How To Be A Value Investor In The 21st Century?
Grahamian value investing don't work anymore, Buffettology works but the alpha is limited; so, what to do instead?
One of the things that has always bothered me most on Substack and Twitter is the constant bombardment of so-called “undervalued stocks.”
If you have been a bit active on these platforms, I urge you to take a look at your home feed, and you’ll see what I mean. There’ll be dozens of “undervalued stock” pitches.
I look at them, and I can’t say all those people are bullshitting. There are indeed very savvy writers, detecting what looks like truly undervalued stocks.
The problem is that most of those people don’t generate alpha when it comes to investing themselves. I see there are a few truly transparent names, and they are returning between 12 and 16%, which is acceptable, but it’s basically the market performance in recent years.
This isn’t something special to them; the broader value factor is also underperforming:
“Wait a minute, don’t you also consider yourself a value investor?”
This is probably the question that has just popped up in your mind if you have been following this publication for a while. I indeed do.
Yet, the way I see value investing in the 21st century is quite different from traditional value investing. This is what I’ll try to explain today.
We’ll tap into two main branches of value investing:
Grahamian value investing
Buffett-Munger Style (Buffettology)
I’ll go through what I think is wrong with them and why they don’t actually fit the 21st century.
Then, I’ll try to explain how I see value investing today and the 5 guiding principles everybody needs to understand to become a good investor in the 21st century.
Let’s dive in!
The End Of Grahamian Value Investing
I started as a pure value investor, like most investors.
The Intelligent Investor was the first book I read about investing, and I fell in love with the idea that something could be priced less in the market than it is really worth and that you can spot it by looking at financials.
When you spot one, you bet on it, and when the market corrects it, you pocket the difference. You could literally make money purely by your brain power.
I carried the book everywhere with and Benjamin Graham was elevated to a prophet-like figure in my head. So, I have been there. I fell in love with it. But at some point, I had to accept that there isn’t much room for the style of investing praised in the book.
Let’s start by elaborating on the investing style praised in the book. This is called with names like “Grahamian value investing”, “net-net investing”, “cigar-butt investing”, etc.
It’s called value investing before all because it concerns itself with one question above all—what is the company worth today?
You have to find the value of the company and compare it to the market value.
The value of a company is made up of:
Earning power
Assets
Earning power is hard to assess because conditions change. There are some businesses whose earnings are predictable. Graham calls these blue chips in the book.
The problem is that the blue chips are often appropriately priced by the market, as they are known by everybody, and there are many analysts looking at those stocks. If you exclude blue chips, earnings become too erratic to reliably value companies.
The solution Graham brought to this was completely eliminating the earning power from the relevant factors by looking at the cases where the net current asset value was larger than the market capitalization.
Net current asset value (NCAV) is calculated as: (Current Assets)—(Total Liabilities)
He even refined this further by marking down receivables to 80% and inventory to 66% of their face value, as there was no guarantee of collection for receivables and inventories could be worth less in the market than they did on the balance sheet:
He also completely ignored fixed assets, as they rarely reached their face value in the case of liquidation. The goal here was simple—find how much cash equivalent the company had.
He then subtracted all liabilities from the liquidation value of current assets. If the business were still left with a positive net asset value and that value was larger than the market cap of the company, it meant that the company was trading at pennies on a dollar.
Yet, there were still two problems:
The market didn’t have to take notice and correct the undervaluation.
The company could misuse its current assets, eroding the value.
Graham came up with two solutions for these:
First, he demanded that the companies had no negative earnings in the past 5 years.
Positive earnings ensure that the company doesn’t destroy value, so that what’s once undervalued doesn’t become fairly valued. Further, every positive earnings report acts as a catalyst that prompts the market to take notice of the company and close the valuation gap. Consistent earnings don’t guarantee future earnings, but they at least show a tendency for it. This is the best you could get.
Second, he had a very diversified portfolio and often owned more than 100 stocks:
This had two functions:
First, it was an insurance against catastrophic losses in any position.
Graham was buying stocks on a quantitative basis of undervaluation, as we explained above. He wasn’t very interested in the business model, future prospects, etc. By diversifying, they were taking a measure against their own willing ignorance. If one company got disrupted and went bankrupt, it wouldn’t affect them much.
Second, it was an insurance against uncertainty.
They knew that the undervaluation didn’t mean it would be corrected soon. An issue could remain undervalued forever. Thus, they avoided cherry picking and created large baskets so that the market would notice at least a few of them. They furthered this by exiting when the issue went up by 50% or after two years of holding.
In short, they were blind to the essence of their holdings, and they worked on a quantitative basis almost algorithmically in an era when software-driven algorithmic trading didn’t exist.
Today, there are several problems with this method.
1️⃣ I think the first one is obvious—it was a strategy of the analogue era.
When Graham came up with this method, computers weren’t mainstream, software programs didn’t exist, and you couldn’t screen stocks on large databases.
Information edge existed.
Warren Buffett often mentioned that he read thousands of pages of Moody’s Manuals to find off-the-path stocks that presented asymmetric opportunities. These manuals were printed separately for each sector, and each one took thousands of pages.
It was possible that you could find things simply by turning over more rocks and reading more than other people. It’s no longer possible today.
All the stocks and their fundamentals are in databases in a structured format, making it very easy to screen. There are also very developed software and algorithmic trading programs that can execute trades based on any criteria at a speed that can’t be matched by any human.
As a result, those net-nets are easily detected and traded by algorithms. When there are hundreds of algorithms doing this, money quickly flows to those stocks, and valuation gaps get quickly bridged, making it harder for smaller investors to capitalize on them.
2️⃣ Second, capital markets today are much more developed.
Back in the 1960s, private capital was nascent, so these opportunities could exist in the public markets without becoming targets for acquisition.
Think about Graham’s stock selection criteria:
No negative earnings in the last 5 years.
Net current asset value is larger than the market cap.
Basically, a money-making business selling for pennies on a dollar, a dream investment for private equity, but they existed in public markets because there was very limited private equity activity.
This is no longer the case; private equity volume reached $600 billion last year:
The assets under management by private equity firms are around $10 trillion today…
As Graham’s net-nets are the perfect candidates for acquisition by private capital, their numbers have significantly declined. They are like species at the edge of extinction.
This doesn’t mean they don’t exist, but there are simply way fewer of them now.
3️⃣ Quality Problem and Holding Period Uncertainty
As a result of digitalization, attractive net-net opportunities have become easily detectable, and the best of them have been acquired by private equity as private capital has developed massively since the 1980s.
Thus, the remaining net-nets in the market are of inferior quality. They often don’t have reliable earnings, and the business fundamentals generally keep deteriorating, justifying the discount.
There are still high-quality net-nets, but they are largely in the markets where private capital isn’t much developed. If there were a developed private equity industry, higher-quality net-nets would be acquired.
The problem is that these markets without a developed PE industry generally don’t have developed public markets either. They often have very low liquidity. In the absence of liquidity, even the highest quality net-nets can remain undervalued for very long times.
Indeed, the US market is unmatched in terms of the size-adjusted speed of mean reversion.
According to research by the Utrecht School of Economics, only the South African, Swiss, and Swedish markets had faster mean reversion than the US markets between 1900-2008. However, they are much smaller. When you adjust for the size, the US markets are unmatched:
This is the very essence of Grahamian value investing. You create a well-diversified portfolio of stocks that are susceptible to rerating simply due to mean reversion. When this happens, you harvest, and you don’t wait longer than two years. In the absence of liquidity, the strategy doesn’t work regardless of the quality of the net-nets.
All these three factors are making it very challenging to follow Graham’s way in today’s markets. It’s almost impossible for a retail investor to:
Create a portfolio of 100 high-quality net-nets.
Observe each one of them and exit either after 50% upside or two years.
Add new stocks instead of the exits and repeat the strategy indefinitely.
It’s simply not possible; there isn’t enough supply for the reasons I have explained above. And it’s already been perfected by the algorithms to the degree, it’s possible.
This is the problem with the deep value hunters on Substack or Twitter.
They find the deep value, but they are just throwing them on us as if they are saying, “Here is a deep value stock, do what you want to do with it.”
People are taking these stocks and mingling them in their portfolios, mixing them with other high-quality stocks, growth stocks, etc. What often happens is that the “deep value pick” stays a laggard, dragging the performance of the portfolio down.
They weren’t meant for those portfolios. They were meant for the portfolios made up of over 100 similar stocks, so you could systematically benefit from the mean reversion.
Those writers aren’t giving it. And even if they tried, they would likely fail miserably, as there is no way anybody can do it better than the machines. The essence of the strategy is that you don’t care about the business. You are buying things on a quantitative basis. Nobody can do it better and faster than the algorithms.
In short, Grahamian value investing works when it’s applicable, but there isn’t much room left to apply it.
Enter Buffett.
The Buffettology: Modern Value Investing
If the defining question of Grahamian value investing was “what’s it worth today?”, the defining question of the Buffettology is “what will it be worth in the future?”
Benjamin Graham didn’t trust earnings, and didn’t want to think like a businessman, look at the business qualities, and predict the future earnings. He wanted to buy a dollar for cents now; he didn’t want to pay two dollars today for the potential of four dollars tomorrow.
Thus, in Graham’s way, future earning power was disregarded in valuation, and earnings were demanded solely to make sure that the value gap wouldn’t close downward as a result of the business’s losing money.
Buffett had internalized this way, but it started to change when he met Charlie Munger in 1959.
Buffett was a Grahamian, but Munger was more impressed by Philip Fisher’s ideas. Fisher had published a book called “Common Stocks and Uncommon Profits” in 1957, just 8 years after the Intelligent Investor.
Fisher’s method was different than Graham’s. His simple idea was that if you find a business that will make more money a few years down the road, it should be priced higher than today, assuming no multiple contraction.
This has a few implications:
It emphasizes durability by assuming no multiple contractions. If the business is valued at 15x earnings today and again valued at 15x earnings 5 years from now, its business should be a durable one.
It champions the concept of moat. Businesses that can grow earnings consistently are those that deploy increasingly more capital at higher rates of return. Businesses that can achieve this have some sort of unfair competitive advantage, or moat.
It emphasizes long-termism. You need time to see the effect of durability. A business can make $1 today and be priced at $15 or 15x earnings. It can make $1.15 next year and can still be valued at $15. But, if it keeps growing like this for 10 years, it’ll generate $4 in earnings at $15 stock price would be ridiculous.
It cautions against overpaying. If you pay, let’s say 15x for a growing business, there isn’t much room for multiple contraction. But if you pay 60x, you can still lose money just due to multiple contraction even if the business kept growing in the 10 years after you bought it.
Fisher shifts the focus from tangible value to earning power.
If you are looking at a business with unfair competitive advantages, thus predictable earnings, you can rely on forward-looking assumptions to calculate its future value. You discount it to the current day by using an appropriate discount factor. If what you find is higher than the market price, you buy and wait for it to play out.
Buffett was clearly impressed by this idea and very well understood its tenets. He put this on a test with his American Express investment in 1964.
Buffett explicitly said this in a CNBC interview back in 2017:
“I had learned that from a fella named Phil Fisher who wrote this great book called “Common Stocks and Uncommon Profits.” And he calls it the scuttlebutt method. And Phil was a remarkable guy. And I first used it back in 1963 when American Express had this great Salad Oil Scandal that people were worried about it bankrupting the company. So I went out to restaurants and saw what people were doing with the American Express card, and I went to banks to see what they were doing with travelers’ checks and everything. And clearly American Express had lost some money from this scandal, but it hadn’t affected their consumer franchise.”
Warren Buffett on CNBC, February 2017
(You access the full interview here: https://www.cnbc.com/2017/02/27/billionaire-investor-warren-buffett-speaks-with-cnbcs-becky-quick-on-squawk-box.html)
So, he decided that American Express was one of those exceptional businesses with a wide moat. Financials were approving Buffett’s findings. From 1954 to 1963, the company grew its revenues from $37 million to $100 million, achieving a 10.45% CAGR over that period.
He decided that the company could sustain a similar growth rate for at least a few more years.
I don’t think Buffett did a DCF valuation at the time for this investment, but if he did one, it would have looked something like this:
Buffett exited this investment partially in 1967 and fully in 1968. According to some sources, he doubled his money; others say he did more than 3x.
In any case, this taught Buffett that Fisher’s method works.
However, I think Buffett comprehended the real power and potential of this style back in the early 1970s. He had acquired the Berkshire Hathaway textile company in 1965, and he didn’t know what to do with it. It is said that Charlie Munger advised him to turn it into an investment vehicle.
Munger reasoned that there weren’t many opportunities to generate high returns on capital for a textile company. Instead, Buffett advised Buffett to draw the earnings and reinvest what’s left of maintenance capex into exceptional companies that actually have opportunities to generate high returns on capital.
Munger’s advice resonated with Buffett, and he closed his partnership in 1969 and decided to continue with Berkshire as his primary investment vehicle.
At around that point, Buffett should have noticed that, or he had always been aware after his exit, American Express was still growing 10% annually. He must have understood that exceptional businesses could improve their earning power consistently for way longer periods than he initially thought.
From then on, he dedicated Berkshire to invest only in exceptional companies that would improve their earning power for decades to come.
This was a significant shift from Graham’s approach, which emphasized current value based on assets and saw earnings just as an anchor of value to Fisher, who incorporates future earnings as the main value driver.
He said openly in Berkshire’s 2000 Annual Meeting that they ignore book value and look at the company’s earning power in valuation:
This worked exceptionally well for Buffett and Berkshire, but there are a few problems that obstruct individual investors today from generating similar results.
First, Buffett could tap into private markets and reallocate his permanent capital base when the return on capital declined on a particular business.
The example is a coffee shop with a strong name in its neighbourhood. That coffee shop can likely succeed with the second store in the outer corner of the neighbourhood, and that investment would generate a high return on capital. However, that return drastically declines if it opens a store in another neighbourhood that has its own popular coffee shop.
Buffett could buy See’s Candies, generate a high return on capital by growing it in California. However, that return would decline if it tried to expand to Texas. When the opportunities for high return on capital declined, Buffett took the capital from that business and used it to buy something else, like Nebraska Furniture Mart.
Public market investors can’t do this. When you exit a stock, it doesn’t keep sending you checks to finance your next positions. Thus, you are confined to truly exceptional companies in public markets, i.e, Google’s, Visa’s, American Express’s, etc.
The second problem is that those exceptional companies in public markets are very rare, thus they are very well known.
As a result, they are almost always expensive. You can find them cheaply only in three circumstances:
Market crashes.
Industry recessions.
Temporary company-specific problems.
You see this clearly when you look at Buffett’s investing in public markets for the last 3-4 decades.
He builds massive cash positions, waits for them to become cheap due to some systematic stress, and deploys massive amounts to well-known exceptional companies:
The problem is that these situations also happen rarely. It’s fine for Buffett because he is managing massive amounts, and he readily accepts that stellar results are not possible for him due to his size, and this strategy is reliable in generating something between 10-15% annual return.
But for individual investors who are managing way smaller sums, this method is too restrictive, and opportunities here will also dry rapidly as systemic crises get rarer due to our improving sophistication and institutional risk management mechanisms.
So, what to do instead? How to be a value investor in the 21st century?
Post-Modern Value Investing: No-Rules Wilderness
Above, I have explained that value investing has evolved from a rule-based discipline to more of a principle-based one.
In Graham’s way, you can invest simply by the rules, not knowing a dime about the business. However, in Buffett’s way, you need to understand the qualities of exceptional businesses and use them as guiding principles in your decisions.
I think modern value investing is almost completely detached from the rules, and it’s based purely on principles. The difference from Buffett’s way is that these principles are also fluid. They should be rethought all the time in conjunction with the business, industry, etc..
Let me give you an example. Being a bit better than others doesn’t translate into a moat in the coffee shop business, but it can translate to a moat in digital platforms due to the tournament effect (I’ll explain this in a bit).
This makes qualitative assessment, intellectual capacity, and decision quality of paramount importance. It can’t be taught anymore like Graham’s way; you can only try to show your thinking, and others should be inspired by it, but they should also create their own way to succeed.
That being said, I think some guiding principles are integral to success in investing in the 21st century.
Let me give a few of them as a starting point:
1️⃣ Growth is integral to value.
This is already well established in the Munger-Buffett approach.
Since net-nets have largely dried up, and we are trying to buy businesses based on their future earning power, growth is essential for success, as the market readily prices current earnings power with perfection.
Growth also supplies the catalysts for the re-rating of stocks.
If a business is trading at 15x earnings and growing earnings 10% a year, there is no reason the market shouldn’t keep the multiple steady and bump up the stock by just 10% a year as long as growth endures. However, if the growth surprises on the upside, then the price jumps substantially, as the market revises future growth expectations.
Thus, growth is an integral part of value. You have to insist on growth instead of buying something at 7x earnings and relying on mean reversion. In the former, you are looking for exceptional companies. In the latter, you may be looking at a deteriorating business, and 7x earnings today may mean 15x forward earnings.
2️⃣ Decision quality is the edge.
Markets today are very sophisticated. There is no way you can have an informational advantage.
You won’t likely find a game changer in the footnotes of the annual reports. Even when you think you have found something, the market might have already incorporated its effect together with something else that you don’t know.
You have to understand that the price of a stock is a product of a confidence-weighted average of estimates that incorporate all the available information. It’s confidence weighted because the market actors make decisions based on their estimates. Some buy, some sell short, some avoid, etc. Their position size is a proxy of their confidence.
You are unlikely to unearth new information, but what you can do is to make better decisions about what the information reflects.
Let me give you an example. This is the earnings forecast for Pagaya:
Top-line growth numbers look reasonable, but the stock today is valued at 10x 2027 earnings. This is very low for a business growing top-line by 15% and the bottom line by 53%.
So, why is it the case? If you know the business, you understand that the discount reflects the balance sheet risk. Now, if you understand the business well and the trends of markdowns, you can think that the markdowns will normalize, and nothing catastrophic will happen.
If this holds, the market will have to bump up the price by 2027 as 10x earnings for a health company growing earnings 50% would be nonsensical.
This is why SoFi was so undervalued in 2023.
Growth estimates held as analysts had projected $2.6 billion of revenue in 2024, and SoFi delivered $2.64 billion. The quintupling of the price couldn’t be due to this small beat. What changed was that the market was concerned about the potential spike in delinquencies. This was why the stock was depressed back in 2023.
If you understood the business well back in 2023, you could reason that it was unlikely for delinquencies to eat away at its earnings as it had enough reserves and its borrowers were actually like prime.
This is how I invested in SoFi and got returns, not because I unearthed some information or because my growth estimates were way higher than the market.
In short, don’t seek information edge in today’s markets. What you can do is to better understand the business and make better decisions about it.
3️⃣ Marginal advantages are the new moat.
In the Buffett-Munger style of investing, a definitive moat is a must for a company to be successful in the long term. That model doesn’t accept small advantages. A bit better product, isn’t a moat.
That is valid for businesses restrained by physical boundaries. Again, imagine a coffee shop that is a bit better than the others in the neighbourhood. Due to its limited physical capacity, the tournament effect will be limited.
What the hell is this? Think of it like a tournament between coffee shops. The winner of the tournament is clear, so people want to spend their $5 in the best coffee shop in town. However, due to its limited capacity, some will go to other coffee shops. The tournament effect will be limited.
However, when there are no physical boundaries, the tournament effect is unlimited.
Think about concerts. There are physical boundaries, but they are offset by repetition. As a result, everybody waits to spend their $100 on top performers, instead of going to concerts by lesser-known artists, the tournament effect is elevated:
In the digital economy of today, there is no limit on the tournament effect.
Everybody can pick the slightly better product. As a result, what was once a marginal advantage can translate into a durable moat. This is key to understanding the giants of today’s economy.
Think about RobinHood. It was a bit better than other online brokerages, but it has won the market because everybody could pick it; there were no physical constraints, so there was no limit on the tournament effect.
You’ll be set for a massive outperformance if you leave behind the definitive moats of the physical economy and look for digital businesses that are marginally better than the competitors and have considerable stickiness. They can win the whole market due to infinite tournament effects.
4️⃣ You have to stick with the base rates.
When you understand the importance of growth, you are naturally inclined to find things that are poised for fast growth in the future. That’s understandable.
However, one of the most critical mistakes I see is people being too generous with their forward estimates. They look at a fast business that grew 35% a year in the last 5 years, and say it can grow 30% a year for the next decade.
This could be true for that company, but that mode of operation completely ignores the base rates. It’s a recipe for disaster.
Look at some base rates for growth:
If you are looking at a technology company, remember that the average annual 10-year growth rate for the sector is 7.3%. Think about this before you assume a 20% average annual growth for the next decade, and maybe revise it to 15%.
If the issue is still undervalued after you readjusted your assumptions according to the base rates, and you are sure of the business’s above-average quality, then you are likely looking at a nice investment opportunity.
5️⃣ Use the barbell approach to risk management.
As I explained above, value investing has become increasingly forward-looking over time. Buffett operated in a more forward-looking manner than Graham; we are operating in a more forward-looking manner than Buffett.
When this is combined with the elevated importance of finding the winners due to tournament effects, we can easily find ourselves in an increasingly speculative environment, exposed to undue risks.
To control this risk, I suggest using a barbell approach. You have to put a higher percentage of your portfolio into low-risk, Buffettesque stocks bought at attractive prices, like American Express, and a smaller percentage into higher potential stocks that have a shot to become a big winner due to tournament effects.
This way, the low-risk side will limit your downside while the high-potential part will maximize your upside, as you’ll benefit from convexity there. This means that you have a higher potential of loss in each stock that side, but your loss will be limited on each position, while potential gains are unlimited. You’ll own the optionality.
(Be careful, I am not saying speculative stocks. Higher potential companies with real earnings and market power, like SoFi in 2023, Lemonade earlier this year, Oscar, etc, not complete speculations like OKLO.)
These concepts are not definitive. There can be an unlimited number of them. I just wanted to give the ones that I always keep in mind to provide some sort of inspiration and explain what type of thinking we are looking for to succeed in investing today.
🏁 Final Words
Value investing has evolved from rule rule-based approach to a principles-based one.
Today, we have internalized that the most important thing is the company's earning power, and we are trying to find companies that are trading attractively today relative to their earning power in the future. When this is the case, the game turns from picking the stocks to picking the businesses.
The market presents us with the information in a sophisticated manner, and market actors reflect that on the stock prices by making decisions based on the information. The stock price you see is a confidence-weighted consensus of the market actors.
In this environment, you can’t count on informational edge or a technical one, as the market is driven by algorithms and complex software. What you can do is to understand where the market downplays the potential for this or that reason.
It may be because of a lack of confidence in execution, as it was with Nebius earlier this year; it may be due to fears around the balance sheet, as it was with SoFi back in 2023; it may be due to concerns about disruption, as it was with Google this year, etc.
When you make these decisions, everything you know about the world and the business gets into the equation. Lateral thinking becomes the king.
For instance, I am not afraid of investing in digital companies, like Robinhood, with marginal advantages, because I know the tournament effect could be unlimited in the digital world. But this is one of my models, this is one of the things I know because I am me. I can’t expect somebody else to see this the same way I do.
This is why you have to develop your own thinking, develop mental models, and understand things in your own unique way.
This is how you make better decisions, looking at the same thing everybody does.
There is no escape from this. Value investing today is no rules wilderness. This is why it’s harder than ever, but this is also why there are more opportunities than ever.
I am sorry if you expected some clear-cut rules, but I would never deceive you into believing something I don’t believe.
It’s hard, no rules exist, and almost everything is circumstantial. We have to evolve and adapt. There is no other way.
Really enjoyed the SoFi example from 2023, the balance sheet risk vs growth potental dynamic is textbook. One nuance worth flagging: lateral thinking is crucial but the barbell approach can get tricky when you're sizing the speculative vs stable split, especially when a tournamnet-effect candidate suddenly runs into execution stumbles before network effects lock in.