Thread by @choffstein

1/ I see a lot of factor dashboards that look at long/short results (e.g. top decile minus bottom decile), but almost no advisor I know actually invests like that.

Rather, it’s long-only implementations benchmarked against the S&P 500.

2/ So I decided to build myself a little dashboard where I take an equal-weight portfolio of investable ETFs representing each factor and construct long/short returns vs the S&P 500.

Just to see what advisors are seeing versus more "pure" implementations.

3/ It's interesting to see the difference in results here vs the traditional approach.

e.g. if we look at counterpointmutualfunds.com/factor-scorebo… (h/t @michaelbkrause), we see something slightly different.

4/ I think it is important to draw this distinction for long-only investors, as factor performance can be heavily influenced by the short side of the factor.

5/ Since long-only implementations can only "short" through implicit underweights, they tend to overly emphasize the top quartile-to-decile.

This can lead to results that do not necessarily match academic factor definitions.

6/ But the real takeaway here is that no matter how you slice it, value sucks.