10x Stocks: The DNA of Multibaggers

When I started investing, one of the books that fascinated me most was 100 Baggers: Stocks That Return 100-to-1 and How to Find Them, by Chris Mayer. The book reads well, it pulls you in, and it touches a nerve that every investor has, even if they do not always admit it: the fantasy of finding a company capable of multiplying your money by 10, by 50, or by 100 while you, in theory, do not have to do too much. But when I finished it, I was left with a slightly incomplete feeling. The idea was brilliant, the examples were suggestive, but I was missing the most important part: data, evidence, charts and a more serious empirical validation. Fewer anecdotes about extraordinary companies and more evidence about what they really had in common before they became stock market monsters.

Because everyone is attracted to multibaggers. That is normal. They are the philosopher’s stone of investing. The problem is that, despite how juicy the topic is, there are not that many studies that have tried to analyze them rigorously. And the few that do exist often fall short: small samples, conclusions that are too narrative, little statistical testing or insufficient methodology to separate what seems true from what simply looks beautiful in hindsight. Because, of course, saying “you should have bought Amazon, Apple or Monster Beverage” after they have already multiplied is very easy. The hard part is studying whether those companies, before exploding, already had detectable traits.

For everyone who has ever felt that way, today I bring gold.

I recently read a paper that tries to do precisely that: study the characteristics of multibaggers and understand which factors explain their extraordinary returns. To do this, it analyzes 464 U.S.-listed stocks that increased in value by at least 10 times between 2009 and 2024. What matters is not only the results, but also the methodology, because it works as a good practical introduction for anyone who has never heard of “factors” in their life. The author begins by explaining the returns of these companies with a model based on the classic factors and then modifies and expands it to adapt it better to a very particular universe: companies that did not just beat the market, but destroyed it.

“when the proposed characteristics are subjected to statistical testing, they frequently fail to hold”

That is where the story changes. The study does not simply confirm obvious ideas, but instead finds characteristics specific to multibaggers that go against what is usually observed in the broader market. That does not mean we now have a magic recipe for finding the next stock that will multiply by 100, unfortunately, but it does mean there are useful patterns. It also questions some fairly widespread myths, such as the obsession many investors have with watching earnings per share under a microscope in this type of company. Maybe, in businesses with this profile, looking only at that is a way of missing what really matters.

So let’s try to dissect their DNA: what these companies had in common, which factors seem to explain part of their returns, which conclusions are useful and which parts of the study should be viewed with some skepticism.

Appendices

If you are only passing through, Appendix III will be the most direct one for you. And if you plan to incorporate any of these strategies into your investing style, Appendix II is mandatory.

Appendix I

Past studies

Go to appendix ->

Appendix II

Limitations

Go to appendix ->

Appendix III

Descriptive statistics of the sample

Go to appendix ->

---

The anatomy of a classic multibagger

So far, the literature seems to broadly agree that a multibagger is usually born from a combination of several ingredients: small size, quality, growth, duration and a reasonable valuation.

Put like that, it sounds simple. But of course, then you open each drawer and the nuances begin. Growth has to be sustainable, quality has to endure over time, size matters but not every small company is worth it, valuation cannot be absurd, management has to know how to reinvest capital, the market must not have already priced everything in...

In other words, finding a classic multibagger is not about ticking five boxes and waiting for the money to appear. Each ingredient has its own sub-ingredients, conditions and traps. I suppose it could not be that easy.

Before getting into the new findings of this study, it is worth reviewing how these companies have traditionally been searched for. Which characteristics appeared repeatedly in previous studies, which signals seemed to matter and what type of company had, at least on paper, the necessary ingredients to become a great winner.

1. Size

The company must be small and relatively unknown. Size matters because of the base effect: it is much easier for a company worth $300 million to become worth $3 billion than for a company worth $300 billion to become worth $3 trillion.

That is why multibaggers usually start from low market capitalizations, still modest sales, little analyst coverage and limited institutional presence. In other words: businesses that are not yet on the whole market’s radar.

That is where the opportunity appears. When a company is small, illiquid and barely followed, it is more likely to be mispriced. If the business improves and starts gaining visibility, the market may recognize its potential and pay higher multiples.

The key is to buy before it stops being invisible.

2. Quality of the business and the management team

A future multibagger cannot be just a small and cheap company. It also has to be good.

The literature usually insists on several traits: a proven business model, high ROE or ROCE relative to the industry, good capital allocation, the ability to reinvest for years at above-average returns, low competitive intensity, sector tailwinds and the potential to become one of the leading companies in its market.

It is not enough to buy a forgotten microcap and pray.

Usually some kind of moat is also needed, as Buffett would say: a sustainable competitive advantage that allows the company to defend margins, grow without destroying profitability and prevent competitors from eating the whole pie. If the business is also asset-light, even better, because it needs less maintenance CAPEX and can devote more cash to growth, buybacks or strengthening its position.

And then there is management, which is not a minor detail. A company can grow a lot and still not make shareholders rich if the management team allocates capital poorly, constantly dilutes shareholders, buys mediocre businesses or plays empire-building. For business growth to turn into shareholder returns, management has to be aligned and know how to reinvest well.

3. Growth in all its dimensions

A multibagger needs to grow. But not just grow revenue and that is it, because anyone can do that by burning money with enough enthusiasm.

Ideally, growth appears across several layers: revenue, cash flow, margins, earnings and, with a bit of luck, multiples too. First the business grows. Then profitability improves. Then earnings increase. And finally, if the market starts to believe the story, it may pay more for every euro earned.

The classic literature usually gives a lot of importance to earnings per share growth, the famous EPS. In theory, it is almost mandatory. If a company cannot turn its growth into higher earnings per share, something is wrong: either it dilutes too much, reinvests poorly, its margins do not hold up, or the business grows but the shareholder does not truly participate.

And if that EPS growth is accompanied by a high or rising ROE, even better. Because then the story starts to make more sense: the company is not only making more money, but it is doing so while using capital well.

4. Duration of growth

It is not enough to grow for two or three years and look pretty in the photo. For a company to become a multibagger, it needs a long runway ahead: a large market, low penetration, the ability to expand into new geographies, new products or new customers, and enough opportunities to keep reinvesting capital for years.

The key is not just growing fast, but growing for a long time. That is the difference between a good trade and a true compounder. Many companies can have a couple of spectacular years because of a cycle, a trend or a one-off recovery. But if growth depends on everything going perfectly during one very specific window, that is not a great machine for compounding capital. It is a bet that looks good.

That is why many multibaggers tend to be in relatively non-cyclical businesses, or at least in businesses with demand that is structural enough to withstand different economic environments. The longer, more stable and more reinvestable the growth, the easier it is for the magic of compound interest to do its job.

5. Attractive valuation at the time of purchase

Future growth cannot already be fully embedded in the price. Because a company can be excellent, grow a lot and still deliver mediocre returns if you buy it too expensively.

That is why the literature insists on buying at reasonable valuations: a low P/E, an attractive PEG, a moderate price-to-sales ratio or any other multiple that does not imply absurd expectations. It is not necessarily about buying extremely cheap companies, but about avoiding paying as if the perfect future were already guaranteed.

The ideal combination is a small, good company with a lot of growth ahead that the market is not yet valuing as if it were already the next great compounder. That is where business growth and multiple expansion can work together.

Buying well does not guarantee a multibagger, but paying too much can kill it before it even starts.

In summary

A big rise in the stock market usually comes from two engines:

1. The company earns more money.
2. The market pays more for each euro it earns.

It can be summarized like this:

Share price = Earnings per share × P/E

Or, in short:

P = EPS × P/E

That is, a stock can multiply because earnings per share grow, because the multiple expands, or, in the best case, because both things happen at the same time.

In logarithmic terms, or as a simple approximation:

Price growth ≈ EPS growth + P/E growth

That is where many great multibaggers are born: more earnings, a better valuation and many years ahead for the math to do the rest.

What is the problem with these studies?

After all of the above, it seems like the recipe is clear: small, good companies with a long growth runway, aligned managers and a reasonable valuation. You buy them, wait twenty years and that is it. Multibagger captured.

If only.

The problem is that the classic studies have several limitations:

1. Lack of empirical validation. Many ideas sound good, but when you test them against data, they start to wobble. Even dogmas such as the need for strong EPS growth are not as clear as they seem.

2. They are more descriptive than analytical. They observe companies that already multiplied by 10, by 50 or by 100 and then look for common traits. Useful, yes, but it does not prove causality.

3. They use criteria that are difficult to measure. “Good management”, “competitive advantage”, “business quality” or “good capital allocation” matter, but they are fairly subjective concepts and hard to test.

4. Many are outdated. Mayer goes up to 2014, and since then the market has gone through zero interest rates, inflation, growth bubbles, aggressive rate hikes, AI and quite a bit of madness.

5. Part of the evidence comes from markets that are not very comparable. Some studies focus on India, a market very different from the U.S. For investors who mostly look at U.S. stocks, something more applicable was needed.

So no, not everything was solved. Previous studies offer good intuition, but it was still necessary to take those ideas, put them up against the data and see which ones survive.

Context of the experiment

The study analyzes companies listed on the NYSE and NASDAQ, including ADRs, during the period 2009-2024.

The choice of period is not random. It starts just after the 2008 financial crisis, almost like a market “reset”, and covers 15 full years. Long enough for a truly good company to multiply several times over.

And it was not exactly a quiet period. It includes bull markets, bear markets, COVID, inflation, rate hikes and cuts, a banking crisis, wars, commodity shocks and major political changes. In other words, quite an entertaining laboratory.

The study identifies more than 500 stocks that reached a 10x return between 2009 and 2024, but only keeps those that maintained that gain until the end of the period. Those that touched the 10-bagger threshold and then fell below it are excluded. Companies with incomplete fundamental data are also removed.

The final sample contains 464 multibaggers.

Here it is worth paying attention to the design: the study does not only analyze the rise between 2009 and 2024, but also the prior history of these companies from the year 2000 onward. That is, it tries to look at what they were like before becoming big winners.

The underlying idea is good: not to stay with the final snapshot, but to look for signals that were already present before the big move. Because finding a multibagger after it has already multiplied by 10 does not have much merit. The interesting part is seeing whether there was any clue beforehand.

Starting point: the Fama-French 5-factor model

The analysis begins with the Fama and French five-factor model (2015), one of the most widely used frameworks for explaining why some stocks generate higher returns than others.

The idea, simplifying a lot, is that a stock’s return can be explained by its exposure to several factors: market, size, valuation, profitability and investment.

The formula is this:

R_it − R_Ft = α_i + b_i(R_Mt − R_Ft) + s_iSMB_t + h_iHML_t + r_iRMW_t + c_iCMA_t + e_it

In other words, the model tries to explain how much a stock has earned by comparing it with what a risk-free asset would have earned and by seeing how much of that return comes from different known factors.

The main ones are:

• R_Mt − R_Ft: the market’s excess return over the risk-free asset. Basically, how much the market pays for taking equity risk.
• SMB, Small Minus Big: the size factor. It measures whether small companies do better than large ones.
• HML, High Minus Low: the value factor. It compares cheap stocks with expensive stocks using the book-to-market ratio.
• RMW, Robust Minus Weak: the profitability factor. It compares companies with high profitability against companies with low profitability.
• CMA, Conservative Minus Aggressive: the investment factor. It compares companies that invest little against companies that invest a lot.

Then come the coefficients:

• b_i measures the stock’s sensitivity to the market.
• s_i, h_i, r_i and c_i measure the stock’s exposure to each factor.
• α_i is the famous alpha: the part of the return that the model cannot explain.
• e_it is the residual, the noise, what is left outside the model.

The beauty of the model is that it allows us to ask a very useful question: did multibaggers earn so much simply because they were exposed to known factors, such as size, value or profitability, or because there was something else?

And that “something else” is exactly what the study tries to find.

Where Fama-French falls short.

RmmarketSMBsizeHMLvalueRMWprofitabilityCMAinvestment

Standard universe

Diversified portfolios

simulated visual fit

Multibagger universe

464 10x stocks

simulated visual fit

Alpha and beta

If you have ever read about investment funds, you have probably heard two words: alpha and beta. Here they appear with a similar idea, but within a factor regression.

Beta, represented by b_i, measures how much a stock moves relative to the market. A beta of 1 means that, on average, the stock moves like the market. A beta above 1 implies greater sensitivity to market movements. A beta below 1 implies less sensitivity. It is the classic way of measuring “market risk”, although here it is only one piece within a model with several factors.

Alpha, represented by α_i, is what remains after explaining returns through those factors. In plain investor language: it is the part of the return that is not explained by exposure to the market, size, value, profitability or investment. If alpha is positive and persistent, the model is basically saying: “something else happened here beyond simple exposure to known factors”.

The important point is that alpha is not an explanation, but a clue. It tells us there is part of the return that the model cannot capture, but it does not tell us why. It may be due to a real characteristic of the company, a missing factor in the model or simply statistical noise. So it should be treated as a useful signal, not as definitive proof.

For now, that is enough. I will assume we already have a reasonable idea of what we are looking at. I still owe myself a post explaining the five-factor model more calmly, where it comes from and what its implications are.

The experiment

As its name suggests, the Fama-French model uses five factors to try to explain stock returns. If those factors explain the historical behavior of multibaggers well, great: the model works reasonably well for this type of company too. But if too much return remains unexplained, then something is escaping us and the model needs to be adjusted.

According to the original Fama-French theory, over the long term small, cheap, profitable companies with a prudent investment policy tend to do better. That is, companies with smaller size, attractive valuation, good operating profitability and more conservative asset growth.

To test this, the study follows a methodology similar to the original one: it independently sorts the companies in the sample into different groups, using data available as of January 1 for the 2000-2024 period.

The classification is as follows:

• 3 groups by size: small, medium and large, according to market capitalization.
• 3 groups by valuation: low, medium and high, using the book-to-market ratio, calculated as shareholders’ equity divided by market capitalization.
• 2 groups by profitability: robust or weak, measured as operating profit divided by book value.
• 2 groups by investment: conservative or aggressive, using annual growth in total assets as a proxy.

By crossing all these factors, we get 36 different portfolios:

3 × 3 × 2 × 2 = 36

The objective is twofold.

First, to check whether the classic theory holds within the universe of multibaggers. That is, whether small, cheap, profitable and prudent companies are also the ones that do best here.

Second, to analyze alpha. If, after applying the model, a lot of return still remains unexplained, that means these stocks generated returns for reasons that the five classic factors do not capture well. And that is where the real work begins: building a model better adapted to multibaggers, with variables that better explain where that extraordinary return really came from.

The results

The table may look like a small visual hell at first, but the logic is fairly simple. Companies are grouped according to the factors we have just seen: size, valuation, profitability and investment. Then the return of each group is colored to make it easy to detect which combinations work better and which ones work worse.

For example, the best portfolio, the greenest of all, appears in the combination formed by:

• Size small: small companies
• Value high: cheap companies, with high book-to-market
• Profitability robust: profitable companies
• Aggres: companies with an aggressive investment policy

In addition, the table also allows us to see the overall effect of each factor separately. For example, the last column shows the average return aggregated by size. There we see that small companies beat medium-sized companies, and medium-sized companies beat large ones. Exactly what one would expect.

The paper includes more tables and a few additional nuances, but to avoid getting lost in the forest, these are the main conclusions:

• Size effect: small companies obtain higher average returns than large ones. However, if we look at the median, the difference is no longer so clear. This suggests that size helps, yes, but it is not enough by itself. Buying small caps blindly is not a winning strategy.

• Value effect: within the universe of multibaggers, valuation still matters. Using the book-to-market ratio as a proxy, cheaper companies tend to offer better returns. Nothing too surprising.

• Profitability effect: controlling for the other factors, companies with weak profitability do worse than companies with robust profitability. In other words, quality matters too.

• Investment effect: this is where the first break with the classic theory appears. According to Fama and French, companies that invest more aggressively, measured as asset growth, should offer worse future returns. The idea is that growing too quickly tends to end badly. But in this sample, almost the opposite happens: controlling for size, valuation and profitability, companies with higher asset growth obtain better returns than conservative ones. It makes sense. A company that wants to multiply cannot stand still. It has to reinvest, expand capacity and build something much bigger.

This is where the story changes.

The study then runs a regression to measure to what extent these five factors explain the returns of multibaggers and, above all, to check what happens with alpha.

Several important signals appear here. The coefficient of operating profitability is close to zero, suggesting that this variable contributes rather little when it comes to explaining future returns within this sample. In addition, the beta of these stocks is high. And, as we already suspected, alpha remains too high.

Translated: the five-factor model does not manage to explain multibagger returns well. Something important is happening here that the model misses. And that is exactly why this analysis is interesting: not because it simply confirms the classic theory, but because it shows where that theory begins to fail for such a particular group of companies.

Improving the model

The study tries to adapt the classic Fama-French model to the specific case of multibaggers. Because if the model leaves too much alpha unexplained, the conclusion is fairly clear: important variables are being left out.

To do this, the author tests different ways of measuring each factor:

• Size: market capitalization, enterprise value, assets, sales, etc.
• Valuation: book-to-market, P/E, price-to-sales.
• Profitability: operating margin, net margin, EBITDA margin, ROE and return on capital.
• Investment: asset growth and new variables that compare that growth with EBITDA growth and free cash flow growth.

Then the study compares the models using different statistical criteria. The best result appears with a fixed effects model, basically because each company has its own story. Not all companies are comparable as if they were identical pieces.

In an intermediate version of the model, the author changes three main variables:

• She uses TEV instead of market capitalization to measure size.
• She uses P/E instead of book-to-market to measure valuation.
• She uses EBITDA margin instead of operating profitability.

This does not mean that P/E ends up being the best valuation measure. In later, more complete models, the author ends up avoiding it as a proxy for value because it did not prove useful in the empirical analysis and introduces too much noise: it does not work with loss-making companies and it explodes when earnings are very small. That is why the valuation variables that end up carrying more weight are B/M and FCF/P. FCF/P, or free cash flow yield, is free cash flow divided by price or market capitalization. It is a simple way of asking: “How much cash does this company generate for every euro I pay?” The study treats it as a valuation and profitability metric based on cash generated relative to price.

But the most actionable finding is in how the study measures investment.

The study introduces a dummy that takes the value 1 when assets grow faster than EBITDA. And the result is quite powerful: when a company expands assets faster than EBITDA, the following year’s return falls by about 22.8 percentage points.

This is one of the most useful ideas in the study.

Multibaggers need to invest. They need to grow. They need to expand capacity, open markets, hire people, develop products, acquire assets or build infrastructure. A company that wants to multiply cannot behave as if it were already mature.

But that investment has to be accompanied by real EBITDA growth.

Investing a lot is not enough. If assets grow but EBITDA does not keep up, the company is probably buying bad growth, inflating the balance sheet or reinvesting capital at mediocre returns. And that, in the long run, does not create a multibagger. It creates a disaster.

In short: the improved model suggests that the best multibaggers tend to be small, cheap, profitable companies capable of investing aggressively, but sustainably. It is not about growing for the sake of growing. It is about growing with earnings behind it.

And this improves the model quite a bit.

But the analysis does not stop there.

Static and dynamic return models

Here the objective of the paper changes. The author is no longer simply trying to check whether multibaggers fit inside the classic Fama-French model, but to build a more complete model to explain their future returns.

To do this, she tests more than 150 variables: growth, valuation, profitability, quality, debt, solvency, momentum, interest rates, analyst coverage, investment, R&D, marketing, sector comparisons and quite a few more things. In other words, much more than the typical “small, cheap and profitable company”.

To separate what is useful from the noise, she uses Hendry’s general-to-specific methodology. The idea is simple: you start with a huge model, full of variables and lags, and gradually eliminate what does not add value until you are left with a smaller, cleaner and more robust version.

First you throw everything into the pot. Then you start removing ingredients until the thing tastes like something.

The model is estimated using data from 2000 to 2022, leaving 2023 and 2024 outside the sample to check whether it really works with new data. This is important, because one thing is to explain the past very well and quite another is for the model to hold up when you show it years it has not seen. They also detect common problems in financial data, such as heteroskedasticity and autocorrelation, so they adjust the statistical errors using clustered errors.

The ambition here is greater than in the classic studies. It is not just about describing what multibaggers looked like after they had multiplied by 10, but about trying to identify which variables best explained their future returns before they happened.

And this paper stands out precisely in that respect.

It moves away from the typical descriptive story we are used to and toward a more serious attempt at a predictive model: more variables, more statistical control, more filters and an out-of-sample test to see whether the result has some substance or whether it was simply overfitting the past.

The result is not perfect, but this is where the conclusions that matter most for investors appear.

Main results

Overall, the model behaves quite reasonably: almost all coefficients have the expected sign. The variables that should help, help; those that should penalize, penalize. There is the occasional exception, especially related to earnings quality, but the general message is fairly clear.

• The market matters. The return of the S&P 500 appears as significant in all models and with a positive sign. Multibaggers are also affected by the general environment. When the market helps, it helps them too. When things get complicated, they feel it as well.

• Size still penalizes. Size, measured with TEV, appears as highly significant and with a negative coefficient. The larger the company, the lower its future return tends to be. It makes sense: it is much easier for a small company to multiply by 10 than for a gigantic company to do so.

• Profitability matters, but less than one might expect. In the dynamic models, EBITDA margin loses strength and ends up being replaced by ROA as the best profitability variable. More profitable companies tend to do better, yes, but the effect is not spectacular. In fact, a variable close to valuation and profitability, FCF/P, ends up carrying much more weight.

• Accounting growth disappoints quite a bit. Here comes one of the most curious parts of the study: many growth variables do not come out as significant. Not revenue growth, not EBITDA growth, not EPS growth, not free cash flow growth, not per-share growth, not 1-year growth, not 5-year CAGR. This clashes with the popular literature on multibaggers, which tends to insist heavily on looking for companies with strong and sustained earnings growth. In this model, that idea is not so clear. Asset growth does appear as significant in some cases, but with a small impact.

  This is not about saying “growth does not matter”. Many of these companies grew a lot: the appendix of the study reports EPS at a 20.0% CAGR, net income at 22.9% and operating income at 17.3%. The right way to look at it is this: once you are already looking at companies that ended up becoming multibaggers, isolated accounting growth does not seem to explain future returns as well as the price paid, free cash flow yield and the quality of investment.

• Investment matters, but with one key condition. The most actionable variable is again the investment dummy. When assets grow faster than EBITDA, the following year’s return tends to be between 4 and 11 percentage points lower, controlling for the other factors. The reading is very useful: a multibagger needs to invest, grow assets and build capacity, but that investment has to come with real EBITDA growth. If assets grow and EBITDA keeps up, good. If assets grow but EBITDA does not, be careful. The company may be buying bad growth, inflating the balance sheet or reinvesting capital at mediocre returns.

• Interest rates carry a lot of weight. In the more advanced models, a rising Fed rate environment appears as a very negative factor. When rates are rising, future returns for multibaggers fall by approximately 8 to 12 percentage points relative to the risk-free asset, controlling for the other variables. This fits quite well with intuition: growth companies depend more on future cash flows. If the discount rate rises, the present value of those cash flows falls and valuations suffer. This variable is not very useful for choosing specific stocks, but it is useful for adjusting expectations about the entire universe.

• Valuation is the main protagonist. The most powerful variables in the model are book-to-market and free cash flow to price. Both are positive, highly significant and have large coefficients. The conclusion is fairly direct: future returns are strongly linked to the initial valuation. Companies that are cheaper relative to their book value and cash flow tend to generate better returns. As we all already knew, the debate between growth and value is poorly framed. The best growth stocks also have to be reasonably cheap. Growing a lot is not enough. The price you pay at the beginning matters, a lot.

• P/E does not work well. Although it is one of the most widely used metrics, or perhaps precisely because of that, in this study it does not prove particularly useful for predicting returns. The problem is that P/E breaks easily: if a company loses money, it makes no sense; if it earns very little, it can shoot up to absurd levels. That introduces a lot of noise into the model. That is why the study prefers metrics such as B/M and FCF/P, which seem to capture the relationship between price and fundamentals better.

• Momentum works in a strange way. Classic momentum says that stocks that have recently gone up tend to keep going up. But in this sample the effect seems very short-lived and with quick reversal. One-month momentum appears positive in only one model, while three- and six-month momentum appear negative. In addition, the closer the stock is to its 12-month high, the lower its future return tends to be. Chasing a multibagger right after a big rise can be expensive.

There are also many variables that, surprisingly, do not add much: leverage, debt coverage, Altman Z-score, increases or reductions in debt, buybacks, dividends, share issuance and R&D intensity. Here we have to be careful, especially with the solvency variables. The fact that they are not significant within a sample of companies we already know survived does not mean they do not matter. It is ex-post selection bias among winners. For the winners that reached the end, debt may not explain much of the subsequent return. But to avoid dying along the way, it may be crucial.

It is worth slowing down a little with R&D. One might expect companies investing more in innovation to generate better returns, but the study does not find a clear relationship. That does not mean innovation does not matter, but rather that R&D spending, measured in a simple way, may not capture the quality of that innovation very well.

As a curiosity, dividends do appear as significant in some static models, although not in the dynamic ones. Even so, the data slightly breaks the cliché: many multibaggers paid dividends. At the beginning of the period, 58% of the sample did so, and in January 2024, 78% did. A great winner does not necessarily have to be a company that reinvests absolutely everything and distributes nothing.

Finally, the study finds that some lagged variables help explain future returns and therefore “Granger-cause” the returns of multibaggers. This sounds stronger than it really is. It does not mean real economic causality, as in a controlled experiment. It means those variables help predict. They may be capturing a real signal, but also sector exposure, liquidity, investor sentiment, credit conditions, hidden stress or any other variable that the model has not measured directly.

In short: the model points to a fairly powerful idea. The best multibaggers are not simply companies that grow a lot. They are small, reasonably cheap, profitable companies, capable of investing without destroying capital and bought before the market prices in too much of the future.

Conclusions

The main findings of the study would be these:

• It questions some dogmas about multibaggers. Not because growth does not matter, but because isolated accounting growth explains less than one might expect within a sample already filtered by winners. In this study, valuation, free cash flow yield, size, interest rates and the discipline between assets and EBITDA carry more weight.

• Size still matters. Small, cheap and profitable companies tend to generate better returns. Nothing too revolutionary, but consistent with previous literature.

• Aggressive investment can be good, but not just any investment. If assets grow and EBITDA keeps up, perfect: the company may be building real capacity. If assets grow faster than EBITDA, be careful. Future returns get worse.

• Free cash flow yield is one of the most important variables. Growing is not enough. The company also has to generate cash and trade at a reasonable price.

• Interest rates carry a lot of weight. In a rising Fed rate environment, next-year returns relative to the risk-free asset are about 8-12 percentage points lower, controlling for the other variables. Multibaggers are not immune to the cost of money.

• Momentum works strangely. It does not seem to simply reward the stocks that have risen the most. In fact, the closer a stock is to its 12-month high, the worse its future return tends to be.

• The entry point matters enormously. The best opportunities seem to appear when the stock is near its 12-month low and, better still, after having fallen quite a lot during the previous six months.

In short: a multibagger is not simply “a company that grows a lot”. According to this study, the most attractive combination would be something more like this: a small, cheap, profitable company, with good free cash flow yield, capable of investing without destroying capital and bought at a time when the market is not yet too excited.

So, yeah, it was never going to be easy.

Congratulations if you made it this far. I hope the read was useful and, if you enjoyed it, you can take a look at the rest of the blog.

---

Appendix I - Past studies

The first serious attempts to find profitable investment strategies go back a long way. As early as the 1930s, Wyckoff proposed a stock selection method based on the evolution of price, volume and market psychology. His approach consisted of detecting accumulation and distribution phases within stock market cycles in order to try to enter before the big moves. In other words, many of the ideas that would later form the basis of modern technical analysis.

Almost at the same time, Graham and Dodd, in 1934, laid the foundations of fundamental analysis. Their idea was very different: look at the intrinsic value of a company, demand a margin of safety and analyze its financial statements to find undervalued stocks that the market was overlooking. Some looked at price behavior. Others looked inside the business. And since then, investing has lived more or less trapped between those two obsessions: what the market is doing and what the company is really worth.

Since then, many authors have tried to answer the same question: what type of stocks beat the market? What characteristics do they have in common? Is there any reasonable way to find them before it becomes obvious to everyone? The problem, of course, is that not everyone has reached the same conclusion. In fact, a good part of financial literature is basically an elegant fight, with papers, formulas and regressions, over whether the market is more or less efficient or whether, on the contrary, it is constantly making mistakes like any human being with fear, greed and too much desire to extrapolate the last quarter.

On one side, defenders of the Efficient Market Hypothesis, associated with Fama in 1970, argue that prices already reflect all available information about companies and their prospects. According to this view, stocks generally trade close to their fair value and it does not make much sense to try to find superior opportunities simply by choosing “the best stocks”. In other words: if the market already knows everything, you are not going to find bargains by looking at four ratios on a website.

On the other side, defenders of the Overreaction Hypothesis believe that markets are full of inefficiencies: information asymmetries, mass psychology, absurd extrapolations, panics, trends, narratives and, in general, humans doing human things. I find this second view much more convincing. Not because the market is stupid, but because sometimes it overshoots. It punishes too much. It gets too excited. It gets too scared. And when it exaggerates, sometimes the opportunity appears.

Yartseva’s study sits quite close to that second idea. Now, although the general debate about the sources of superior stock market returns has generated rivers of ink, the specific analysis of multibaggers remains almost a footnote. There is a huge amount of literature on value, momentum, quality, size, investment, profitability, volatility and the other classic factors. But there is considerably less on stocks that multiply by 10, by 50 or by 100. And what little there is usually moves between two worlds: on the one hand, very rich qualitative stories; on the other, still fairly limited statistical attempts.

One of the best-known works is that of Thomas Phelps, focused on 100-baggers. He analyzed the period between 1932 and 1971, identified 365 stocks that multiplied by 100 and tried to find what they had in common using examples, specific cases and a lot of qualitative analysis. His idea was to look for small, relatively unknown companies, with new products, new materials or new ways of producing. Companies that solved real problems, improved people’s lives, had strong earnings growth, room to keep expanding and good management. And then came the hardest part of all: buying them and holding them for a long time without sabotaging yourself along the way.

Phelps summarized it with a very good phrase:

“To make money in stocks you must have the vision to see them, the courage to buy them and the patience to hold them.”

Although Phelps’s work was more descriptive than statistical, it became almost legendary in the investing world and served as a foundation for many later studies on multibaggers. His great merit was not building a perfect model, but putting a very powerful idea on the table: great fortunes in the stock market usually do not come from correctly timing small moves, but from finding a few exceptional companies and not selling them too early.

Chris Mayer did something similar years later with 100-baggers. Drawing heavily on Phelps’s ideas, he popularized the coffee-can portfolio approach: buy extraordinary companies and leave them alone for many years, ideally at least a decade. The idea is simple, almost insultingly simple: if you really own a company capable of multiplying its value many times over, your biggest enemy is probably not the market, but you pressing buttons.

Mayer suggested paying attention to companies with characteristics such as these:

1. Long periods of earnings growth accompanied by multiple expansion, such as P/E, price-to-sales, etc.
2. Accelerating earnings growth, not merely stable growth.
3. High ROE, usually above 20%.
4. Founders or managers with a lot of talent and a lot of skin in the game.
5. Forgotten or punished companies that turn around and become profitable again.
6. Small companies, not enormous giants, because they have more room to multiply.

The idea that small companies can generate higher returns has quite a bit of empirical support. Not just because of common sense, but also because of works such as Fama and French, many later tests of their factor models and specific studies on multibaggers. Size matters. Or better said: the starting point matters. A company that starts from a small base has more room to grow, more room to surprise and, if everything goes well, more potential to multiply.

Another relevant attempt was Oswal’s, in his 2014 Wealth Creation study, where he tried to build on the qualitative ideas of Phelps and Mayer using a simple statistical analysis of multibagger stocks. In his case, he focused on the Indian market and identified 47 stocks that multiplied their value by 100 over the previous 20 years. The technology sector was the biggest wealth creator, although he also found 100-baggers in pharmaceuticals, banks, consumer retail, automobiles, construction materials and other sectors.

A curious fact: Oswal found that the Indian market itself, represented by the BSE Sensex index, was also a 100-bagger. It multiplied by 100 between 1979 and 2006, with a compound annual return close to 19%. In India, the average time needed to multiply by 100 was about 12 years, equivalent to a CAGR of 47%. Much faster than in developed markets, where Mayer estimated an average close to 26 years.

Oswal’s recommendation was to look for small, little-known companies with sustainable earnings growth, good management and low valuations, often with single-digit P/E ratios. He called this philosophy QGLP: Quality, Growth, Longevity, at reasonable Price.

In other words: quality, growth, duration and reasonable price.

According to his analysis, for a stock to multiply by 100, two engines are needed: earnings growth and multiple expansion. It is not enough for the business to grow. The market also has to start valuing it better. That is a key idea, because we often think of multibaggers as simple stories of business growth, when in reality they are usually a combination of several things happening at the same time: more sales, better margins, more earnings, a stronger narrative and a market willing to pay more for each euro earned.

That is, his idea was that to achieve life-changing returns, one has to look for growth in all dimensions. Growth of the business, growth of earnings, improvement in profitability and, if possible, an initial valuation low enough for the multiple to also play in your favor.

---

Appendix II - Limitations

Before turning this into a screener, it is worth ruining the party a little. The paper offers useful ideas, but it has several important limitations:

1. The sample is already filtered by winners. It studies companies that we already know were multibaggers. That introduces survivorship bias and means the results cannot simply be copied to search for future winners.

2. The “out-of-sample” is not completely clean. Although 2023-2024 are reserved to validate the model, the sample is selected using the entire 2009-2024 period. In other words, there is future information in the initial selection.

3. It mixes pre- and post-success signals. Some companies may have become 10-baggers early and then contributed years of data after the big rise. The model may be learning what winners look like once they have already won.

4. There may be overfitting. The study tests more than 150 variables, lags and different specifications. With so many tests, some relationships may appear significant by chance. More robustness tests would be needed.

5. The risk adjustment is limited. It uses return relative to the risk-free asset, but it does not properly capture beta, volatility, drawdowns, liquidity, execution costs or real implementation capacity.

6. It ignores dividends and costs. It uses price returns, not total returns, even though many multibaggers paid dividends. This is odd. It also excludes spreads, slippage, taxes and transaction costs, which are key in small caps.

7. The period and market are very specific. 2009-2024 in the United States was a special environment: low rates, QE, the tech boom, COVID, inflation and aggressive rate hikes. It is not clear that the conclusions would work the same way in other countries or cycles.

8. Important variables are missing. The model does not properly include sentiment, narrative, positioning, flows, short interest, options or retail investor attention, all of which can matter a lot in big winners.

9. It is still a working paper. It is exploratory evidence, not a validated formula for finding multibaggers.

In other words, the study contributes new and counterintuitive concepts that are worth keeping in mind, but it is not a panacea.

Appendix III - Descriptive statistics of the sample

The study includes some final data on the sample, which I leave here as a curiosity. These data refer to the final sample of 464 multibaggers analyzed.

1. On average, the study reports that these stocks multiplied their price by 26 times over the 15-year period, between 2009 and 2024, with an average CAGR of 21.4%.

Within the sample there were also 24 stocks that multiplied by 100.

2. At the beginning of the period, in 2009, these companies were relatively small:

   • Median market capitalization: $348 million
   • Median revenue: $702 million

   They were small companies, with enough size to have a real business, but still with room to multiply.

3. Growth over the 15 years was good, although not necessarily spectacular across all metrics:

   • Revenue: 11.1% CAGR
   • Gross profit: 12.0% CAGR
   • Operating income: 17.3% CAGR
   • Net income: 22.9% CAGR
   • EPS: 20.0% CAGR
   • R&D expense: 15.1% CAGR

   Revenue growth was not insane. We are not necessarily talking about companies that compounded revenue at 40% per year for 15 years.

   The big difference appears further down the income statement: operating income, net income and EPS grew much faster than revenue. That suggests margin improvement, scalability, operating efficiency or buybacks, depending on the case.

4. In terms of valuation, in 2009 these companies did not look expensive:

   • Median price-to-sales: 0.6
   • Price-to-book: 1.1
   • Forward P/E: 11.3
   • PEG: 0.8

   In other words, many of these multibaggers started out as small, reasonably cheap companies with undemanding valuations.

5. And in terms of profitability, at the beginning, they were rather normal:

   • Gross margin: 34.8%
   • Operating margin: 3.9%
   • ROE: 9.0%
   • ROC: 6.5%

   Not all of them started as perfect compounding machines with extremely high margins and brutal returns on capital.

In other words, many looked like small, cheap businesses, somewhat mediocre in initial profitability... but with plenty of room for operational improvement.

The general picture would be this:

cheap small caps, with real businesses, decent growth, fairly average initial profitability and a lot of room for operational improvement.