched in the face by six losing trades and now I’m telling you to walk back into the same ring. Everything from the previous post about loss aversion and the frequency illusion is screaming at you to sit on the sidelines and “wait for things to calm down.”

But the data is clear. Post-crisis premium selling has historically been among the most profitable environments for short strangles. Implied volatility overshoots realized volatility by the widest margin during and immediately after crisis periods. The fear is priced in at a level that dramatically exceeds the actual subsequent movement (because things usually DO calm down, usually quite quickly, and the options market is slow to reflect that). Selling strangles here is like selling insurance right after a hurricane, when everyone wants it and the premiums are sky-high.

We check the Convexity Score for new Explosion Positions. After a crisis, some tail options are now expensive (the ones in the affected markets, where vol has spiked). But others, in markets that weren’t directly involved, may have actually gotten cheaper (vol spillover is uneven). Maybe copper options or corn options are now screening as the best convexity buys because their tails are still underpriced while everyone’s attention is on equities and crude. We follow the math, not the narrative.

We continue holding the remaining two-thirds of the winning Explosion Positions. The yen calls, the wheat calls. These are still in play. The crisis might deepen. The discipline here is to let the winners run, per the rules, even though the urge to “take what you’ve got” is overwhelming after a stressful week.

We do NOT change the framework. We don’t increase sizing to “make back the losses.” We don’t reduce sizing because we’re scared. We don’t skip the Explosion Positions this month because “we already had a tail event.” We don’t add extra strangles because “premium is so rich right now.” We run the same process, with the same parameters, at the same sizing. The rules were made for moments like this. They were designed by our calmer, more rational past selves. We trust them.

The Months After

Here’s what historically tends to happen in the 2-3 months following a crisis:

Implied volatility slowly bleeds lower as the acute fear subsides. The strangles we entered at peak IVR have extremely high win rates and oversized premium. The strangle sleeve goes on a run. Three, four, five consecutive winners, each one fatter than the winners in a normal environment.

The Explosion Positions from the crisis month either continue paying off (if the crisis extends) or expire with the remaining two-thirds intact (if the crisis was a sharp V-bottom). Either way, the portfolio captured the asymmetric payoff on the way down and now captures the elevated premium on the way back up.

The calendars likely suffer from the declining vol, but again serve as a hedge since we cannot predict when the crisis will end.

Within 2-4 months, the portfolio has typically recovered the drawdown and then some. The crisis that felt like an existential threat to the strategy in real time turns out to have been the strategy’s best friend, because it created the conditions (high VRP, tail payoffs, elevated premium) that the framework is designed to exploit.

The Scenario We Don’t Survive

I want to be honest about this too, because intellectual honesty is the whole point of writing in public.

The framework has a failure mode. It’s unlikely, but it exists.

If a crisis produces extreme moves in every market simultaneously AND those moves are sustained (not a V-bottom) AND the Explosion Positions happen to be in the wrong three assets that month (the top 3 on the Convexity Score were, say, copper, corn, and the Australian dollar, but the crisis hits equities, bonds, and energy), then you can have a month where the strangles stop out, the calendars lose their structure, AND the Explosion Positions expire worthless because the moves happened in markets you didn’t have tail exposure to.

In that scenario, we could be looking at maybe -20% to -30% in a single month. That’s a serious drawdown. Not portfolio-ending (the 2% sizing caps individual losses), but bad enough that recovery takes quarters, not weeks.

The mitigation is the diversification rule in the Explosion Position selection. By requiring positions across different sectors, you reduce the probability that the tail event misses all three of your positions. But you can’t eliminate it entirely. With many markets in the universe and only 3 selected per month, there’s inherently a chance that the event strikes where you’re not positioned.

This is the real risk. Not the risk that the thesis is wrong (I’m confident in the data so far, that’s why I’m writing). The risk that in any given month, you’re buying the right thing at the wrong time, or in the wrong place. Over many months, this averages out (you cycle through the entire universe and the tails appear where the data says they appear). But in any single month, you can be unlucky.

The poker analogy from last post applies perfectly. You can make the correct call with 38% equity and lose. You can make the correct Explosion Position selection based on the Convexity Score and have the tail event miss you this month. The decision was right. The outcome was bad. Variance is not signal.

The question is whether you can make the same call next month.

The Takeaway

The framework isn’t designed to win every month. It’s designed to lose in a controlled, survivable way when things go wrong, and to recover aggressively when conditions normalize.

The 2% strangle sizing means that even a cluster of simultaneous stops is a manageable hit, not a catastrophe. The calendar sleeve’s long vega partially absorbs the shock. The Explosion Positions provide asymmetric upside during exactly the moments when the other sleeves are hurting. And the post-crisis environment creates the richest premium-selling opportunities, which the framework re-enters per its rules.

The hardest part of all of this isn’t the math or the portfolio construction. It’s the discipline to follow the process during the worst moments. To re-enter strangles after getting stopped out six times. To buy more Explosion Positions after watching a month of them expire worthless. To hold the winners when every instinct says to take the money and run.

That’s why I wrote the psychology post before this one. Because this post is, ultimately, about what happens when the theory meets reality and reality punches you in the mouth. The framework is designed to take the punch. The question is whether you are.

Let’s see what happens.

Subscribe to Leptokurticapital. Theory and Practice.

It’s the final table of the World Series of Poker Main Event. Nine players left. Millions of dollars on the line. ESPN cameras everywhere. Your mom is watching.

You look down at your cards. Two sevens. Not great, not terrible. You’ve got a medium stack of chips. The big blind is getting expensive, and eventually you’ll be eliminated. You need to make something happen soon.

The player in front of you, naturally wearing sunglasses indoors, as one does, shoves all-in for a 66% pot sized bet. He has the same number of chips as you. Yuck.

You think about it. Your 77s are pretty good here. Based on his range of hands (the hands he’d do this action with) in this spot (he’s been aggressive all night, shoving all in a lot), you estimate pretty confidently you’ve got about 38% equity (aka 38% of the time you call, you’re winning against his hands).

Here’s some math to ponder. Let’s say the bet is offering you 2.5 to 1 on your money. You need about 28% equity to make this call profitable. You’re pretty sure you have 38% based on all information it is possible to have here. With these assumptions, this is a clear, mathematically correct call. Positive expected value. Not even close, actually. Any competent poker player takes this in their sleep.

And yet.

You’re going to feel sick making this call. Your hands will shake. The commentators might even question it. Your mom will definitely question it. Because you’re about to put your tournament life on the line in a spot where you lose 62% of the time. Almost two out of three times, you’re walking away from the table.

The call is correct. The math is unambiguous. But every atom of your psychology is screaming at you to fold and “survive.” To wait for a “better spot.” One where you’re “more sure.”

This is the exact same psychological trap that makes tail options chronically underpriced. And understanding why is, I think, pretty crucial to holding the positions that make this strategy work.

The Frequency Illusion

Here’s what your brain does with probabilistic information. It takes the frequency of an outcome and treats it as the importance of that outcome.

If you make the correct poker call and lose (which happens 62% of the time, remember), your brain records: “That was a bad decision. I lost.” It doesn’t record: “That was a good decision that happened to have a negative outcome this particular time.” It especially doesn’t record: “If I make this exact decision 1,000 times, I make a fortune.” Your brain doesn’t think in distributions. It thinks in narratives. And the narrative of “I called all-in and lost” is a much more vivid story than “I made a positive-EV decision.” Positive EV decisions are often boring, painful, or both. But they are correct. And we should seek them.

Now multiply all this by the stakes involved in trading. You buy Explosion Positions for three months straight. Each month, they expire worthless. Three months of seeing money evaporate. Your brain is screaming: “This doesn’t work. Stop lighting money on fire.” Or the less intense but more insidious “Maybe my models are just wrong”. (Yes, they are. But that’s actually the point sometimes). It doesn’t matter that the math says the expected value is positive. It doesn’t matter that you’re buying tails that are empirically 10-20x more frequent than the pricing assumes. What matters to your lizard brain is that you have lost money three times in a row, you’re likely to lose 3 more times in a row, and it would very much like you to stop because losing money now = pain = wrong. Ouch.

This isn’t a character flaw. It’s a well-documented cognitive bias called the frequency illusion (closely related to availability bias). We judge the probability and importance of events based on how easily we can recall them happening. Three consecutive months of worthless options are three vivid, painful memories. The hypothetical future payoff of 50x is an abstraction. The memories win every time. Our brains are designed this way, and it’s hard to combat.

This bias isn’t unique to trading. It shows up everywhere in how humans process risk.

We Are Famously Terrible at Estimating Danger

Quick: which kills more Americans per year, sharks or vending machines?

Vending machines. And it’s not close. Vending machines kill about 2-3 people per year (they tip over on people, probably while they’re shaking it to get a candy bar, which is a frankly a rough way to go). Sharks kill about 1 person per year in the US, and even that is on the high side. And yet Jaws didn’t generate $470 million at the box office because people are terrified of vending machines.

This is availability bias in its purest form. Shark attacks are vivid, terrifying, heavily covered by media (Shark Week is an entire television franchise). Vending machine deaths are mundane and unreported. So we massively overestimate shark risk and massively underestimate vending machine risk, despite the data being clear.

Some other fun ones:

People are more afraid of flying than driving, even though driving is roughly 100 times more dangerous per mile traveled. The vivid mental image of a plane crash overwhelms the statistical reality of car accidents. Plane crashes make the news. Car accidents don’t (unless they involve celebrities, in which case we’re back to availability bias).

People in the US consistently overestimate the probability of dying from terrorism and underestimate the probability of dying from heart disease. Heart disease kills roughly 700,000 Americans per year. Terrorism has killed about 100 Americans per year on average over the past two decades (heavily skewed by a single catastrophic event). But terrorism is vivid, emotional, narratively compelling. Heart disease is slow, boring, and happens to everyone’s uncle.

People dramatically overestimate the danger of nuclear power and dramatically underestimate the danger of air pollution. Nuclear accidents are spectacular and terrifying (Chernobyl, Fukushima). Air pollution silently kills millions per year globally. The expected deaths per unit of energy generated are far lower for nuclear than for coal or natural gas, but try telling that to someone’s amygdala, especially when they’re being paid to not believe you.

The pattern is always the same: vivid, narratively compelling, emotionally charged events are overestimated. Slow, boring, statistically significant events are underestimated. Frequency of experience trumps magnitude of impact in our intuitive risk assessment.

How This Shows Up in Options Pricing

Now here’s where I bring this back to what we’re trying to do (sorry for the detour, but I think the groundwork matters, and I had fun).

Options pricing is, at its core, a collective statement about the probability of future events. Implied volatility is the market’s aggregate estimate of how much something will move. The skew (how much more expensive puts are than calls, or vice versa) reflects the market’s estimate of directional tail risk. The smile (how much more expensive deep OTM options are relative to at-the-money) reflects the market’s estimate of tail fatness.

These aren’t computed by a single rational agent with perfect information. They’re the emergent result of thousands of participants making decisions, each of whom is subject to the same cognitive biases we just discussed, especially when volume is thinner and market makers are less efficient.

The result is predictable:

Vivid, emotionally charged tail risks are overpriced. SPX puts after a recent crash. VIX calls when volatility has been elevated. Oil options right after a geopolitical event (see: Iran recently). In each case, the recent vivid experience of a tail event makes participants overweight the probability of another one. The options get bid up above fair value. (This is actually great for the strangle sleeve. When everyone is scared, premium is expensive and we sell it.)

Invisible, boring, historically-documented-but-not-recently-experienced tail risks are underpriced. Wheat calls during a quiet growing season. Yen calls during a period of stable BOJ policy. Natural gas puts when winter has been mild. Nobody is thinking about these tails. They’re not on CNBC. There’s no vivid recent memory driving the pricing. The models fill the vacuum with their Gaussian assumptions, and the options sit there, underpriced, waiting for the tail event that the data says will come far more often than the price reflects.

This is why the mispricing persists even though the data showing fat tails has been publicly available since Mandelbrot in 1963. Knowing the data is not enough. You have to be psychologically capable of acting on it. And acting on it means buying things that lose money most of the time, in markets nobody is talking about, based on statistical frequencies that your brain is literally evolutionarily wired to underweight.

Most people can’t do it. Not because they’re dumb, but because they’re human.

We can do it with math, and some discipline. Light work.

The Poker Players’ Edge (And Ours)

Back to the poker table.

The thing that separates professional poker players from amateurs isn’t hand-reading or bluffing or any of the stuff that makes for good television. It’s the ability to make positive-expected-value decisions over and over, regardless of recent outcomes, regardless of emotion, regardless of anything. To call all-in with 38% equity and lose, and then do it again the next time the math says to. Without flinching. Without “adjusting” your strategy based on a sample size of one.

Professionals call this “playing your A-game.” It means executing the mathematically optimal strategy, informed by all information you have access to and can process, regardless of whether you’re running hot or cold. It means understanding that variance is not signal. A string of losses (often) doesn’t mean the strategy is broken. A string of wins (usually) doesn’t mean you’re a genius. Both are just expected fluctuations in a probabilistic system.

This is almost exactly the discipline required to run the Leptokurticapital framework.

The strangle sleeve will have losing months. Sometimes several in a row. This doesn’t mean the variance risk premium has disappeared. It means you hit the wrong side of the distribution for a while (which happen about 17% of the time, the win rate is 83%, not 100% after all).

The Explosion Positions will expire worthless most months. Month after month of premium evaporating. This doesn’t mean the tails aren’t fat. It means the tails haven’t been triggered yet (which is what “infrequent but more frequent than expected” means in practice).

In both cases, the edge is real but the experience of the edge is psychologically adversarial. The strategy is designed to be profitable. Your brain is designed to make you abandon profitable strategies that feel bad in the short term. The tension between these two facts is, in my opinion, the hardest part of systematic/quant trading. Not the math. The math is the easier part (relatively speaking, anyway, I did go to school through 18th grade, and have always struggled with math).

Loss Aversion and the Asymmetric Experience of P&L

There’s one more cognitive bias worth naming, because it’s especially relevant to how we experience the three-sleeve portfolio.

Kahneman and Tversky (the titans of behavioral economics, if you’re not familiar, their book “Thinking Fast and Slow” is very much worth a read) demonstrated that humans feel the pain of a loss about twice as intensely as the pleasure of an equivalent gain. Losing $100 feels roughly as bad as gaining $200 feels good. How interesting. This is called loss aversion and it’s one of the most replicated findings in all of behavioral science. Most people act this way.

Now think about how this interacts with a portfolio that wins small (consistent strangle premium) and occasionally loses big (strangle stops hit) while simultaneously bleeding small (Explosion Position decay) and occasionally winning huge (tail event payoff).

In a typical month, you collect, say, $500 in strangle premium and lose $300 in Explosion Position decay. Net $200. Fine, but not exciting. The $300 loss on the Explosion Positions feels about as significant as the $500 gain on the strangles, even though the net is positive. Your emotional P&L is roughly zero even though your actual P&L is positive.

In a bad month, you lose $1,000 on strangle stops and $300 on Explosion Position decay. The $1,300 total loss creates a vivid, painful memory. Loss aversion amplifies it. This single month weighs more in your psychological accounting than three good months of +$200.

In a great month (tail event), you lose $1,000 on strangle stops but make $5,000 on an Explosion Position. Net $4,000. This should feel amazing. But loss aversion means the $1,000 strangle loss still hurts, even as the Explosion Position is printing. You’re up huge on the month but you’re fixating on the losing trades.

This is not rational. But it is human. And last I checked, we’re all still human here. And it’s why many people who try strategies like this eventually abandon them. The psychological experience of the strategy is worse than the mathematical reality of the strategy. The numbers say it works. Your nervous system says it’s torture.

The only solution I know is radical transparency with yourself (and, in this case, with all of you). Track the numbers obsessively. Look at the portfolio-level P&L, not the individual position P&L. Review the data regularly. And remind yourself that you’re playing a game where the edge reveals itself over hundreds of trades, not in any single month.

Or, failing all of that, become a robot. I hear they’re great at executing systematic strategies without emotional interference (I may end up doing this one day, honestly, it’s a natural evolution of the framework).

Why This Means the Edge Persists

Here’s the optimistic conclusion to all of this.

The mispricing we’re exploiting isn’t just a mathematical artifact. It’s a psychological artifact. It persists because the cognitive biases that create it are features of human neurology, not temporary market conditions. Availability bias isn’t going away. Loss aversion isn’t going away. The frequency illusion isn’t going away. As long as humans are making the marginal pricing decisions in options markets (even with algos, a human set the parameters, decides when to turn it on, how much capital to allocate, etc etc. We’re far from true “efficient” AI pricing in this context), these biases will continue to distort prices in predictable, exploitable ways.

The vivid tails (equity crashes, recent crises) will continue to be overpriced because participants can easily recall them. The boring tails (wheat supply shocks, currency interventions, energy disruptions) will continue to be underpriced because participants aren’t thinking about them.

And the majority of traders who stumble onto positive-EV strategies like this one will continue to abandon them after a string of losses, because their psychology won’t let them continue doing something that feels bad regardless of what the data says.

Which is fine. If everyone could do this, the edge wouldn’t exist.

The question isn’t whether you understand the math. If you’ve read this far, you almost certainly do. The question is whether you can sit with the discomfort of a strategy that’s correct on average but painful in the moment. That’s the real barrier. And it’s one that no amount of quantitative sophistication can solve. Only self-awareness and discipline.

Or robots. Robots could also work.

Next post: when everything goes wrong at once. A realistic scenario where multiple sleeves bleed simultaneously, and how the framework is designed to survive it (spoiler: it mostly involves following the rules even when the rules feel insane, since we made the rules for a data-backed reason).

Subscribe to Leptokurticapital. We talk about feelings here, apparently. But also math. Mostly math.

I think one of the gaps in trading education is that you either get the theory people (professors, quant bloggers, people who’ve read a lot of papers but don’t show you their account) or you get the execution people (here’s my P&L screenshot, trust me bro, no explanation of why). I want to do both. Every trade, the thesis AND the receipt.

Let’s get into it.

The Trade

Underlying: /ZC (Corn Futures) Structure: Short strangle (20-delta both sides) Strikes: 425 put / 495 call Expiration: April 24, 2026 (52 DTE at entry) Entry date: March 9, 2026

Premium received:

• Put: 3.25 ($0.0325/bushel × 5,000 bushels = $162.50)

• Call: 4.625 ($0.04625/bushel × 5,000 bushels = $231.25)

• Total credit: $393.75

Management rules (same as always):

• Profit target: 50% of credit received → close at ~$197 profit

• Stop loss: 2x credit received → close at ~$787.50 loss

• Time stop: close at 21 DTE regardless of P&L

Quick Return Math

A note that my average hold time for these trades is around 25DTE (since sometimes we enter a bit longer than 45DTE depending on expiration availability). This means:
Margin from my broker = ~$780, * 2 for safety = $1560.

197 / 1560 = ~12.6% over 25 days or so, which is about 0.5% return per day on a successful position, less some for fees.

Note also that this strategy has performed about 0.19% return on capital per day for me, accounting for losing positions!

Why This Trade, Why Now

If you read the oil post earlier this week, you know the world is in the middle of a genuine geopolitical crisis. Oil is flirting with $100/barrel. The Strait of Hormuz is (maybe?) closed. Markets are in risk-off mode.

So why am I selling a strangle on corn?

Because IV on corn is elevated right now, and not because of anything specific to corn. The global instability from the Iran situation has spilled over into implied volatility across the commodity complex. Corn IV spiked in sympathy with the broader fear, even though the fundamental supply/demand picture for corn hasn’t changed dramatically. Corn isn’t shipped through the Strait of Hormuz (most US corn moves domestically or through Gulf of Mexico ports).

This is exactly the kind of IVR environment the framework targets. Implied volatility is high relative to what corn is likely to actually do over the next 52 days. The variance risk premium is wide. We’re getting paid a fat credit for a strangle that has wide strikes.

The 425/495 strikes give us a range of $70 on either side of the current price. Corn needs to move roughly 15%+ in either direction before either strike is breached. Could it? Sure, anything can happen (we literally just watched oil move 70% in a week, so I’m not going to claim anything is impossible). But the base case for corn over the next 7 weeks is that it chops around in a range, IV bleeds back toward normal, and we collect our premium.

Current Status (as of 3/15/2026)

The position is roughly breakeven, maybe slightly positive. Two things have happened since entry:

IV has come down about 15 points. This is good for us. We sold the strangle at elevated IV, and as the initial panic has subsided slightly, the options have lost some of their vol premium. This is the variance risk premium doing its thing in real time. We sold expensive insurance, and it’s getting cheaper.

The underlying has moved up a bit. This is slightly against the call side but nothing threatening. We’re well within the expected range. The call at 495 is still far from being tested.

Net effect: roughly flat. The IV decline helps us, the directional move hurts us slightly, they roughly offset. This is a normal early-life strangle. Nothing exciting, which is exactly what we want.

Account Update

This is the first position in the live tracking. The account is small (10k doesn’t give us much margin to play with yet), so position sizing isn’t at the ideal 2% per position that I’ve discussed in the theory posts. At this account size, each position is a larger percentage of the portfolio than I’d like. That’s a known limitation I’ve been upfront about, and it’s something that improves as the account grows.

I’ll be tracking cumulative P&L, win rate, and per-sleeve performance as the trades accumulate. One trade isn’t a track record. But every track record starts with one trade.

What I’m Watching

IV continuing to normalize. If the geopolitical situation stabilizes (or at least stops escalating), commodity vol should continue bleeding lower. Every point of IV decline benefits this position. This is the most likely path and the one where we collect our 50% target in the next 2-4 weeks.

Corn-specific catalysts. USDA Prospective Plantings report comes out March 31. This is a significant report for corn and could cause a directional move. We’ll be at roughly 26 DTE at that point, which is getting close to the 21 DTE time stop. If the position hasn’t hit the 50% target before the report, I may close before the number drops rather than holding through a binary event at reduced DTE.

Broader market contagion. If the Iran situation escalates further and commodity vol spikes again, this position could come under pressure. The 2x stop is there for exactly this scenario. If it triggers, we take the loss and move on. That’s what mechanical management is for.

The Template Going Forward

This is what trade recaps will look like. Entry parameters, thesis, current status, what I’m watching. When trades close (win, loss, or time stop), I’ll post the exit with final P&L and any lessons learned.

Over time, these recaps build into a real track record. Every trade documented, every decision explained, every outcome recorded. That’s the commitment.

More soon. If nothing dramatic happens, the next update on this position will be in next week’s recap, alongside any new entries.

Subscribe to Leptokurticapital for weekly trade recaps alongside the theory posts. This is where the framework meets reality.

Subscribe now

One of the most common questions I get when I explain the Explosion Position framework is some version of: “If the tails are really underpriced, why hasn’t someone already arbitraged that away?” (or the more common “what are you talking about” that’s common too).

It’s a good question. And the answer is one of my favorite topics in all of finance, because it has almost nothing to do with math and almost everything to do with understanding who is actually in these markets and what they’re trying to accomplish. Human psychology and motivation.

In short: the people trading commodity and currency futures options are doing something fundamentally different than the people trading equity options. Their motivations are different. Their risk tolerances are different. The flow patterns they create are different. And those differences show up directly in how the tails are priced.

This is, I think, one of the most underappreciated edges available to anyone willing to look beyond equity markets. And conveniently, you can actually see the evidence for it in public data every week.

Two Kinds of Hedgers, Two Kinds of Prices

Let’s compare two participants.

Participant A is a pension fund. They manage $50 billion in equities. Their board has a mandate to limit drawdowns. Their risk management team buys SPX puts every quarter, systematically, as a matter of policy. They are not price sensitive in any meaningful way. They need the protection. They’ll pay whatever the market charges. If the puts are expensive this quarter, they buy them anyway. If they’re even more expensive next quarter, they buy them anyway. The mandate says “hedge tail risk.” So they hedge tail risk, in a way that makes sense and is executable at billion dollar scale.

This creates a persistent, price-insensitive bid for SPX downside. Market makers know the flow is coming. They price accordingly. SPX put skew stays steep. The tails stay expensive. It’s well documented. It’s been this way for decades and it’s not going to change because the structural incentives are permanent.

Participant B is a wheat farmer in Kansas. They grow wheat. They want to lock in a price for their harvest so they can plan their finances. In the fall, they sell wheat futures or buy wheat puts to hedge their crop. Their hedging is concentrated around planting and harvest seasons. They’re hedging the body of the distribution (normal price fluctuations that could affect their margins), not the tails. They are not buying far out-of-the-money options because a 5-sigma wheat spike is not what keeps them up at night. What keeps them up at night is wheat dropping 15% before they can sell their crop.

The wheat farmer’s hedging creates demand for at-the-money and slightly out-of-the-money puts during specific seasonal windows. It does not create demand for deep out-of-the-money options. Nobody in the physical wheat supply chain is systematically buying 5-delta wheat calls as tail protection the way pension funds systematically buy 5-delta SPX puts.

And that’s the gap (I think, as always). The deep tails in wheat (and soybeans, and corn, and natural gas, and copper, and the yen) don’t have, or don’t have as many, structural buyers driving up prices. The options that expire far from the money are priced primarily by market makers using models that, as we’ve discussed at length, underestimate tail frequency. Without a flow of price-insensitive institutional buyers pushing those prices higher, the model-based price persists.

The tails are underpriced not because markets are inefficient in some abstract sense, though humans are bad at estimating and pricing tail risk generally. They’re also underpriced, I argue, because less size is buying them.

The Commitment of Traders: A Window Into Who’s Doing What

You can actually observe these dynamics in a public dataset released every week by the CFTC (Commodity Futures Trading Commission). It’s called the Commitment of Traders report, or COT.

Every Friday, the CFTC publishes the aggregate positions of three categories of participants in every major futures market:

Commercials. These are the hedgers. The wheat farmer. The oil producer. The airline hedging fuel. The mining company hedging copper. They have a physical business that’s exposed to the commodity price, and they use futures and options to manage that exposure. They are in the market because they have to be, not because they think they’re smarter than the market.

Large Speculators (Non-Commercial). These are hedge funds, CTAs (commodity trading advisors), and other institutional speculators. They’re expressing directional or volatility views. They’re the closest thing to “smart money” in these markets, though (hot take) that label is often less deserved than it sounds.

Small Speculators (Non-Reportable). Retail traders. You and me (well, sort of). Smaller positions that don’t meet the CFTC’s reporting threshold.

The interesting thing for our purposes is the Commercial positioning, and specifically how it differs from what you see in equity markets.

In equity index options, the dominant commercial flow is buy-side. Institutions are buying puts for protection. This creates the net buying pressure on tails that makes SPX puts more expensive, and often, more efficiently priced.

In commodity futures, the dominant commercial flow is often sell-side. Producers hedge by selling futures or selling calls. A soybean farmer is selling forward production. An oil producer is selling crude futures against their reserves. A gold miner is selling gold calls against their output. The commercial flow is net short, putting downward pressure on futures prices and creating selling pressure (not buying pressure) in the options market.

This is a structurally different ecosystem. In equities, the commercial flow inflates tail option prices. In commodities, the commercial flow either deflates them or simply doesn’t interact with the deep tails at all. The 5-delta soybean call or the 5-delta natural gas put exists in a relative pricing vacuum, populated mainly by market makers and their models.

You can watch this play out in real time. Pull up the COT data for wheat, soybeans, or crude oil. Look at the Commercial net position. It’s (almost) always net short (producers hedging output). Now look at the Commercial net position in S&P 500 futures. It oscillates, but the options flow is dominated by put buying. The difference in who is doing what, and why, directly maps to the difference in how the tails are priced.

Seasonal Patterns and the Timing of Hedging Flow

Here’s another layer that I find fascinating and that plays directly into the Explosion Position framework.

Commercial hedging in commodities is seasonal. Farmers hedge around planting and harvest. Natural gas producers and consumers hedge around winter (heating demand) and summer (cooling demand). This means the hedging flow, and therefore its impact on options pricing, is concentrated in specific calendar windows.

Outside those windows, options activity thins out. Market makers still quote prices, but the flow they’re pricing against is lighter. The models dominate the pricing even more than usual. And the models, as always, assume thinner tails than reality.

This creates periodic windows where tail options are even cheaper than their already-cheap baseline. In the late spring, after the major grain planting decisions are locked in but before summer weather risk materializes, grain options can get particularly cheap. In the shoulder seasons for natural gas (fall and spring), when neither heating nor cooling demand is driving hedging activity, NG tail options can be almost laughably underpriced relative to the actual frequency of extreme moves. And since we can’t predict Black Swan moves, we don’t try to. We just by the tails when they’re underpriced consistently.

I’m not (yet) systematically timing my Explosion Position entries to these seasonal windows, but it’s something I’m looking at. The Convexity Score already captures some of this implicitly (when options are cheaper, the score goes up), but there might be additional edge in being more deliberate about when in the calendar you’re buying specific commodities. This is one of those areas where the framework is still evolving.

The COT as a Signal (With Caveats)

Some traders use the COT data as a direct trading signal. The classic interpretation: when Commercials are extremely net short, they’re hedging at levels that historically represent price ceilings. When they’re less net short (or net long, which is rare), the market has room to run higher. The theory is that the people with the most intimate knowledge of the physical supply/demand balance (the commercials) are expressing that knowledge through their hedging activity.

I think there’s something to this, but I want to be careful about how much weight I put on it until I prove it out with my own experience over time.

For the Explosion Position framework specifically, the COT data is more useful as a structural indicator than a timing signal. What I mean by that:

If Commercial positioning in a given market is heavily net short (lots of producer hedging), it tells me that the at-the-money and near-the-money options are being actively traded by hedgers, but the deep out-of-the-money options are probably NOT being actively traded by hedgers. The tails are in model-pricing territory. The Tail Richness Ratio is more likely to represent a genuine mispricing rather than an efficiently priced risk.

Conversely, if I ever saw a commodity market where Commercials were actively buying deep out-of-the-money options (as a group, systematically), I’d be skeptical of the Convexity Score for that market. The flow would be pushing tail prices toward fair value, or even above it, the same way institutional put buying pushes SPX tail prices above fair value.

So far, I haven’t seen that in any of the commodity markets I screen. The Commercial hedging flow is overwhelmingly concentrated in the body of the distribution. The tails remain a model-priced, low-flow, structurally underpriced corner of these markets.

A Quick Detour on Market Makers

It’s worth briefly talking about the people who actually set the prices on these deep out-of-the-money options, since in most cases it’s not Commercials or Speculators. It’s market makers.

Market makers in commodity options are running books. They quote bids and offers, they manage their Greeks, and they try to be flat at the end of the day (this is their version of “hedging”, except they’re hedging the flow of options orders, not a physical commodity). They price the deep tails using models. Usually something like Black-76 (the futures variant of Black-Scholes) with some skew adjustment.

Here’s the thing about market makers: they’re not incentivized to price the tails accurately. They’re incentivized to make the bid-ask spread and manage their risk in the near term. If they set the price of a 5-delta natural gas call too low by a factor of 3 (relative to actual tail frequency), they don’t realize that error for months or years. The option either expires worthless (most of the time, even with the mispricing) or the market maker gets hit on a tail event and chalks it up to bad luck.

The feedback loop for correcting tail mispricing is incredibly slow. The model says the price should be X. The option expires worthless. The model is “confirmed.” Repeat for 50 months. On month 51, a 4-sigma event happens, the market maker takes a loss, and they maybe adjust their skew parameter slightly. Then the slow cycle of confirmatory expirations begins again.

Compare this to at-the-money options, where the feedback loop is fast. The market maker sells an at-the-money straddle and within days knows whether the implied vol was too high or too low. They adjust constantly. The pricing converges toward efficiency quickly. For tail options, that convergence barely happens. The sample sizes are too small and the feedback is too slow.

This is another structural reason the tails stay mispriced. It’s not that market makers are stupid. It’s that the incentive structure and information environment in which they operate doesn’t force them to price tails accurately.

What This All Means for the Framework

Let me tie this back to the portfolio.

The strangle sleeve works because the variance risk premium is a structural feature of all options markets. Hedgers overpay for insurance. We collect that overpayment.

The Explosion Position sleeve works (in theory, and I believe in practice, but I acknowledge I’m still building the evidence) because the specific structure of commodity and currency hedging creates a pricing environment where the deep tails are systematically underpriced. The participants who would correct this mispricing (by buying the cheap tail options until the price rises to fair value) largely don’t exist, or don’t exist in big volume. Model adjustment here is a very slow feedback loop for tails. Pension funds don’t buy 5-delta wheat calls. CTAs who might have the sophistication to identify the mispricing are usually focused on trend-following or mean-reversion, not systematic tail buying. Retail traders who buy options are overwhelmingly focused on equities (where the tails are expensive, not cheap).

So you’ve got a mispricing that persists because:

The models used to price it are wrong (Gaussian assumptions). The hedgers who dominate the market don’t interact with the tails. The market makers who price the tails have slow feedback and no large incentive to correct the error. The speculators who could correct it aren’t looking at these markets for this purpose, aren’t sophisticated enough, or do it poorly.

Is this a permanent edge? Probably not in the truly long run. If enough systematic capital flows into cross-asset tail buying, the prices will adjust. But “enough capital” is a lot. The commodity options markets are (relatively) small compared to equity options. The flow required to meaningfully reprice the tails of all 18 futures in my universe would be substantial. And right now, as far as I can tell, very few people are doing this systematically (in public, anyways).

Which is part of why I’m writing about it, honestly. If I’m right, the more people who understand the thesis, the more capital flows toward the mispricing, and eventually it gets corrected. But the correction would be slow and I don’t think writing a Substack about it is going to move the market any time soon (unfortunately for my ego, fortunately for my portfolio). My hope is that the community here is smarter than me. Please poke holes. We’re all students. So…

Homework (Yes, Really)

If you want to start exploring this yourself, here’s a concrete thing you can do this week:

Go to the CFTC’s Commitment of Traders page (it’s public, free, and updated every Friday: https://www.cftc.gov/MarketReports/CommitmentsofTraders/index.htm). Pull up the disaggregated report for a few of the futures I trade: wheat (/ZW), soybeans (/ZS), crude oil (/CL). Look at the “Producer/Merchant/Processor/User” category (these are the Commercials). Note their net positioning.

Then think about what that positioning implies for who is buying and selling the deep out-of-the-money options in that market. Are the commercials likely bidding up 5-delta puts? Or are they focused on near-the-money hedges that have nothing to do with the extreme tails?

I think you’ll start to see what I see: a structural vacuum in the tails of these markets that the Convexity Score is designed to exploit.

Next time: why humans are terrible at thinking about tail events (spoiler: it involves the same cognitive biases that make us terrible at everything else probabilistic), and how that creates the psychological conditions for the mispricing to persist indefinitely, even though the data has been publicly available for sixty years, and Wall Street really likes money.

Subscribe to Leptokurticapital. We’re building this thing one post at a time. We are all students.