Decades' worth of psychological studies show that people are extremely good at figuring out which information they need for a decision—but do a poor job of using that evidence methodically over time. You are likely to draw divergent conclusions from identical data on different occasions, even when nothing fundamental has changed, because of variations in context, alterations in your mood, shifting demands on your attention and memory, and so forth.
No wonder John Mihaljevic, editor of the Manual of Ideas, a website for value investors, says wryly that he uses checklists to combat his tendency to make "the same type of mistake again and again."
Structuring your decisions this way, says Michael Shearn, author of the book "The Investment Checklist," forces you to take "a holistic view" of a stock or other asset. That should reduce your odds of being flummoxed by the unexpected.
"When we look to make an investment, the greed part of the brain is turned on," says Mohnish Pabrai, managing partner of Pabrai Investment Funds in Irvine, Calif., a group of private portfolios with assets of approximately $700 million. "A checklist is like a circuit breaker that helps prevent the brain from being able to flip that switch."
To build his list, Mr. Pabrai studied his mistakes and those of great investors like Warren Buffett. Anyone "can build a customized checklist based on your own history of your own failures," he says. Mr. Pabrai advises investors to review their past decisions that lost money.
"Rub your nose in your own failures," he urges. "Avoiding the mistakes you've made in the past will take your error rate way down in the future."
Mr. Pabrai says he believes that the flubs made by great investors fall into five groups: valuation, or how cheap an investment is; leverage, or risks associated with borrowing; management and ownership; "moats," or how well-fortified business are against competition; and personal biases.
First he does all his other research; then he works through the checklist to make sure he didn't miss anything.