The TACO trade meets the fog of war

Counting on Trump to "always chicken out" is a risky strategy. And war in the Middle East is a different game.

Still from War Games (1983). Blu-Ray.com.

Now that our NCAA tournament forecasts are out — please go check them out for men and for women — I’ll have a little more time over the next few weeks to catch up on the news and some other stories that are half-finished in the drafts folder. But we’re also well overdue for the March SBSQ. There’s still time to submit questions in the comments of last month’s post.

---

I’m obviously not privy to President Trump’s thinking on why he decided to go to war with Iran.¹ But even among well-connected reporters, there seem to be conflicting accounts on whether the White House and the Department of War anticipated that Iran would seek to effectively block shipping traffic in the Strait of Hormuz, causing oil and gas prices to spike. (Though it’s hard to think they were totally unaware, given that this has been a well-known consequence of attacking Iran since my high school debate days.)

But maybe it’s as simple as this. Trump is a man who has faced remarkably few consequences for his own actions. It’s easier to do what you “feel in your bones” when you don’t bear the downside risks.

Trump has usually gotten away with it

“When you’re a star, they let you do it” has basically been Trump’s superpower. For instance, his strategy of telling off the entire Republican establishment in 2016 actually proved popular with GOP primary voters, defying the conventional wisdom from idiots like me who claimed the primaries are mostly about building intraparty consensus. Then he won the general election when polls had him losing.

Not only were there no real legal consequences to Trump from January 6, but he actually got re-elected four years later! (And everyone seemed to have forgotten about his mishandling of COVID.²) Meanwhile, in the second term, being a lame duck has arguably been freeing for Trump. It will probably be bad for Republicans at the midterms, but Trump has never seemed to particularly care how other Republicans fare when he’s not on the ballot himself.

On the foreign policy front, Trump didn’t face any particularly adverse consequences for nabbing Nicolas Maduro under cover of night. On domestic policy, the Supreme Court sometimes bails him out.

Indeed, “you can just do things” is often a sound approach when you’re playing on a low difficulty level. In poker, we’d call this an exploitative strategy. Game theory will tell you that, if your opponent is playing optimally, you have to make some effort to balance and disguise your strategy. You can’t always bluff or the other guy will wise up. But some guys do always fold.³

And if we’re being honest, Democrats are often like that player who falls for the same trick every time. (I mean, this is literally a party that might nominate fellow Californian and electoral underperformer Gavin Newsom four years after Kamala Harris’s loss.) Furthermore, there’s some degree of context collapse in what news stories draw sustained public attention. The sense one gets is that there’s always a rising tone, an escalating crisis, whether or not that’s actually the case. Breathless coverage of inconsequential stories⁴ blow out the speakers for when there’s a story that should truly raise alarms like war in the Middle East.

The game theory of market behavior isn’t well-resolved

“Markets” sometimes provide more discipline to Trump, whether because of his personal financial interests or because he watches a lot of TV and red downward arrows don’t look pretty on the screen. But I put “markets” in scare quotes because I’ve struggled in this newsletter to operationalize how this actually works in practice:

Wednesday evening’s headlines after the bump in the market were full of happy talk about the “Trump put”. But the celebratory tone already looks premature. The term is borrowed from options trading — a “put” is an option to sell⁵ a stock at a specified price that’s typically lower than its current value, which caps your downside risk. So more broadly, the “Trump put” is the idea that Trump will back down if markets have too much of a tantrum.

I’ve expressed skepticism of this idea before because it anthropomorphizes “the market” into an entity that has agency and is capable of strategic behavior — when, in fact, the market is composed of individual firms and investors who are on a financial and emotional roller coaster.

TACO (Trump Always Chickens Out) has become the slogan for the “Trump put” thesis that I described above. Trump does something that imperils the United States’ economic interests, whether tariffs or threatening to invade Greenland. The Dow sheds 1,000 points, and he reverses course. This doesn’t seem like a very stable equilibrium, however. If traders know that Trump is going to chicken out, they shouldn’t sell off in the first place; otherwise, you could always profit by “buying the dip”. But if markets don’t panic a little bit, how does Trump get the signal that he needs to TACO?

A game-theory equilibrium would almost certainly reveal that both sides are supposed to employ mixed strategies. In other words, sometimes they might be bluffing, but they can’t always be bluffing or there would be no deterrence. Some percentage of the time, they have to follow through with their threats: Trump to do the thing that markets don’t want, and the markets to actually get past the “freak out” stage into sustained, full-blown panic that might cause irreversible damage.

In a true mixed strategy, the participants in the “game” are supposed to be literally randomizing their actions.⁶ It might actually help Trump in a weird way that his behavior is effectively random in some ways based on the last person he talked to or the last TV segment he watched. Markets, though, would seem to be at a disadvantage because they’re composed of thousands of individual participants and there’s no way for them to coordinate:

Still, even other non-zero-sum “games” like nuclear deterrence rely on some degree of implicit randomization — what Thomas Schelling called “the threat that leaves something to chance”. (Basically, you don’t want to escalate when nuclear weapons are involved because mistakes can be made in the fog of war.)

If investors could get together and say: “every week you keep up with this tariff crap, Donnie, there’s a 5 percent chance we’ll have a panic that triggers a global financial crisis, with unrecoverable long-term damage to the economy,” then maybe that would work if Trump had read his Schelling, which he surely hasn’t. But that’s not how markets work. You can’t half-panic any more than you can be half-pregnant. And even markets could work this way, the strategy entails sometimes pulling the trigger, so you’re playing Russian Roulette.

But in that earlier story, I think I gave short shrift to the idea of Thomas Schelling’s idea of “the threat that leaves something to chance” as it applies to market behavior. Schelling, an economist who was one of the early developers of game theory, especially around nuclear deterrence, proposed the “threat that leaves something to chance” as a mechanism to explain why you don’t want to fuck around and find out when a country has nuclear weapons. It might be true that it would be irrational for them to retaliate with a nuclear strike for some lower-magnitude, more conventional escalatory move. But there can be misunderstandings in the fog of war. The world has only narrowly averted an inadvertent nuclear crisis before.

Back to markets. It might be the case that, even though individual market participants can’t coordinate on a strategy, their behavior is nevertheless effectively chaotic enough to serve as a deterrent. (In the literal sense of Chaos theory: i.e., small changes in initial conditions can produce highly variable and unpredictable results in a sufficiently complex system.) Thus, the market effectively does have a “mind of its own” and behaves randomly for all intents and purposes. There’s a lot that can be said for this theory. But if markets’ behavior is essentially random, it implies that markets sometimes will escalate an initial sell-off and it will cascade into something worse.

Oil prices have been fluctuating wildly, of course, from a steady state of about $65-$70 barrel before Iran⁷ to as high as almost $120, before settling into something in the $90-$100 range recently as of this writing. But some analysts think oil could reach as high as $200 a barrel if the crisis in the Persian Gulf persists for more than another few weeks. At $95 a barrel, or even $120, markets actually are still hedging their bets. These prices imply that Trump probably will chicken out: $120 is closer to the baseline of $70 than to $200-a-barrel oil. But there’s a credible threat that he does not. I’m not sure that’s so irrational, even if prices at any given moment can become unmoored based on market psychology.⁸

Trump and markets aren’t the only players with something at stake

Or, the matter might be out of his hands. Tariffs, a previous source of market anxiety, are unusual to some degree because, especially before the SCOTUS ruling, they’re something that more or less could be turned on or off with the literal stroke of the executive’s pen.

Sure, there might be some purely market-based mechanisms for moments of anxiety over tariffs to spiral into something more, like from bond markets panicking. But when you can’t just press the “UNDO” button — we’ve already killed Iran’s leader — there are far more ways for things to go wrong, especially in a multilateral “game”.

Iran has a say, for one thing. If it believes the best way to deter Trump is by triggering a decline the markets and/or a spike in his unpopularity ratings because oil and gas prices are surging, it has every incentive to keep oil prices surging. Or a country like China could try to take advantage of overstretched American military capabilities. And the United States didn’t go to war alone; we’re partnered with Israel, which reportedly threatened to proceed unilaterally with or without us.

Administration officials, including Secretary of State Marco Rubio, briefed lawmakers on Monday about the “imminent threat” Iran posed and that Israel would strike first, prompting the U.S. to launch its own attack. Some lawmakers found this to be suspect.

I’m not going to predict how this will go. Maybe the Department of War has some sort of coherent plan (I suppose I’d have more confidence if Pete Hegseth weren’t leading it, but I know what I don’t know). The base case of “cooler heads will prevail” or “nothing really matters” is usually a reliable assumption in a $100 trillion global economy: markets have been incredibly resilient to COVID, the invasion of Ukraine, and other things. But usually isn’t always, and war in the Middle East is the very opposite of an easy-mode problem.

1. Though maybe I should just call him?

2. Although there was a lot of blame to go around for COVID, which is why Harris and Biden also didn’t want to talk about it much in 2024.

3. Or if they call, it’s with a remarkably strong hand that wasn’t even among your “bluff targets”. If you bluff and your opponent reluctantly calls with a full house because he’s afraid of a higher full house, this sometimes almost feels better than actually getting away with a bluff.

4. What’s “inconsequential”, I suppose, is debatable. Certainly something like the White House ballroom story. Epsteingate is a tricky case; I wouldn’t say it’s inconsequential so much as that we’re probably past the point of diminishing returns. I couldn’t disagree more with media critics who say that the media has focused too much on Iran and not enough on Epstein!

5. This was wrongly described as “buy” in the original post I was quoting and has been corrected.

6. Many poker players actually do this. For instance, they’ll bluff if the last digit on the clock is even but check and give up if it’s odd.

7. Although even that may have priced in some chance of future Trump actions in Iran.

8. Just as an observation over the years, often these freak-outs seem to be worse overnight in futures markets when there’s little actual news to trade on and markets are basically tripping out on the feedback.