In partnership with

My first real exposure to the debate about market efficiency was when I realized that both sides of the argument think they're the scrappy underdogs, but each side was talking past the other. I read the Roger Lowenstein biography of Warren Buffett at an impressionable age, in the summer of 2001. And the next summer I happened to read this early New Yorker profile of Nassim Taleb. In the Buffett book, academic finance was universal in its agreement that Buffett was merely lucky, not good, and only a handful of independent thinkers were able to point out that Buffett's approach made a) sense, and b) a statistically improbable sum of money. (Since the book was published, Berkshire has compounded at 12.0% annualized, compared to a 10.0% return for the S&P 500. And while Berkshire is more volatile than a diversified portfolio of 500 large-cap stocks, it's also a lot less volatile than the median S&P 500 member. It's a good track record!) The New Yorker piece informed me that, on the contrary, everyone treats Buffett's track record as evidence for investment skill, except a handful of skeptics, led by the daring Diogenes, Taleb, who asks: what if it was dumb luck and we just fit a narrative to the story after the fact?¹

Part of the reason it's contentious is that you can state the efficient market hypothesis in two ways, depending on what kind of fight you're interested in starting:

1. EMH is the claim that all information, whether publicly-known or secret, is instantaneously incorporated into asset prices at all times. Beating the market is impossible, though you can achieve a superior risk-adjusted return by buying a diversified portfolio. That's the only thing you should ever do.

2. The market is efficient in the sense that there is usually a good reason for the weird stuff that makes it look inefficient. It's easy to see a big one-day move, or a company outperforming over the course of a year, and point to this as evidence that the market is missing the obvious. It's a lot harder to compile a track record of spotting this in advance.

Point 1 is strong enough to be absurd, but point 2 is just the wishy-washy observation that investing, like any other competitive field, is challenging. And yet, there's a use for each.

In academic finance, the strong form of EMH is often a useful starting assumption because it allows you to reduce the degrees of freedom enough to get useful answers in other areas. For example, suppose you're talking about how to fairly value an option to buy Tesla in a year for $250/share. There are formulas that can get you to a reasonable answer based on the current price, strike price, volatility, and risk-free interest rate. But these formulas are all wrong if Tesla is really worth $50, or $5,000. So if you want to discuss how to value an option assuming you don't have a view on the stock, you might as well apply EMH. This also helps for discussing broader macroeconomic points about how capital markets develop, and how some countries have bank-oriented and others market-oriented financial systems—it needlessly complicates the discussion to also incorporate inefficient markets into this.

On point 2, there are different takeaways for the average saver and for professionals and market hobbyists. For the average saver, the (generally correct!) takeaway is that it's probably best to just buy a balanced portfolio and not think about it too often. Yes, sometimes stocks are overpriced or underpriced, and sometimes there are sectors that are artificially cheap or expensive. But the only way they get that way is if someone thinks the opposite is true—you can easily end up in a situation where you're persistently buying statistically cheap stocks in shrinking industries and systematically selling expensive-looking ones to people who recognize that, accounting for qualitative factors, those are actually the cheap stocks.

If you're going to be an active investor, you turn point #2 around and ask: why do I deserve to make money on this trade? That's not a moral question, but a practical one: you found a growth stock that seems cheap, you plan to buy it, and now you need to ask why other people aren't buying it, i.e. why it got so cheap. That's an especially useful exercise to apply to certain corners of financial advice: yes, there are people who can point to a track record of making 80 basis points a week writing options. Someone is buying those options—why is that someone paying enough to give their counterparty such a high return?² Sometimes, there's a good answer for this, like:

• This company operates in two different industries, and the one that accounts for most of the revenue historically is not the one that accounts for most of the growth. (There was a time when the leading cloud computing company was tracked by a large number of analysts who also published research on Walmart and Sears.)

• It's listed in a market that US investors don't pay much attention to.

• There's hair on the deal—maybe they had accounting problems in the past, or have recently switched management. If you're very lucky, there were professionals paying attention to it during the bad times, and they dropped it from their watchlist before the good times started.

• The company is selling a product that you're in a position to know is trendy, and average investors don't understand it. This can work, in both directions, for products that are popular with non-English-speaking populations, big in the Midwest, a hit with stay-at-home-moms, better for large families than for single people, etc. Lululemon, Peloton, and DoorDash all sell disproportionately to the "analyzes stocks for a living" demographic, or their significant others. Torrid, Planet Fitness, and Cracker Barrel don't. This doesn't mean that those stocks are perennially undervalued, of course: plenty of money was lost on Zulily and Wish, both of which sold to non-hedge-fund-analyst demographics. But it does mean that these companies are less likely to be on the radar.

• Some investors have liquidity constraints, which preclude them from investing in tiny companies with low volumes. But beware! There are plenty of professional money managers whose mandate at work confines them to mid- and large-cap stocks, but whose hobby is to manage their personal account and do the more interesting stuff there. Just because it does $20k/day in volume, you can't be certain that $5k/day of that isn't some hedge fund analyst selling shares to you from their personal account because the easy money has been made.

• Sometimes there are institutional limitations that keep people from exploiting opportunities. Zero-days-to-expiration options are messy to trade, and returns from selling them are lumpy, but an individual investor who cares about overall returns and not day-to-day volatility can sell them. On the other hand, that individual investor needs to know exactly what they're doing. Always assume that orders of magnitude more thought have been put into the restriction you're exploiting than into your own exploitation of it—to the extent that you try to take advantage of these situations, try to frame it as "here's why this is a good trade for me and not for a bigger investor" rather than "here's a dumb mistake I spotted and am taking advantage of." There might be such a dumb mistake in the end, but you don't want to find out the hard way that you're the one making it.

The strongest form of EMH is fundamentally a paradox. Markets won't stay efficient for long if there's no money in doing research and making trades. Weaker forms are a good starting point. The goal is to find the messy border between a useful descriptive theory and the facts on the ground, and then to convert that more academic observation into your own P&L. Or to conclude that you can free-ride on lots of smart people working hundred-hour weeks to eke out extra returns and just buy the index while you enjoy the rest of your life.

Read More in The Diff

The Diff has covered topics related to market efficiency in many different places. We’ve looked at:

• How some investors assume that one set of markets is efficient in order to exploit a different market ($).

• Rentec as the sum of many different edges, or one big one. 

• We’ve told the story of why IAC stock once dropped on the news that the company had $450m more cash on hand than investors had thought a few hours before ($).

• Here’s more on spotting and exploiting tiny inefficiencies, including another example of a strategy that reliably produces excess returns—but usually maxes out at a few thousand dollars a year.

And asking about efficiency means asking what we mean when we talk about investing and trading, anyway. If you lose a few thousand dollars a year trading stocks, but you also have a good time with your friends, maybe it’s just a convenient alternative to Vegas with a little less debauchery.

Join the discussion!

1  There is plenty of luck involved, as Buffett points out. There are few better ways to ensure a high net worth than to be born on the autism spectrum, with a special interest in stocks, at a time when the start of your career coincides with a stable political and economic situation coupled with a wildly mispriced market. Perhaps the only better recipe for wealth is to be on the spectrum, with a special interest in computing machines, and to start your professional career just when Moore's Law has made enough progress that a "personal computer" makes sense. I don't know that either of them have been formally diagnosed, but read a biography of either of them with this thought in mind and it's obvious that you're reading a story about extremely high-functioning autism. And that the late twentieth century is, in part, a story about the economy and society finding good ways to cope with and benefit from these traits.

2  In case you're curious, the answer is that an option seller is really selling insurance against big moves, especially big correlated moves that wipe out all of their positions at once. If there weren't capital requirements in insurance, you could make amazing returns offering cut-rate earthquake insurance. Until the earthquake hit.