Skip to content
All library documents

Why P/E Screens Can Return Fewer Stocks Than Expected

Article Quant Q&A · Author: Andrew
Read the original

Summary

The document examines why selecting the top decile of S&P 500 stocks by price-to-earnings ratio can produce fewer names than an expected count. One proposed explanation is that some companies had negative earnings during the measurement period. Since the earnings value is the denominator of P/E, data vendors commonly leave the ratio undefined when that denominator is negative, so those companies may be absent from a direct P/E field.

The suggested diagnostic is to examine price and earnings separately and calculate the ratio independently. This can help distinguish missing vendor ratios from other reasons for a screen’s count, such as ties, data coverage, index membership, or percentile methodology. The note does not establish which explanation caused the specific result, and a negative-earnings company’s calculated P/E should not be treated as an ordinary positive valuation multiple. Researchers should inspect the vendor field definition and selection rules before interpreting percentile portfolios.

Key ideas

  • A direct P/E field may be undefined for companies with negative earnings.
  • Stocks with undefined ratios can be excluded from percentile screens, reducing the returned count.
  • Calculating from price and earnings separately can help identify this data-handling issue.
  • Other causes, including screen rules and ties, may also explain a count difference.

Tags

EquitiesFactor investingStatistics
Full text
# S&P 500 P/E percentile


# S&P 500 P/E percentile












I am researching the past five year return for the securities in the top and bottom 10 percentile of the S&P 500 on date 5 years ago. I used Bloomberg to get this data. When I searched for the top 10 percent of stocks based on P/E ratio in the S&P 500 5 years ago, my searched returned 48 stocks. Why was it 48 instead of 50?

## Answer by Viquar (score 1)

https://quant.stackexchange.com/a/4212

This is possible for many reasons; one of which was explained by John. Another reason could be what you used for your P/E multiple. If it was a direct field providing you the ratio, then there is a possibility that a couple of those stocks had a negative EPS that period. Bloomberg (and many similar data vendors) do not calculate a ratio if the denominator is negative. A workaround to this is to have the P/E calculated yourself by looking at Price and EPS individually.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.