Trade the Congressional Effect: How To Profit from Congress's Impact on the Stock Market

MPT Assumes All Daily Pricing is Random, but the Congressional Effect Shows it is not

MPT, like most theories, is only as good as the assumptions behind it. The Congressional Effect shows that MPT's key assumption—that daily returns are randomly distributed—is wrong. The statistical data in support of this is overwhelming. Unlike the Super Bowl indicator, which involved 45 data points, the Congressional Effect is confirmed by 11,832 daily observations over 47 years. On the 7,767 days from 1965 through 2011 when Congress was in session, the stock market went up in price on days at the annual rate of 0.72 percent, whereas on the 4,065 days when Congress was out of session it went up at the annual rate of 16.60 percent. These figures ignore transaction costs, tax effects, and dividends.

These data were further confirmed by Lamb, Ma, Pace, and Kennedy in 1997. Their study, which covered the period from 1897 through 1993, found that “almost the entire advance in the market since 1897 corresponds to periods when Congress is in recess.”⁶ They looked at a total of 26,337 trading days, comprising 9,950 in recess days and 16,387 in session days. Using the DJIA, they found the average daily return on in-session days was .0042 percent and on in-recess days 0.0541 percent. This is a difference of over 13 times, over a period representing over 90 percent of the available meaningful data on the United States stock market. It is an extraordinary magnitude for an effect to be so consistent over ...