Efficient Market Hypothesis (EMH) [Basic Investing I Studied #12]

Alright, let’s talk about the Efficient Market Hypothesis (EMH)!!!

Back in the 1950s, computers showed up and economists started actually using them for research. And one of those guys was Maurice Kendall, who basically grabbed a computer and went:

“I am going to find SOME kind of Rule in this stock market, no matter what!!!”

So he dove in. Tried his absolute hardest to dig up some regularity, some pattern, anything.

He failed. Hard. And his conclusion was basically: “Aw damn — you can’t predict a thing in here. I think this stuff just does a Random Walk.”

…which was, uh, a kinda dangerous thing to say.

Because if stock prices do a random walk, that means the people in the market aren’t acting “rationally” — and that little observation could be read as evidence that humans are, you know, kind of irrational creatures. Hella inconvenient if your entire field is built on the opposite assumption.

So picture every existing economist at the time at full red alert:

“Ah sh*t lol my whole research has a busted premise sitting at the bottom of it… no no no no, that can’t be right.”

And so, other economists swooped in and reinterpreted the random walk thing in a completely new way — one that let them keep the “humans act rationally” assumption alive. How’d they pull it off?

“It’s not that people are acting irrationally. It’s that they’re acting too rationally — and that’s why the market does a random walk.”

Wait, what? Let’s unpack:

“Hey hey hey, listen up — there’s a TON of information out there in the world that’s relevant to the stock market, right?”

“Now assume all of that scattered information is something anyone can get to.”

(So whatever I know, you also know, and that other guy knows too. Not a crazy assumption.)

“Since everyone has access to all the info, and humans are uniformly rational, they’re all gonna use that info to make whatever choice benefits them — agreed?”

“Which means: all the information out there has already done its full job on the market. That stock price chart you’re staring at? It’s the perfect distillation of human rationality. Every bit of it baked in. Cool?”

“I’m calling this the Efficient Market Hypothesis (EMH).”

“And from this angle, yes — stock prices doing a random walk is exactly correct.”

“Because — when does the price you’re staring at right now actually move?”

“It moves the moment new information shows up. And here’s the key: ’new’ information, by definition, can’t be foreseen by anyone. The second you can foresee it, it stops being new.”

“So stock prices move when info nobody knew, nobody could find, nobody could predict suddenly drops — which means, of course, to our eyes, the prices look like they’re doing a random walk!!! Make sense?”

…apparently that’s the move, lol.

And from there, you can extend it: depending on the speed of information flow, price behavior is going to differ. Which means market efficiency varies country to country.

Places like the US and South Korea where information moves fast? Yeah, plausible that all the world’s info is sitting in the prices. Emerging markets where the internet and other info channels still aren’t fully built out? More room for prices to be set inefficiently compared to the developed world.

Same logic: a country that’s a little less transparent in terms of politics or corporate governance has a higher chance of inefficient pricing than a fully transparent one. And small-cap stocks, which get way less analyst attention than the big names, are also more likely to be priced inefficiently.

But hey — even cold porridge has layers (so to speak). EMH itself comes in layers.

We classify EMH by what counts as “all available information.”

Weak-form EMH:

Past price data, short-selling balances, etc. — basically all the “available information” you can scrape out of historical trading data is already baked into the price.

Semistrong-form EMH:

Everything from above, plus a company’s product pipeline, management quality, balance sheet info, patents, etc. — i.e., all publicly available information about a company’s prospects is already in the price.

Strong-form EMH:

Everything from the two above, plus every bit of insider-only info is also already in the price.

OK so now that you’ve heard all that — how do you feel? Do you actually buy EMH? I do not, lol. Like, at all. And I’m definitely not alone — plenty of well-known economists are in the “nope” camp too.

Anyway, let me walk through some stuff that pushes back on EMH.

Technical analysis:

This is the school of “let’s hunt for repeating patterns in stock prices.”

And it’s kinda funny — because the people doing technical analysis are commonly called “chartists,” and the thing they use to find these so-called repeating patterns is (surprise) past stock price records. Which means they are, basically, EMH deniers in their soul.

The classic example of technical-analysis tools is resistance and support levels. The idea: there’s an upper and lower band that prices tend to bounce between. And the reason these levels are even meaningful is that they’re set by people’s psychology — so it’s not totally nothing.

There’s a step beyond that, though.

Fundamental analysis:

This one uses past records plus corporate earnings, dividends, future prospects, interest rate expectations, risk — basically all available info — to compute the “present discounted value” of a single stock.

If the present discounted value > current price → tell people to buy. If < current price → tell people to sell. Something like that.

Of course EMH would totally reject this. But — if some analyst happened to have access to special information that no other analyst could touch, then their analysis would actually work without breaking EMH, right?

(Of course, the very existence of such information would itself be a violation of EMH, so… yeah.)

Anyway, those are some of the things standing in opposition to EMH. And honestly, the very existence of these analysis methods feels like it’s quietly arguing that EMH can’t be 100% right.

(Saying it again though — true believers in EMH will tell you all that analysis is completely meaningless.)

So: someone who believes in EMH ends up doing passive portfolio management — no attempts at outperforming. And someone who doesn’t believe in EMH ends up doing active portfolio management, where you try to beat the market via analysis. Right?!

Buuut — being a passive manager doesn’t mean you do zero analysis. You still gotta:

Properly analyze stock selection for diversification (the thing from the previous chapter). Mix stocks and bonds smartly to minimize risk.
Think about taxes. Pick tax-advantaged stocks, or just eat the tax to chase return — that’s still a real analysis problem.

So even passive managers analyze. They just don’t analyze the market.

Terms, terms

Serial correlation: the tendency for stock returns to be related to past patterns.

↓

Momentum effect: stocks that did well (or poorly) in one period tend to keep doing well (or poorly) in the next.

Reversal effect: stocks that did well (or poorly) in one period tend to flip in the next.

The momentum effect is weak in the short term, and apparently the more reliable and up-to-date the price data, the smaller the momentum effect gets.

But over the medium term (3–12 months), there was a 1993 study by Jegadeesh & Titman saying it’s a stretch to claim the effect is gone.

That research gave birth to the fads hypothesis!!!

The fads hypothesis says: in the medium term, stock prices may overreact to information. But in the long run, you can expect the fad (the overreaction) to get corrected — so over the long run you’d expect a reversal effect kicking in to undo the overreaction.

Which is why there’s this thing called the contrarian investment strategy: avoid the recent winners, buy the recent losers.

(Apparently the returns reverse this way, that triggers another overreaction, then it reverses again further out. The cycle just keeps going.)

Small firm effect:

Small-cap stocks tend to post abnormal returns specifically in January. Follow-up research after this was announced apparently nailed it down further — most of the small-cap effect happens in the first two weeks of January.

Book-to-market effect:

In 1992, Fama & French reportedly said the ratio of book value to market value of equity is a powerful predictor of returns. The group with a high book-to-market ratio had higher average annual returns than the group with a low ratio.

This one was apparently a terrifying result — the kind of theory that could blow up basically all of existing market theory. Why? That’s a whole conversation for the next chapter.

Anyway — whether all these effects are showing up because the market is genuinely inefficient, or whether they’re just a perfectly legit financial risk premium, is still open research.

And whether these effects are real anomalies that show up anywhere, or just needles-in-a-haystack you only ever find by data mining — that one’s still open research too.

Originally written in Korean on my Naver blog (2016-05). Translated to English for gdpark.blog.