In 1981, a 24-year-old, Joel Greenblatt authored an essay in the Journal of Portfolio Management titled ‘How the small investor can beat the market’ alongside Richard Pzena and Bruce Newberg. Together they shared outlined their process for buying stocks selling below their liquidation value. Inspired by Benjamin Graham, Greenblatt & Co fathomed what they might find if they refined Graham’s ‘net-net’ universe by removing the crud; a combination of Graham’s approximation of liquidation value and the PE ratio. The results, calculated over a six-year period, are impressive, and I have some thoughts. The essay is notoriously difficult to attain (seriously, I tried), but I eventually found one through a connection. If nothing else, it might provide some food for thought or serve as a summary of this well-regarded paper.
Ignore the academics
The essay starts off with a dismissal of the efficient market hypothesis.
“Past experience and common sense tell us that the individual investor has little chance of outperforming in the stock market. After all, even the billion dollar institutional funds, with their greater access to timely information and comprehensive analysis, have been notably unsuccessful in their efforts to achieve superior long-term investment results”.
Ironically, the expertise of Wall Street portfolio managers has its own part to play in their disappointing performance. According to the academia of the time, competition between these professionals ensures that stocks remain efficiently priced at all times (an idea which still persists in 2023). If all publicly available information is absorbed into equity valuations, then there are no bargains to be had. Attempting to beat the market is, therefore, a loser’s game.
“We should recall, however, that Wall Street research houses limit their coverage to fewer than 500 actively traded issues. Meanwhile, the NYSE trades 2,000 stocks, the Amex trades 1,000 companies, and the OTC market trades another 7,000 issues that are required to provide relatively full disclosure to the SEC”.
While the numbers may differ today, something which holds true that is institutions are still mandated to neglect certain stocks. Many of them are too small, or illiquid, to be covered by these firms. For this reason, many investors believe there will always be unrecognised value in the market. This is where Benjamin Graham comes in.
"In his textbook, Security Analysis, he outlines in little more than a page the opportunities to be found in stocks selling below their liquidation value. In studies between 1923 and 1957, Graham reported superior results when market levels enabled him to buy a diversified list of these bargain stocks”.
In an effort to discover how to find these inefficiently priced companies, the trio utilise Garham’s approximation of liquidation value.

The assumption is that all current assets less all liabilities and preferred stock approximate the liquidation value because any loss on the liquidation of inventories and receivables will be made up through the liquidation of fixed assets (land, machinery, etc). This process was intended only to be a “rough screening device” to sort out the likely prospects from the thousands available.
“Clearly, on a theoretical level, it is possible for a stock to be selling below liquidation value if the discounted cash flow of its future earnings results in such a lower valuation and if there are no immediate prospects for liquidation. Firms losing money, or those firms expected to lose money in the future might reasonably be selling at sub-liquidation prices”.
In an effort to eliminate the crud, the study excluded companies showing a loss over the trailing 12-month period. Combining liquidation value and the PE ratio, Greenblatt constructed four unique portfolios and compared their performance to that of the OTC1 and Value Line indexes across 18 four-month periods from 1972 to 1978.
A computer program would purchase any stock2 that met the above criteria and the team would sell each stock after a 100% return or after 2 years had passed; whichever came first.
The Results
An initial $100 investment across the OTC and Value Line indexes would return $88 and $79, respectively. After six years, portfolios 1 through 4 returned a range between $248 and $517.
Portfolio 4 ($517) was the highest-returning vehicle and selected only those stocks which were valued the lowest in the sample with relation to PE (≤ 5) and LV (≤ 0.85). Portfolio 3 ($363) allowed for the same PE ratio (≤ 5) and a higher LV value (≤ 1.0) and was the second-highest returning vehicle.
Portfolio 2 ($331), despite restricting the universe to a lower LV (≤ 0.85) of stocks than portfolio 3, underperformed slightly. Portfolios 1 and 2 set PE thresholds at a floating rate, equal to “twice the prevailing triple A rated yield” in the period. E.g, if triple A rated yields were 8%, the PE would be ~6.25. As it happens, yields ranged from 7% to 9% during the period, so we can assume that the PE value was considerably higher for portfolios 1 and 2.
Portfolio 1 had, collectively, the highest values for LV (≤ 1.0) and PE (5.5 to 7) and performed the worst. The study suggests that searching for profitable stocks that have an LV ≤ 0.85 and PE ≤ 5 is the most attractive fishing spot for stock pickers. This backtest appears to have yielded impressive results, but I have a few comments about the legitimacy and reliability of the study; abstract from the typical flaws of backtesting such as human error, period bias, survivorship bias, etc. Below is the return data in raw form.
Reconciling returns: Note that during the six-year period3, the paper provides annualised compound returns for each portfolio. These do not reconcile with the performance across the 18 four-month segments. What's more, the paper repeatedly refers to there being "15" four-month segments. Unless I am missing something, six years split into four-month segments equals 18, not 15. The paper claims that "the percentage gain for the entire portfolio was merely an average of the percentage gain of the individual stocks”. In the first 3-4 periods of the study, the authors claim they were “unable to find any “bargains” in the market” during this time.
Dividends & fees: The return data is understated due to the absence of dividends (averaged 3.5% per annum over the time period). In the modern day, where entry costs to investing are lower than they were in the 80s, the influence of commissions may be lower today. In this study, the authors assumed a 2.5% commission on the purchase and sale of a stock (totalling 5%) as well as a 2.5% bid/ask spread (applied to the 60% of the stocks they purchased over the counter) and a 25% capital gains tax (of which 90% of the stocks they bought were held long enough to qualify for capital gains treatment).
Sample period: In Greenblatt’s words, this timespan was particularly volatile and would make for a robust testing period under extreme market conditions. While that may be true, I am not sure if six years is sufficient enough time to be considered reliable. I personally assume a six-year stretch to be “medium term”.
Sample size: Representing just 15% of the total population of data (footnote 2), and including only those stocks beginning with letters A or B was a decision to create an unbiased sample and is understandable. While appropriate for a study, investors typically don’t limit their investable universe in accordance with the first letter of a stock’s holding company.
Portfolio size: Next to the period returns in the table you will see each data entry has a “/ number” next to it. This implies how many stocks were in the portfolio during the period. Ranging from 0 to 52, the average portfolio contained 15 stocks in it at any given time. Concentrated by some people’s standards, but then again this comes from the guy who said4 this in 2022:
“I don’t consider six to eight names of making up 80% of your portfolio, particularly concentrated. One name for your whole portfolio, that’s pretty concentrated.”
Joel Greenblatt
There is also something to be said for the market cap threshold ($3 million) employed in the study. By today’s standards that would equate to a little over $20 million. Most companies of this size tend to be extremely illiquid. Some might even be ineligible to trade as and when they crop up on Greenblatt’s screener. Excluding such outputs (with stricter screening) could improve the results and accuracy of the study.
Why Did It Work?
Greenblatt feels this selection criterion produces results because of the “large second tier of stocks” that are forgotten and inefficiently valued by the market. While there were no magic qualities associated with buying stocks trading below liquidation value, the study sought to exploit stocks that were undervalued and protected by large asset values and strong balance sheets. These companies, according to Greenblatt, are “often the object of takeover initiatives”.
“Simply stated, by limiting our investments to stocks that according to fundamental notions of stock valuation appear severely depressed, we were able to locate more than our share of these inefficiently priced, undervalued securities. In other words, there are probably many more undervalued stocks that are not selling below liquidation value”.
He concludes by questioning, with the advancement of technology, whether more retail investors will possess sophisticated enough tools in the future to conduct their own screening processes. In his opinion, so long as the structure and focus on the securities industry remains fixated on primary issues, the individual investor can benefit by searching in untapped corners of the market.
Overall, I think there are limitations to any backtest. I have a few issues with this one; the time period, the inclusion of potentially illiquid stocks, and the decision to segment the performance in the way they did. Nonetheless, it made for a unique paper on the process of screening and stands as a valid reminder that there is still, fifty years later, a similar dynamic of institutions and fund managers who are mandated to ignore certain areas of the market. I am not a bargain basement hunter myself. The idea of having to rotate into and out of stocks, low-quality ones at that, every few years exhausts me. But I do like undercovered stocks. There are many ways to skin the cat, and this proved to be an engaging study of one of those methods. I will leave on one final caveat that I believe is summed up well in Morgan Housel’s The Psychology of Money.
“Graham’s classic book, The Intelligent Investor, is more than theory. It gives practical directions like formulas investors can use to make smart investing decisions. I read Graham’s book when I was a teenager, learning about investing for the first time. The formulas presented in the book were appealing to me, because they were literally step-by-step instructions on how to get rich. Just follow the instructions. It seemed so easy. But something becomes clear when you try applying some of these formulas: few of them actually work”.
Morgan Housel, The Psychology of Money
In a chapter which handles our tendency to look to historians as prophets, Housel alludes to the fact that the formulas used to identify stocks in Graham’s classic, The Intelligent Investor, changed in every updated edition of the book. Graham was a practical man, not wedded to his prior ideas, and so he would alter his formulae when he felt they were no longer the best way to extract value from the market.
“Graham advocated purchasing stocks trading for less than their net working assets—basically cash in the bank minus all debts. This sounds great, but few stocks actually trade that cheaply anymore—other than, say, a penny stock accused of accounting fraud. One of Graham’s criteria instructs conservative investors to avoid stocks trading for more than 1.5 times book value. If you followed this rule over the last decade you would have owned almost nothing but insurance and bank stocks. There is no world where that is OK”.
Morgan Housel, The Psychology of Money
The point is, things change; markets, investing themes, accounting rules, economies, monetary policy, interest rates. They all change. We are approaching 40 years since Greenblatt published this paper. In the preceding four decades the US endured eight recessions. In the proceeding four decades they endure just four. When the paper was published, tangible assets would account for ~70% of the assets across the constituents of the S&P 500. Today, it’s less than 10%.
Investing history is a funny thing. The more years you peel back, the more likely you are to analyse a version of the world that is no longer applicable today. Sure, it's important to study the past to learn from our mistakes. But it's also important to remember that the world is constantly changing. What worked in the past may not work in the future. The further back you go, as Housel suggests, the more general your takeaways ought to be. People’s love affair with fear and greed is a persistent trend. So too are the formulations of bubbles, albeit with different catalysts. But everything else, including the use of fixed parameters from which to screen stocks, is likely to have a short shelf life.
Thanks for reading,
Conor
For context, the “OTC” was used interchangeably with the “NASDAQ” until the mid to late 80s.
To create an unbiased sample of stocks, they compiled data on all firms listed in the Standard and Poor’s stock guide with market values over $3 million beginning with the letters A and B. This represented ~15% of the stocks listed in the guide or about 750 companies.
Following the six-year period, the OTC and Value Line indexes would report a CAGR of 1.3% and -0.3%, respectively. Portfolios 1 through 4 would report CAGRs ranging from 20% to 42.2% (11.3% to 29.2% after tax and commissions). In other words, each of the portfolios more than doubled their initial stake, and then some.





