Why Longing a Strong Factor Group Does Not Guarantee Short Profit
Summary
The document explains why positive returns from buying stocks with the strongest exposure to a factor do not imply that shorting stocks with the weakest exposure will also be profitable. The bottom group may still rise, only by less than the top group. A short position in that group would then lose money even as the long group performs well.
A long-short factor portfolio aims to capture the return spread between the strongest and weakest exposure groups. Whether the short leg earns a positive standalone return depends on the factor and market behavior, so it should be evaluated separately rather than inferred from the long leg. The response also notes that the broad upward tendency of stocks can make shorting less profitable. The exchange provides reasoning rather than empirical comparisons or a particular testing procedure, and its comments do not show that any specific factor will work.
Key ideas
- A profitable long position in the strongest factor group does not guarantee that the weakest group will fall.
- The weakest group may rise by less than the strongest group, causing a loss on a standalone short.
- A long-short factor strategy seeks to earn the return spread between the groups.
- Short-leg profitability varies by factor and should be tested directly.
- The general upward tendency of stocks can hinder short strategies.
Tags
Full text
# Factor performance in different groups # Factor performance in different groups While we are testing performances of factors, it is common for us to put stocks into different groups based on their exposure to certain factor, such as size, value, momentum,etc., Then we can test the performance of the trading strategy that buy/long the group of stocks with the largest exposure, or the strategy that short the group of stocks with the least amount of exposure. For example, we can have group 1 to group 5, group 1 have the largest exposure, and group 5 have the least amount of exposure. When we generate positive returns from buying group 1, does it necessarily means that we will also have positive returns by shorting group 5? I think for some factors this might be true, but it will not work for all factors. So, when a factor is effective in the case of buying/longing the group with the largest exposure, does it means that shorting the group with the least exposure can also make money? Thank you! ## Answer by KaiSqDist (score 1) https://quant.stackexchange.com/a/80943 > When we generate positive returns from buying group 1, does it necessarily means that we will also have positive returns by shorting group 5? No. This is because, it is (almost always) the case where group 1 performs better than group 5, which means that when group 1 has a large positive return, group 5 can either have a smaller positive or negative return. The former which contradicts your case as a smaller positive return results in a loss in your short. The whole idea behind a long (short) top (bottom) deciles is just to earn the spread in returns between the two groups, which reflects the factor returns. ## Answer by Brian from QuantRocket (score 0) https://quant.stackexchange.com/a/80932 There's no general answer to this question; you have to test each factor. Sometimes shorting group 5 will be profitable, but often it won't be, due to the long-term upward tendency of stocks. The best library in Python for analyzing factors is Alphalens. You can see example analyses of a variety of different factors on QuantRocket's blog.
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.