Structural Market Inefficiencies
Twelve market inefficiencies to improve your investment process: value, momentum, profitability, low beta, small caps, illiquidity, and other useful signals for investing better.
There is no free lunch.
Or is there?
Investors have been told many times that the market is more or less efficient. That there may be occasional mistakes, sure, but they usually get corrected quickly, and exploiting them systematically is very hard. In the abstract, the idea sounds reasonable. If everyone is looking for free money, the normal thing is that the free money disappears.
And yet, the market does not work exactly like that.
Some inefficiencies do not disappear entirely because they do not depend only on someone discovering them. They depend on human biases, institutional constraints, misaligned incentives, liquidity problems, implementation costs, and something much simpler: many strategies are psychologically devastating even when they work.
In other words, these are not inefficiencies that will vanish overnight. They are structural inefficiencies.
You already know many of them. Value, momentum, small caps, illiquidity, quality, buybacks. I am not discovering America here. But maybe you have not stopped to think about why they still exist, how they combine with each other, and above all, what they can add to a real investment process.
But careful: contrary to what it may look like, these factors are not a shopping list.
Just because a company scores well on a factor does not mean it will go up. Just because it is cheap does not mean it is a bargain. Just because it has momentum does not mean it is good. Just because it is high quality does not mean any price is justified. We are talking about patterns that work in aggregate, not an algorithm for picking individual companies.
That said, understanding them helps a lot. It lets you know what kind of tailwind you have, what risks you are taking without realizing it, and what traps are worth avoiding. So here are 12 structural inefficiencies that, used well, can meaningfully improve any investment process.
Value
The best-known one.
The academic definition of value does not always match the one used by old-school value investors. Academics talk about buying companies that are optically cheap, meaning cheap relative to earnings, book value, sales, FCF, EBITDA, assets, or any other reasonable valuation metric. In other words, companies trading at low multiples relative to their fundamentals.
It can look like a fairly crude way to value businesses. And very often it is. A single multiple tells you almost nothing about a specific company. But in aggregate, historically, companies that are cheap on multiples have tended to outperform expensive companies. We usually call the first group value; the second, glamour.
And here is the good part. The interesting question is: if everyone knows that buying cheap makes sense, why does it still work?
Because people love convincing stories and hate companies that stink.
The market tends to overpay for companies with a promising narrative, visible growth, and an earnings presentation that says all the right things. At the same time, it overpunishes boring, cyclical, dirty, disliked companies that are simply uncomfortable to own.
Fama and French formalized this effect using book-to-market, because it was a fairly stable metric over time. Then the world changed, less tangible-asset-heavy business models appeared, like software, and that metric lost part of its power. That is why many other valuation measures are used today. You may also have heard of HML, high minus low, which is basically a factor built as a strategy that buys value or cheap stocks and sells expensive stocks; its return measures the difference between the two.
Now, careful.
The fact that the value factor works does not mean buying anything at a P/E of 2 is a good idea. Many companies are cheap for a perfectly reasonable reason: because they are bad, because the business is deteriorating, because the accounting is misleading, or because capital will be destroyed before you can get paid.
That is the big problem with value: value traps. Optically cheap companies that are not actually cheap, but rather garbage businesses with justified low multiples.
Also, value can have brutally long periods of pain. You can spend years looking like an idiot while the market pays absurd multiples for growth companies. You can bleed slowly against the index and, right when everyone decides value is dead, get a brutal rotation that recovers years of underperformance in a few months.
It works, yes, but it is not glamorous. You look like an idiot most of the time, and it is a bet that is hard to hold for years.
And that, precisely, helps explain why it still exists.
Value: cumulative factor return
Long cheap stocks and short expensive stocks, according to the value cluster in the U.S. sample of the Global Factor Data.
Price momentum
Price momentum is one of the most hated factors among many value investors and one of the most loved by technical investors. The idea is to buy the stocks that have done best over the last 6-12 months, usually skipping the most recent month to avoid short-term noise and reversal.
Fundamental investors often look down on it because it is based only on price.
And of course, if you understand the stock market from a business-owner point of view, this does not quite fit your mental model. How can price tell you anything useful while ignoring earnings, margins, debt, returns on capital, or anything that looks remotely fundamental?
Well, it does.
The market takes time to incorporate information. Analysts adjust estimates gradually. Flows chase winners. Narratives build in layers. And prices, whether we like it or not, sometimes start reflecting changes in expectations before those changes become obvious in the numbers.
Also, investors are not robots. We struggle to sell winners, we chase what is going up, we extrapolate trends, and we suffer from FOMO with embarrassing ease. Momentum does not work because the market is "dumb" in some simplistic way; it works because price discovery is slow, social, and emotional.
The problem is obvious: in euphoric phases, momentum can push you into speculative garbage. Companies with no fundamentals, no cash, no profits, and a wonderful story that only needs 17 more funding rounds to become real. And when the market turns, the famous momentum crashes arrive: violent drops that can evaporate a large part of the accumulated return.
It is a very powerful factor, but it is not free.
Price momentum 12-1: cumulative factor return
Long winners from the last 12 months and short losers, skipping the most recent month to reduce very short-term reversal.
Value + momentum
Value and momentum get along better than it seems.
In fact, they tend to be negatively correlated. Value buys what is hated; momentum buys what is rewarded. When the market is in love with growth and value suffers, momentum often helps. When momentum crashes because losers rebound, value often has more exposure to those cheap or cyclical losers that bounce hard.
In practical terms: momentum can help you avoid value traps, and value can cushion part of momentum crashes.
That is why, even if Warren Buffett himself attended your christening, you should understand momentum. You do not have to become a trader or start drawing lines on a chart. It is enough to recognize that price contains information, and that buying cheap while everything keeps getting worse is usually not optimal.
Fundamental momentum
This is the fundamental cousin of price momentum.
Here you are not buying because the stock has gone up, but because expectations about the business are improving. Upward estimate revisions, positive earnings surprises, better guidance, margin expansion, revenue acceleration, debt coming down faster than expected. Things that indicate the business is doing better than the market believed.
The inefficiency exists because the market rarely adjusts everything at once.
One analyst raises estimates. Then another. Then the company beats again. Then the market starts believing the story. Then a new narrative appears. And meanwhile, the price incorporates that improvement in phases.
The classic case is PEAD, post-earnings announcement drift. After a positive earnings surprise, the stock tends to keep rising for a while. After a negative surprise, it tends to keep falling. In theory, the market should adjust the price almost instantly on earnings day. In practice, it often does not fully do so.
This factor is especially useful for value investors, because it helps distinguish between cheap companies that are starting to improve and cheap companies that are still digging downward. If the fundamentals keep deteriorating, maybe you are not buying a bargain; maybe you are buying trash.
Fundamental momentum: earnings surprise
Standardized earnings surprise: long companies with positive earnings surprises and short companies with negative earnings surprises.
Profitability
Business profitability is one of those ideas that seems too obvious to be an inefficiency. It is also one of the main metrics behind "Quality" companies.
Everyone wants companies with good margins, high returns on capital, recurring profits, and the ability to turn sales into cash. A company that earns a lot on the capital it needs to operate has more options: it can reinvest, buy back shares, pay dividends, withstand bad cycles, and grow without depending as much on external financing.
The question is the same as always: if everyone knows profitable companies are better, should that not already be reflected in the price?
Yes. But often not enough.
The market struggles to value the persistence of profitability. Many companies look expensive if you only look at P/E or EV/EBITDA, but if they maintain high returns for many years, reinvest at good rates, and do not need much incremental capital, they can compound value in a way static multiples do not capture well.
The inefficiency is not "buy any good company at any price." That would be an elegant way to overpay. The idea is that, in aggregate, the market tends to underprice how much it matters that a company earns a lot on its capital and that this profitability is more stable than it looks.
As a filter, profitability combines very well with value. It helps you avoid filling your portfolio with companies that are cheap for horrible reasons. It also combines well with momentum, because it reduces the risk of ending up buying purely speculative companies just because they are going up.
Lovely when factors help each other.
Gross profitability: cumulative factor return
Gross profits-to-assets: long companies with higher gross profitability on assets and short companies with lower profitability.
Conservative investment
This inefficiency is fairly counterintuitive.
You would think that the companies investing the most, expanding assets the most, and chasing growth the hardest should do better. After all, they are investing to grow. Was that not what we wanted?
Not necessarily.
In aggregate, avoiding companies that aggressively expand their asset base, sharply increase capex, issue equity, or chase growth at any price tends to work better. The reason is very mundane: many companies invest badly. They buy expensive assets, build capacity they do not need, enter projects with mediocre returns, or make acquisitions so the CEO can say they run a bigger company.
The market, meanwhile, tends to applaud those expansion signals. More growth, more sales, more geographic presence, more verticals, more acquisitions. It all sounds wonderful until you look at the return on capital.
Reinvestment is necessary to build great businesses, yes. But it only creates value when it is done at attractive returns. This factor helps filter empire builders: companies that grow for the sake of growing without translating that growth into better returns for shareholders.
There is an important nuance. In the post about 10 baggers, we saw that many big winners did invest aggressively and expand their asset base. But they did it with one key difference: that growth was backed by real EBITDA growth. It was not empty expansion, but healthy growth.
The line between quality reinvestment and capital destruction can be thin, but it exists. That can be a topic for another post.
Asset growth: cumulative factor return
Long companies with more conservative asset growth and short companies that expand their asset base more aggressively.
Accruals / earnings quality
This one is less popular among retail investors, but I find it very useful.
The idea is that it is always better to trust profits that come with cash. If a company earns a lot on paper but converts little into cash flow, it is worth looking twice. Sometimes the difference is normal. Other times it is a sign that those profits are less solid than they look. In the end, what pays debt, dividends, buybacks, and reinvestment is not adjusted EPS; it is cash.
Nobody should be surprised that it pays to distrust companies that report beautiful profits but mediocre cash flows. Nor should anyone be surprised that it pays to distrust companies that live among adjustments, extraordinary items, invented metrics, and accounting reconciliations designed so nobody reads them.
The problem, classically documented by Sloan in 1996, is that the market focuses too much on accounting EPS and does not always distinguish between high-quality earnings and low-quality earnings. Sometimes it punishes a company for an ugly quarter even though cash generation remains intact. Other times it rewards an accounting improvement with no economic backing.
This factor is mainly useful for filtering out the worst. Companies with high accruals, weak cash conversion, and too much accounting creativity. It is not for finding jewels, but for avoiding problems.
And that is not nothing.
Operating accruals: cumulative factor return
Long companies with lower operating accruals and short companies with a larger accrual gap between earnings and cash flow.
Shareholder yield
Value investors tend to like this factor a lot, and for good reason.
The general idea is to buy companies that return capital to shareholders intelligently and avoid companies that systematically dilute them. Broadly speaking, this can include buybacks, dividends, debt reduction, and share issuance. I have already written quite a bit about buybacks in the introduction to buybacks, the post on the math of buybacks, and the empirical backtest, so there is no need to rehash everything.
But the chart below measures something more specific: net payout yield. That means companies that return capital through dividends and net buybacks, versus companies that issue more equity. It does not measure debt reduction. So it should be read mainly as a signal of capital return through equity and anti-dilution.
The base rate is clear: issuance is usually a bad signal.
Maybe the company thinks its shares are expensive. Maybe it needs to fund losses. Maybe it is paying for acquisitions with stock. Maybe it has no other reasonable way to raise capital. It is not always bad, but as a starting point, it is ugly.
Buybacks, by contrast, can be very good. But only if they are real and done at a good price. Buying back shares at absurd multiples while funding it with debt can destroy value. Buying back shares only to offset stock-based compensation is not exactly returning capital to shareholders either; it is more like running on a treadmill to stay in the same place.
The interesting thing about shareholder yield is that it looks very good in aggregate, even if case by case it is less obvious. Sometimes issuing shares is necessary or even positive. Sometimes a buyback looks reasonable and ends up being bad. And of course, when a company dilutes its shareholders, it usually comes with a wonderful narrative about growth, strategic opportunities, and long-term value creation.
There is always a nice narrative to justify every screw-up.
Net payout yield: cumulative factor return
Net payout yield: long companies that return capital through dividends and net buybacks after share issuance; short companies that issue more equity.
Illiquidity effect
Illiquidity is one of the best-known inefficiencies and one of the few real advantages individual investors have.
Less liquid stocks, less covered by analysts, less present in institutional portfolios, and harder to buy and hold tend to offer better opportunities. The market pays you for bearing friction: wider spreads, lower coverage, less processed information, more difficulty getting in and out, and less capacity to deploy large amounts of capital.
For a huge fund, many of these companies are basically invisible. It cannot buy enough without moving the price, cannot justify the research time, or cannot assume the liquidity risk. For a patient individual investor, by contrast, that same discomfort is an advantage.
This factor fits especially well with long time horizons. If you do not need immediate liquidity, do not use leverage, do not depend on stops, and do not hold positions that are too large, you can afford to look where others do not.
In other words, it is perfect hunting ground for the retail investor.
And, as almost always, it gets more interesting when combined with other factors: illiquidity plus quality, illiquidity plus value, illiquidity plus insider ownership, illiquidity plus a special situation nobody is looking at.
That is where the little gems tend to show up.
Anti-lottery / betting against beta
This is an often ignored gem and one of the pieces behind Buffett's success, as we discussed in the post on whether Buffett is really a good investor.
Boring, defensive, less volatile, lower-beta stocks have historically offered better risk-adjusted returns than their more exciting counterparts. This is where BAB, betting against beta, comes in: buying low-beta stocks and selling high-beta stocks.
The explanation makes a lot of sense.
First, investing in boring companies is not sexy. Everyone wants to find the next Tesla, the next NVIDIA before it was NVIDIA, or that stock that can multiply by 100. Lottery-like stocks attract capital because they promise an asymmetric story, even though the price often already embeds too much fantasy.
Second, many investors have leverage constraints. If you cannot use leverage and you want to increase the expected return of your portfolio, a simple way to try is to buy high-beta stocks. That pushes those stocks to prices that are too high and leaves low-beta stocks relatively ignored.
Third, benchmarks distort incentives. A professional manager does not live only by maximizing risk-adjusted return; they live by being compared against an index. A low-vol portfolio may have a better Sharpe, but if it has less beta than the market, it may lag in bull years. And lagging the benchmark is a very reliable way to lose your job.
That is why this structural inefficiency appears: knowing about it is not enough for it to disappear. Many market participants cannot or do not want to exploit it well.
Now, careful.
Low vol is not always cheap. Sometimes these stocks become bond proxies: utilities, staples, REITs, infrastructure, dividend stocks. When everyone wants safety or yield, safety also becomes expensive.
Even boring can get expensive.
Reversal
Here it is worth separating horizons.
There is short-term reversal: stocks that fall a lot over days or weeks can rebound because of microstructure, liquidity, rebalancing, or overreaction. The problem is that it is usually hard to exploit because trading costs, spreads, and taxes eat a large part of the opportunity.
And there is long-term reversal: stocks that have been losers for 3-5 years can rebound if the market overestimated bad news. This effect fits quite well with the value intuition: the market overpunishes something, the narrative becomes too negative, and when reality stops getting worse, the price adjusts.
Reversal complements and fights momentum depending on the horizon. Over one month there can be reversal. Over 6-12 months there is usually momentum. Over 3-5 years, reversal can appear again.
That is why classic momentum usually skips the most recent month. And that is why value works better when you are not buying a recent fall simply because it looks cheap, but when there is some signal of stabilization.
And this is where much of the debate about catching falling knives comes from.
Short-term reversal 1M: cumulative factor return
Long last-month losers and short last-month winners, a signal with a lot of practical friction but a clear statistical history.
Industry momentum
As you can see, momentum keeps showing up in different disguises.
Industry momentum buys winning industries because winning industries tend to keep winning for a while. It is the more macro part of individual-stock momentum and, in fact, can explain a meaningful part of the momentum in many stocks.
Sometimes you think you have found a company with a lot of momentum, but in reality you only bought the right sector. That can be good or bad depending on what you are trying to do.
There are two reasonable ways to use it.
The first is to do sector-neutral momentum: you look for the best names within each sector to reduce macro or sector bets.
The second is to explicitly accept that winning sectors keep winning for a while and incorporate that as a signal. This exposes you more to the cycle, sector narratives, and rotations, but it can also capture very powerful moves.
If you do stock picking, what you need to do is separate "the company is good" from "the whole sector is going up."
They are not the same thesis.
Small caps / size factor
We leave the most famous factor for last.
The intuitive explanation for the small-cap effect is simple: small companies live in a part of the market with less attention, less capital arbitraging mistakes, more friction, more real risk, and more chaff mixed in with the wheat.
It is the kind of place where, in theory, a patient investor with a strong stomach can find opportunities that do not exist in mega caps covered by 40 analysts.
But the size factor is also one of the most debated and least clean.
There are several important nuances:
- The effect is not linear and is concentrated mostly in the smallest companies.
- The classic size effect weakened quite a bit after being discovered.
- Part of the historical premium seems concentrated in January. In some analyses, the size premium outside January was practically zero.
- Market equity does measure size directly, but the size premium is not a pure inefficiency like value, profitability, or momentum. It is often mixed with illiquidity, lower coverage, higher business risk, value, beta, sector composition, and a lot of other things.
In a sense, it is not a free inefficiency. It is compensation for putting up with real crap.
Fama and French incorporated it as SMB, small minus big, but the broad small-cap signal without filters is dirty. The small-cap universe is full of bad companies: companies with no profits, high leverage, constant dilution, binary biotechs, mediocre roll-ups, weak cyclical businesses, expensive growth with no FCF, and projects that only exist because someone is still willing to finance them.
The good news is that size matters if you control your junk.
When you filter by quality, profitability, balance sheet, dilution, and other junk signals, the small-cap premium starts to look more interesting. Again, not because all small companies will do better, but because good, underfollowed small companies can sit in an area of the market where important mistakes still exist.
So, does the small-cap effect still exist?
Yes, but not as a pure, clean, automatic factor. The broad version of "small minus big," without filters, is weak, unstable, expensive to implement, and probably contaminated by liquidity, quality, January, beta, and sector composition.
Is it simply a proxy for liquidity?
Partly yes, but not entirely. Liquidity explains much of the raw effect. But when you control for quality or junk, something unique seems to remain: a premium associated with small companies that are better than the market is willing to look at.
That is why small caps are also interesting territory for the individual investor.
Size: market equity
Market equity: long smaller stocks and short larger stocks in the U.S. sample of the Global Factor Data.
How to use all this
The tempting thing after reading a list like this is to open a screener, add 14 filters, and think you now have a return-printing machine.
Spoiler: you do not.
The right way to use factors is not as a recipe, but as a map of probabilities. They tell you what type of company tends to have a statistical tailwind and what type of company tends to have a statistical headwind. Then comes the hard part: understanding the specific case you are studying.
For a retail investor, this is especially useful because the scarcest resource is usually not capital, but attention. You cannot analyze 5,000 companies in depth. You need to decide where it is worth spending hours and where it is better to move on quickly. That is where factors help a lot.
The first practical use is cleaning up the universe. If a company looks cheap but dilutes a lot, converts earnings poorly into cash, invests aggressively at mediocre returns, and also has terrible momentum, the burden of proof is extremely high. It can work, of course. But you already know you are fighting several forces at once.
The reverse also works. A small, undercovered, reasonably cheap, profitable company with a good balance sheet, no dilution, and signs of improvement deserves more attention than a beautiful story with only a good narrative. It does not mean buying it. It means it probably deserves a bit of your time to study it.
The second use is combining signals. One factor is a clue. Several factors pointing in the same direction are more interesting. Value works better when you do not buy garbage. Momentum works better when you do not pay outrageous prices. Small caps are more attractive when you filter for quality. Shareholder yield is more powerful when the company buys back cheap and is not just dressing up dilution.
Some interesting combinations:
- Value + quality: cheap companies, but with decent profitability, cash, and balance sheets.
- Value + stabilization: punished companies where the price stops getting worse and fundamentals are no longer deteriorating.
- Small caps + illiquidity + quality: underfollowed businesses where the individual investor can actually look.
- Shareholder yield + reasonable valuation: buybacks or dividends that genuinely create value.
- Momentum + fundamentals: companies where price confirms a real improvement, not just a fad.
You can imagine there are a thousand attractive combinations.
The third use is understanding your portfolio. Often you think you have a diversified portfolio because you own 15 different stocks, but in reality you have a single bet: all expensive, all growth, all high beta, all from the same cycle, or all dependent on high multiples. Factors help you name those hidden exposures. This is where many investors are naive, thinking that diversifying across sectors and countries means they are safe.
They also help you stop cheating at solitaire. If you say you are value but all your positions depend on perfect growth for 10 years, maybe you are not that value. If you say you are quality but your companies do not convert earnings into cash, maybe you are buying accounting quality, not economic quality. If you say you are contrarian but you only buy stocks that keep going up because of momentum, maybe you are on the consensus side.
And I could keep going... in the end, knowledge has a thousand branches and applications.
To finish, I want to make one thing clear: this should not be misunderstood. No factor works all the time. No good company scores perfectly on everything. And no backtest replaces understanding the business. This is just one more tool to add to your investor toolkit, not the Ten Commandments.
Appendix: all factors on the same scale
As a final extra, here is a small summary of all the factors mentioned in the post so you can compare them easily. Remember that I omitted the charts for some factors, so they will not appear in the summary.
All factors in the post, on the same scale
Comparison from the first common year available. All series use JKP. Each series starts at 0% so the chart shows relative cumulative growth, not absolute levels inherited from different samples.
| Factor | Total return | CAGR | Volatility | Sharpe | Max drawdown |
|---|---|---|---|---|---|
Mom 12-1Momentum | +2,117% | +5.9% | 15.0% | 0.46 | -49.0% |
Op. accrualsQuality | +1,076% | +4.7% | 6.5% | 0.74 | -35.7% |
PayoutShareholder yield | +897% | +4.3% | 14.1% | 0.37 | -60.3% |
ValueValue | +643% | +3.8% | 11.7% | 0.38 | -62.0% |
Asset growthInvestment | +478% | +3.3% | 10.0% | 0.37 | -34.3% |
Rev. 1MReversal | +413% | +3.1% | 12.2% | 0.31 | -40.1% |
GP/AProfitability | +342% | +2.8% | 8.2% | 0.38 | -33.6% |
SUEFund. momentum | +284% | +2.5% | 5.1% | 0.51 | -14.5% |
SizeSize | -10.0% | -0.2% | 10.2% | 0.03 | -62.9% |
Enjoyed this post?
Leave me your email and I'll let you know when I publish something new.
Rate this post
If you liked it, give it a rating. It helps me improve.
1 = weak, 5 = excellent
Keep reading
More posts you might enjoy.
Is Buffett Really a Good Investor?
Nobel Prizes, statistical anomalies, orangutans flipping coins, leverage, and the paper that tries to explain the Berkshire miracle.
13 min read- fundamentals
- quantitative
- opinion
You Don't Understand Share Buybacks
Why cancelling 17% of the shares does not mean creating 17% of value, and how to separate EPS accretion from real value creation.
21 min read- fundamentals
- quantitative
10x Stocks: The DNA of Multibaggers
Analysis of 10x stocks and multibaggers: size, valuation, profitability, growth, investment, free cash flow and the signals that best explain their returns.
45 min read- fundamentals
- quantitative