Leverage Isn’t the Problem. Concentration Is. -

The Rational Reminder recently released another episode on leveraged ETFs. I didn’t want to listen (and do not care to link it :)). They had done one not long ago, and I was fairly sure I wouldn’t care for the angle they would take.

It’s almost 2027. It shouldn’t be news to anyone that humans like to gamble. I imagine a Bill Burr bit: “OMG! They REALLY did that?!” Then he switches character and, in a perfectly anticlimactic voice, says, “Yes. They did.”

Humans like to gamble, and the financial industry has produced, and will keep producing, ETFs for them. There is nothing new under the sun. There is demand, and a capitalist society supplies it. Pretending to be shocked, or recording yet another episode on why gambling is financially stupid, is…a reliable way to keep your show going.

(The other way, and here I’m looking at Animal Spirits, is the Bridget Jones approach: “Yes, gambling is stupid, but I’m an imperfect human myself. The sign on the door says ‘evidence-based shop,’ but the kids have to eat, so here’s our sponsor, an autocallable ETF.” Sorry, I digress.)

Yes, if the Rational Reminder is targeting a wider audience, that’s an understandable way to go. But isn’t a productive use of my time. And yet I listened to the first 20 minutes while commuting yesterday, mostly out of laziness. I needed to find something in the time it took me to run three flights of stairs, and I went with the safe choice.

The guests spent most of those 20 minutes dissecting the issues with the lifecycle investing model. Your job is correlated with the stock market. Most ways of getting leverage are expensive. Nothing new.

Then, to my surprise, one guest conceded that moderate leverage is fine. He meant something like 1.5x (I’m fairly sure about this) at a reasonable cost (I’m less sure about this).

That’s when I arrived at the office, so I was left with a question. Why isn’t that the central point of the episode? Why talk about how people do stupid things when you could talk about this? Wouldn’t that make for a better episode of a show called The Rational Reminder? Maybe they did. I’ll find out.

The stress test

I decided to listen to that episode for another reason. A week ago, someone posted a series on the LEFTS subreddit about stress-testing levered, diversified portfolios. In short: yes, you can beat SPY with less volatility if you combine assets and borrow at a low cost.

What I found most interesting is that the author shows how bonds, gold, and trend following all add real diversification benefits, even after you strip out their excess returns. Many combinations of these diversifiers plus stocks match SPY’s long-term return with a lower maximum drawdown. I think it’s a neat way to show the power of diversification.

Every dot below and to the left of SPY is a worse combination. Every dot above and to the right is a better one. If you don’t see it in this chart, you’re wasting your time reading this blog.

Yes, the sample size is limited. But look at what we’re comparing: a portfolio that goes all-in on one thing (hello, 100% stocks) versus one that doesn’t. That is the real benefit of diversification, and owning 6,000 stocks isn’t the same thing. I don’t know what tomorrow looks like, but I’d rather bet on four horses than one.

I’m also glad the author looked at maximum drawdown, because it’s one of the key risk metrics in retirement. The lower the max drawdown, the lower the sequence-of-returns risk, and the better you sleep while your portfolio funds your life.

Surpriiiiise

I shared the exercise on our subreddit and, to my surprise, a French user replied with a paper he had written. He’s also one of the people behind Cayas.fr, a really nice platform in France. Given how Italians tend to react to English content, I’m glad he spends time on a subreddit that is mostly in Italian.

The paper is a jewel. It addresses every issue I had with the Cederburg series of papers on lifecycle investing. It also puts to bed the objections BigERN raised about diversified portfolios in retirement. To jump to the conclusion: a levered, well-diversified portfolio beats a 100% stock portfolio throughout your life.

Levered means borrowing at a reasonable cost. In 2026, anyone who can type “How do I set up a box spread in IBKR?” into Claude has access to one.

Well-diversified means not stopping at bonds.

If you’re ever asked about the difference between using leverage on a single asset and using it on a diversified portfolio, show them the chart above. This is especially true if the person asking is BigERN. Add leverage, and the diversified portfolio can generate enough return to match 100% stocks (how many times have I written this?).

I’d like to see how many people, Ben Felix included, will address this research.

I expect the main criticism to land on the trend-following strategy built for the paper. Is it too good to be true? Maybe. It’s impossible to say for sure, but some data helps.

The SG Trend and SG CTA indexes are about 25 years old (?). Did they show the useful traits described in the paper? Yes. And the indexes don’t suffer from survivorship bias. If anything, the bias runs the opposite way, because funds tend to join after good performance and leave after bad.

Do you need to be a multimillionaire to invest in this? No. DBMF and RSST show that. You might even capture a bit of “fee alpha” on top.

What about bonds?

The most interesting part of these studies is watching how readers react to the optimization. Of course they do. When you have the data, you have to optimize and find the best mix, right? RIGHT?

Bonds don’t end up in a nice place in these analyses, and their medium-term outlook isn’t bright either. So I often wonder why I’m “defending” them, and why I do it with my money more than with my words. Am I biased? A boomer attached to that glorious 40-year run?

The most recent answer I gave myself is this: we know there is one economic environment, deflation, where bonds work great. Can I predict whether deflation is coming? Probably not.

If assets perform differently in different environments, then the result of any optimization depends on how often each environment shows up. Unless it’s clear that human societies are more likely to produce high inflation than low inflation (and maybe they are), I still see a role for bonds.

And if I kick bonds out?

Then there’s the other side of the decision: what do I put in their place? As much as Nicolas Mirjolet holds a special place in my heart, I can’t see myself going 50/50 with stocks and managed futures, even if I diversify the managed futures sleeve across many providers.

Maybe I’m too attached to the idea of owning lots of candies (diversifiers). The solution might be to go full Roger Nusbaum and use many different alternatives. But that raises three problems:

Each alternative has a short history. I’d like to hold a global macro ETF at some point, and probably will, but I want to see some out-of-sample results first.
Each alternative has its own model risk. I don’t mind holding many ETFs, but even I have a limit.
Bonds are very capital efficient. If I remove them, I need another source of cheap financing. Box spreads solve that, but if possible I’d rather diversify my funding sources too.

The main lesson of these studies is that it doesn’t matter much how much you put in each diversifier, or even which diversifier you pick. What matters is that you own some and keep the cost of your leverage under control. Do that, and you have a good chance of ending up in a better place than the plain-vanilla stock or stock-and-bond investor.

It takes more work, research, and maintenance. Sure. It probably limits where you can live, since it’s not a great strategy in places with high capital gains taxes. But it gives you a lot back in flexibility. Ask BigERN. He’s living it.

What I am reading now:

Follow me on Bluesky @nprotasoni.bsky.social