A perfectly efficient market

Cannot be a meritocracy!

Imagine a marketplace so perfect, every scrap of knowledge, every rumor, every insider tip is instantly baked into asset prices. This is the dream of efficient-market theorists, and it goes by many names.

But let’s call it what it is: an environment where being “smarter” than everyone else fails to guarantee better results. In fact, in the theoretical extreme, no amount of raw talent or skill can systematically outperform the market’s collective intelligence. Outcomes start to look random. That’s what Grossman and Stiglitz (in their 1980 paper, “On the Impossibility of Informationally Efficient Markets”) pointed out decades ago. If markets were truly “perfect,” prices would reflect all available information, and no one would gain an edge. The moment you discover something juicy, it’s already in the price before you can click “buy.”

This sets up a curious puzzle: how can we talk about a market rewarding merit if any advantage vanishes faster than we can exploit it? If you’re the one doing the analysis, spending time and money to gather data, the instant you try to trade on that data, others see your action, and the price adjusts. The reward you get must be large enough to cover all the time and expense you sank into research, or else you’ll stop doing the research. But in a perfectly efficient market, that reward is supposedly zero. You’re left spinning your wheels.

That contradiction is at the heart of Grossman-Stiglitz: an absolutely efficient market destroys the profit motive behind analysis—thus, ironically, it erodes the mechanism that keeps it efficient in the first place. There has to be some “slack,” some friction, or people stop looking for edges, and the market stops reflecting new information accurately. Perfection, taken to an extreme, undermines itself.

You might ask, “So does that mean nobody is rewarded for skill?” The short answer: in the idealized scenario, yes. The presence of skill becomes nearly impossible to distinguish from luck. It’s a bit like an unyielding slot machine that’s perfectly rigged to pay out exactly its statistical average over time. Even if you uncover a clever pattern, the machine’s math has already accounted for it. Each outcome is still a coin toss in disguise.

Here’s a thought experiment. Picture the market as a quantum system. The very act of measuring it—say, you learn a new piece of data—alters its state. Trying to gain knowledge from a quantum particle changes how that particle behaves. Likewise, in a perfectly efficient market, your attempt to trade on information changes the market’s “state” immediately. There is no quiet moment to strike. It’s done before you even realize it.

Now, some folks find this depressing. Doesn’t it mean skill is meaningless and your hard work is in vain? Let’s not jump too quickly. Remember that in reality, markets are never truly perfect. They can be nearly efficient, but not all the way. Traders do exploit “glitches,” from mispriced emerging-market bonds to subtle patterns in earnings announcements. They can earn profits, though the highest profits often demand the highest risk. That doesn’t guarantee a wonderful existence for everyone, but it does mean there’s still a marketplace for information-gatherers. The system breathes because of those small imperfections.

However, it’s still instructive to think about what full efficiency implies. No matter your brilliance, you can’t beat the market if everything is instantly reflected in price. And if, hypothetically, you did better for a while, statisticians would argue that some small fraction of people always land a lucky streak. It’s the broken clock phenomenon: across many participants, someone’s bound to be “right” a surprising number of times in a row, but that doesn’t prove skill is the cause.

We see echoes of this in professional sports drafts. Imagine a perfectly efficient draft in which every athlete’s hidden potential is immediately spotted, priced, and distributed across the teams. Every manager pays precisely the correct cost in trades or draft picks for that future superstar. No team manages to “steal” a gem in the late rounds. There’s no way to “beat” the system, because the system sees everything in real time. No manager is lauded for clever scouting. What remains? Pure, random assignment. In practice, that never happens. Human biases, incomplete data, and weird intangible factors create inefficiencies. But in a frictionless universe? A draft might look random.

Similarly, in neuroscience, we sometimes talk about the “predictive brain”: it’s constantly updating expectations in real time, so your perception of the world is always a hair’s breadth behind. A perfect brain that instantly integrated all signals would leave little room for learning, because it would already know everything the moment those signals arrived. Real brains aren’t perfect. Their biases drive them to update in slow, partial ways, which allows them to discover new strategies (and also make mistakes).

All of this suggests that if markets were truly perfect, there’d be no discernible measure of “merit.” It’s random outcomes all around. Paradoxically, the quest to make markets as efficient as possible might annihilate the very merit-based performance that many analysts and fund managers prize. Ironically, we need small pockets of inefficiency to reward insight. That’s what keeps people hunting for better ways to interpret data, pick investments, or scout that overlooked basketball phenom.

Now, let’s think about the advent of more and more AI, in every kind of market: from asset markets, to spot commodities markets, to futures markets, to car markets, to candy markets! The more AI there is, the more “opportunities” are going to get “exploited.” And the more opportunities that get exploited, the more efficient the market is, and the less opportunity there is for others to exploit. That means one of the likely outcomes of AI is that market results, in every kind of market, become more random. “Merit” (however you define it) might be less rewarded in the age of AI!

In everyday life, we appreciate the opportunity to be recognized for our talents. We want to know that if we work hard and sharpen our skill set, some sort of advantage awaits. In a theoretical universe of absolute market efficiency, that advantage dissolves. There’s a profound lesson about incentives embedded in that. The engine of discovery runs on the possibility of gain. Remove the spark, and the engine sputters.

The next time someone claims the markets are “too smart” or “too perfect,” remember Grossman-Stiglitz. Remember that perfect efficiency is its own downfall. And think about how that principle might apply to your own pursuit of knowledge or personal goals. Perfection looks appealing. But it can rob us of the fun in being clever. Sometimes, a little noise in the system is what keeps us alive and engaged.

So where does that leave us, especially in an era when AI accelerates this trend? Perhaps it leaves us grateful for small imperfections—tiny gaps where genuine skill can still shine. Keep that in mind as the age of AI dawns upon us!

---

Check out my book, “The Science of Free Will: How Determinism Affects Everything from the Future of AI to Traffic to God to Bees,” which has all kinds of cool stuff: https://amzn.to/4aMQJD1.