Things That Come Back

Why mean reversion is the engine under the strangle sleeve, and why it quietly stops working at the worst possible time

Apologies for the delay in posting! I do in fact work a full time job, and that’s been sucking up most of the bandwidth lately. I owe trade updates, and those are in progress. I aim to have 1 theory post and 1 trade recap post each week going forward.

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.