The Psychology of Trading

In March 1940, eight months before he shot himself in a New York hotel, Jesse Livermore published the only book he wrote himself, How to Trade in Stocks. Its rules are short and clear. Buy only when a stock "acts right", crossing a pivotal point on volume. Build a position in steps, each purchase at a higher price than the last. Decide in advance how much you are willing to lose if you are wrong. "It is foolhardy to make a second trade, if your first trade shows you a loss. Never average losses. Let that thought be written indelibly upon your mind." And each time a successful deal is closed, take half the profit and lock it in a safe deposit box. They read like the manual of a modern trend follower (1940 first edition, as quoted in the companion's profile of Livermore).

The same book is also a confession. "Every time I lost patience and failed to await the Pivotal Points and fiddled around for some easy profits in the meantime, I would lose money," Livermore wrote. A whole chapter, "The Million Dollar Blunder", describes buying 20,000 bales of cotton, being stopped out, and buying them back "five times in six weeks", losing $25,000–30,000 each time and about $200,000 in all, then missing the 500-point rise that followed; the episode is undated. Six years before the book appeared, Time had reported his bankruptcy of March 1934, with liabilities of $2,259,212.48 against assets of $184,900, "mostly life insurance" (19 March 1934). By Time's count it was his fourth; Forbes and Wikipedia call it his third. Chapter 2 told his career and its lore, including the "$100 million" of 1929 that traces only to a biography of 2001. What matters here is the gap that his own book admits: the rules were written down, by the man who broke them.

Livermore's case is anecdote, but the habit it shows has been measured. In 1998 Terrance Odean examined the trading records of 10,000 accounts at a US discount broker. On days when investors sold, they realised 14.8% of the gains sitting in their portfolios but only 9.8% of the losses: they were about half as likely again to sell a winner as a loser (Journal of Finance, October 1998). Hersh Shefrin and Meir Statman had named the pattern thirteen years earlier as the disposition effect, the disposition to sell winners too early and ride losers too long (1985; companion, section 20). It is the opposite of Livermore's rules, and of every trend-following system in this book.