Stay Away From Negative Earnings

is it better to invest in stocks with zero or negative recent net income? The answer is no. Consider the monthly data from 1954 ( $N = 838$ months). Compute total returns of negative earnings stocks and subtract risk-free returns (measured by 3-month Treasury rates), this is the equity premium $P(t)$ . Then do the same for the benchmark: top large $30\%$ stocks, get $P_0(t)$ .

Simple linear regression with Capital Asset Pricing Model gives us $P(t) = -0.0022 + 1.36P_0(t) + W(t)$ , where $W(t)$ are residuals. Statistical tests show that W are close to independent identically distributed, but not quite Gaussian. The $R^2 = 63.7\%$ . Here, $\alpha = -0.22\%$ monthly (so $-0.22\%\cdot12 = -2.64\%$ annually), thus excess return is negative. But $\beta = 1.36 > 1$ , so market exposure is much greater than 1! More risk, less reward!

A corroboration of value investing. This portfolio is capitalization-weighted: Each stock is weighted according to its market weight. This means we invest in a slice of the total market. Most index funds are capitalization-weighted. Rate data is from Federal Reserve Economic Data (FRED) web site, and stock data is from Kenneth French’s Data Library at Dartmouth College.