Benjamin Graham: the man who wrote the rules lost 70% breaking them
Almost every trading account on the internet has quoted him. Margin of safety. Mr. Market. Price is what you pay, value is what you get. What the quote threads leave out is where the man got those ideas: from four consecutive losing years in which he lost roughly seven dollars out of every ten he managed — because he had borrowed to buy things he was sure were cheap.
KEY TAKEAWAYS
- Four losing years in a row: −20%, −50.5%, −16%, −3%. Graham remembered it as a 70% loss; the figures compound to −67.7% — our arithmetic.
- His unhedged stock was 1.8× his capital. A 38.9% fall in it accounts for the whole 70% — the Dow fell far more than that.
- At 2.57× and above, the same portfolio in the same market takes the entire account. That is a setting any crypto exchange offers in one tap.
- Getting back from 32.3 cents needed +209.9%. That asymmetry, not the crash, is the reason he wrote the books.
Who was Benjamin Graham?
Born in London in 1894, raised in New York, and poor by nine — his father died, the family business went with him, and his mother lost what was left on margin in the panic of 1907. Graham finished near the top of his class at Columbia, turned down teaching offers in three different departments, and went to Wall Street as a runner in 1914. By 1926 he was running the Benjamin Graham Joint Account, and over its first three years it compounded at about 25.7% a year against roughly 20.2% for the Dow. He was beating the market on the way up, and he had begun to borrow to do it.
The rest is the part people quote: Security Analysis with David Dodd in 1934, The Intelligent Investor in 1949, a teaching post at Columbia from 1928 that produced Warren Buffett, Walter Schloss and Irving Kahn. He died on 21 September 1976. What earns him a page on a site about risk is not the books. It is the four years that came before them.
How much did he actually lose?
Graham wrote the numbers down himself. The loss for 1929 was 20%; for 1930, 50.5%; for 1931, 16%; for 1932, 3% — which he called, without irony, a comparative triumph. Set against the Dow’s roughly −17%, −34%, −53% and −23% over the same four years, his stock selection was better than the market in every single year, including the worst one.
What exactly went wrong?
Not the picking. The financing. The clearest account of the position comes from James Grant’s address to the Graham and Dodd Breakfast at Columbia in October 2008, reading from Graham’s memoirs: $2.5m of capital, with $2.5m of long positions hedged by $2.5m of shorts — a sound, market-neutral book — and then, on top of it, as much as $4.5m of unhedged long stock against which he had borrowed $2m.
We were convinced that all of our long securities were intrinsically worth more than the market price. Although many of our issues were little known to active Wall Street hands, similar ones had previously shown a praiseworthy tendency to come to life in a decent interval after we bought them and give us a chance to sell out at a nice profit.Benjamin Graham, memoirs — quoted by James Grant, Graham and Dodd Breakfast, Columbia Business School, October 2008.
Read it twice. Every clause is a reason to be patient, and patience is exactly what borrowed money does not let you buy.
That is the whole mechanism, and it has not changed in a century. A method that had worked repeatedly produced a conviction; the conviction justified size; the size removed the ability to wait for the method to work. The 1930 figure — a 50.5% loss in one year — is what happens when the third step arrives before the first one is wrong.
How much of the loss was the borrowing?
This is the number nobody runs, so we ran it. Take the hedged book as roughly neutral and attribute the damage to the unhedged $4.5m — an approximation, stated so you can argue with it. That book was 1.8 times his $2.5m of capital. To erase 70% of capital, or $1.75m, it needed to fall $1.75m / $4.5m = 38.9%.
Thirty-nine per cent. The Dow fell about 89% peak to trough over the same stretch. Graham’s holdings did far better than the market and the account still lost seven dollars in ten, because every 1% move in that book moved his capital 1.8%.
So here is the uncomfortable version of the story. At 1.8× Graham survived, wrote two of the most influential books in finance, and taught the people who built modern value investing. At 2.6× — a setting that is one tap away on any crypto exchange, and far below the 20×, 50× and 100× on the menu — the same four years take the account to zero. No Security Analysis. No Intelligent Investor. No Buffett.
Leverage is usually sold as a multiplier of returns. It is more useful to think of it as a multiplier of time pressure: it decides how long you are allowed to be wrong before the decision is taken out of your hands. Graham bought himself four years. At 10× he would have had about a 10% move.
PRACTICE CORNER
Graham’s ruin came from a number he could have written on a napkin: borrowed exposure divided by capital. Yours is on the order screen right now, and most beginner accounts are set to a default nobody chose deliberately. Open your exchange, find the leverage selector on the futures screen and the margin mode beside it, and write both down — then read what you are really borrowing with those two numbers in front of you.
These are affiliate links: we earn a commission if you sign up, it costs you nothing, and it does not change what we write. See our disclosure and the full exchange comparison.
What did the recovery cost?
Getting back is not the mirror image of falling, and this is the single most useful piece of arithmetic on the page. From 32.3 cents you do not need +67.7% to be whole. You need 1 / 0.3227 − 1 = +209.9%. Graham had to roughly triple what was left just to arrive back at the starting line.
He got there, and quickly by the standards of the period — his biographers put the Joint Account back at its 1929 level by 1935, six years after the peak, though we could not trace that date to Graham’s own figures and flag it as reported. Compare it with the alternative: the Dow did not close above its September 1929 peak until November 1954. Twenty-five years.
Two things bought him those years, and neither is available to most people reading this. He had a salary — he had begun teaching at Columbia in 1928 and lived on it while the fund recovered. And nobody could force him to sell at the bottom. On a crypto perpetual, the position closes itself at the liquidation price whether or not you were eventually right.
Did he ever break his own rule again?
Once, and it made him rich. In 1948 Graham-Newman put about $712,000 — roughly a quarter of the fund — into half of a small car insurer called GEICO, in flat violation of the diversification limits Graham himself taught. He was honest about it in the postscript to the 1973 edition of The Intelligent Investor: the profits from that single holding eventually exceeded the sum of all the other gains the partnership made over the following twenty years, from hundreds of decisions taken strictly by the rules.
Secondary accounts put the stake’s value at over $400m twenty-five years later. If that round number is right it is a 562× return, about 28.8% a year for 25 years by our calculation — but it is a secondary figure and we flag it as one.
The wrong lesson to draw is concentrate. Graham did not draw it. His own conclusion in the postscript was closer to the opposite: chances like that are rare, mostly unrecognised at the time, and no basis for a method. The diversified, unlevered process is what kept him in business long enough for one such chance to show up.
What did he say at the very end?
In his last published interview, months before his death, Graham went further than most of his followers have ever been willing to:
I am no longer an advocate of elaborate techniques of security analysis in order to find superior value opportunities.Benjamin Graham — “A Conversation with Benjamin Graham”, Financial Analysts Journal, vol. 32 no. 5, September–October 1976.
The man who wrote the 700-page textbook on security analysis, in the year he died, saying the elaborate version was no longer worth its cost. The risk discipline he never withdrew.
That is the shape of his whole career, and the reason he belongs on this site rather than only in a value-investing one. He revised the strategy repeatedly across fifty years. He never revised the part about not being destroyed.
Where this lesson stops working
His holdings had a floor; tokens do not. When Graham said his securities were “intrinsically worth more”, he meant a balance sheet you could read — by 1932 he was finding companies trading below the cash in the till. A 68% fall in that kind of portfolio was followed by a recovery. A 68% fall in a token is frequently followed by another one, because there is no number underneath it doing any work.
Nobody liquidated him. Graham kept discretion over when to sell. A perpetual position does not offer that, so “he was right eventually” is not a plan you can copy — see avoiding the big loss (Stage 11) for what replaces it.
Low leverage is not the same as safe. Read the first bar again: unlevered, the same fall still costs 38.9% and needs +63.7% back. Leverage decides whether a bad stretch is survivable, not whether it happens. Position sizing is what governs the size of the hit.
He had an income that was not trading. The Columbia salary is not a footnote; it is the reason he could wait. Someone living on the account has a shorter fuse than Graham had, at the same leverage.
Common mistakes when learning from Graham
Quoting the margin of safety and skipping the tuition. The phrase is chapter 20 of The Intelligent Investor; the reason for it is 1929 to 1932. A quote posted without the cost attached teaches nothing. Reading GEICO as permission to concentrate. One rule-break in a fifty-year career, acknowledged as luck by the man who made it, is not a strategy. Hearing “value” and thinking “floor”. Cheap things fall, and Graham’s own account is the proof. Treating leverage as a return multiplier. It multiplies how fast you run out of time, which is why risk comes before strategy (Stage 8) and not after it.
FAQ
How much did Benjamin Graham lose in the Great Crash?
In his memoirs Graham puts the damage at 70% of the $2.5m his Joint Account held in January 1929, built from yearly losses of 20% in 1929, 50.5% in 1930, 16% in 1931 and 3% in 1932. Those four figures actually compound to a 67.7% loss, leaving 32.3 cents on the dollar rather than 30. Either way he beat the Dow, which fell about 80% over the same four years.
Did Benjamin Graham use leverage?
Yes, and he said so. By James Grant's reading of the memoirs, Graham held $2.5m of long positions hedged by $2.5m of shorts, plus a further $4.5m of unhedged long stock against which he had borrowed $2m. That unhedged book was 1.8 times his capital. He swore off borrowing afterwards and Security Analysis, written in the years that followed, treats margin as a risk to be avoided rather than a tool.
Did Graham ever recover from the 1929 to 1932 losses?
Financially, yes. His biographers put the Joint Account back at its 1929 level by 1935, and Graham-Newman went on to compound at roughly 20% a year from 1936 to 1956. We could not trace the 1935 date to Graham's own figures, so treat it as reported rather than documented. For comparison, the Dow did not close above its 1929 peak until November 1954.
What did Graham mean by margin of safety?
In chapter 20 of The Intelligent Investor he writes that if forced to distil sound investment into three words, the motto would be MARGIN OF SAFETY: buy only at a wide enough discount to your estimate of value that being somewhat wrong is survivable. He arrived at it after being wrong with borrowed money, which is why the idea is about surviving error rather than about finding bargains.