Price to Current Fiscal Year Earnings Backtest
Another widely used valuation ratio that Richard Tortoriello examined in his book Quantitative Strategies for Achieving Alpha is the P/E ratio. It is probably the most recognizable and used stock valuation metric. There are several variations on this popular fundamental ratio, including using trailing twelve-month earnings per share, one-year forward earnings per share estimate, and the use of enterprise value instead of market price. Tortoriello decided to present the backtest result for the P/E ratio that uses current fiscal year earnings, since he found it to have the most consistent performance.
For ease of backtesting, I used the inverse of the price to current fiscal year earnings, which is current fiscal year earnings per share divided by price. I used StockScreen123 to conduct a 10-year backtest of the current fiscal year earnings to price ratio. I filtered out ADRs, non-US companies, companies in the miscellaneous financial services industry category (to mainly filter out closed-end funds), stocks trading below $2, market caps less than $433 million (approximately matching the average cut-off Tortoriello used), and companies that did not have a current fiscal year earnings to price ratio due to missing data. The results are as follows:
Current Fiscal Year Earnings to Price Ratio : Rolling 3-Yr Periods Excess Returns vs. Universe
The 2001 to 2011 backtest results for the 1st quintile are very similar to what Tortoriello found. Tortoriello estimated that the average annual excess return versus the universe from 1988-2007 was 4.7%, while I found the average excess return versus the universe was 4.13% from 2001 to 2011. This excess return is less than the 4.80% I found for return on enterprise value (ROEV) and 4.59% for enterprise value to EBITDA ratio for the 1st quintile. The Sharpe ratio for the top quintile of the 2001-2011 price to current fiscal year earnings estimate backtest was also lower than for the other previously mentioned backtests.
The 5th quintile average excess returns versus the universe were much lower than what Tortoriello found, -5.2% versus -2.2%. This is probably due to the different time periods. My backtest included the major market decline in 2008 and 2009. However, when comparing different valuation ratios over the same time period, the bottom quintile of enterprise value to EBITDA had average excess returns of -5.84% and ROEV had average excess returns of -5.29%.
In general, this backtest of price to current fiscal year earnings indicates that in the past ten years both return on enterprise value and enterprise value to EBITDA ratio models outperformed the P/E ratio using current fiscal year earnings for both the top and bottom 20% of stocks. Given these results, one has to wonder why the P/E ratio is still so popular. Do you use the P/E ratio? If so, why? Please leave you responses in the comments section below.