Using quarterly rather than TTM figures

In many of the ranking systems used most frequently by P123 users appears the factor NPMgn%Q, higher better, compared to universe. This seems to work very well as a part of an integrated ranking system. In fact, it appears to work better than NPMgn%TTM. I’ve also seen in P123 ranking systems the formula FTFQ/AstTotQ, which works pretty well. And I have found that approximating PE based on an average of this quarter’s and last quarter’s EPS works better than using TTM figures.

What I don’t understand is the reason here. Net profit margins, free cash flow, and EPS vary wildly from one quarter to another, and I can’t see how anything derived from quarterly figures can have any true relevance to anything. Perhaps these factors are working simply because the quarterly figures are the ones the big institutions rely on when deciding to begin investing in a microcap? I have no clue. Anyone have any insights into this?

Some companies will have seasonality while others not, if you can separate the two, it would make sense that it works on non-seasonal and not so much on seasonal. Personally, I think many users are using it because of the reason you state “it works better” or better said, it has worked better, it does not mean it will work better, it does not mean it will work worse going forward. It looks like it is curvefitting.

Might there be a behavioral explanation for why it works better? Some combination of anchoring and recency bias?

Here you go: http://faculty.chicagobooth.edu/workshops/finance/pdf/hartzmarkseasonality.pdf

Well, if some of the other factors in the ranking system deal with asset turnover and debt leverage, then you could build a Dupont equation-like model. Maybe the ranking system author stumbled upon it from trial and error. It is always hard to reverse engineer the initial strategy (if there was one).

What people fail to understand is that it doesn’t matter whether it makes sense or not. The market is driven by group psychology, and the group on average is not a sophisticated investor. So if a stock has a great quarter, investors will buy regardless of whether it is a seasonal stock or not. This can be determined by empirical tests, no different than value/momentum.

BTW, I think you will find that using quarterly factors works better when ranking against same industry.

Cheers,
Steve

Thank you Matthew for the great article and Steve for the behavioral wisdom.