My blogging ground to a halt the last few weeks because I was completing a paper “Measuring Portfolio Valuation. “ Will put link to paper here shortly.
My new paper looks at two issues. The first issue involves correct and incorrect ways to measure the PE ratio of a portfolio of stocks. The second issue involves the correct way to conduct statistical tests on valuation measures for groups of stocks.
The paper starts with a discussion of the limitations of the PE ratio, the most commonly used valuation measure for common stocks. The PE ratio is undefined when earnings are negative and unstable when earnings are small. By contrast, the ratio of the difference between market cap and earnings to market cap (denoted (MC-E)/MC) has a clear economic meaning when earnings are negative and is not an outlier when earnings are low. In addition, there is a one-to-one relationship between this ratio and the PE ratio.
Many investment firms use a weighted average of firm PE ratios to measure the PE ratio of their ETFs or mutual funds. The firms often discard observations from firms with negative earnings and cap the PE ratio of firms with high PE ratios. These methods are arbitrary and often tend to understate the valuation of stock prices relative to earnings.
The ratio of the sum of market caps of firms in a portfolio to the sum of earnings of firms in the portfolio is the correct way to measure the PE of a portfolio. This measure of PE can include all firms even firms with negative earnings. Moreover, small changes in earnings for firm with high PE ratio do not have a large impact on the overall portfolio PE ratio.
A second way to measure the PE ratio of a portfolio, which relies on the weighted average of the statistic ((MC-E)/MC) is presented and shown to be equivalent to the ratio of the sum of market cap to sum of earnings. This result is motivated in the following blog post.
Two Ways to Calculate a Portfolio PE Ratio:
The paper contains a formal proof demonstrating the two methods of constructing a portfolio PE are identical.
Often analysts conduct hypothesis tests on portfolio financial ratios. Tests based on PE ratios often provide misleading results because of problems measuring the PE ratio when earnings are negative or small. Firms with negative earnings are routinely omitted from the sample. The standard deviation and skew of portfolio PE ratios are often large making it difficult to reject a null hypothesis.
By contrast, statistical tests based on (MC-E)/MC do not require the omission of firms with negative earnings. Moreover, the distribution of (MC-E)/MC appears normally distributed with few outliers. As a result, statistical tests using this ratio are more reliable than statistical tests using PE ratios.