WHICH FACTORS GET PICKED?
Our proposed remedy for failed factors is to use a three-factor portfolio each month that is derived using the optimization process described in the previous section. Exhibit 15.6 gives a visualization of the history of which factors get picked for the 29-year period from 1982 through 2010. Factors that can be picked are labeled on the right and left vertical axes, with time along the horizontal axis. For each date on the horizontal axis, there are three points on the chart signifying what was selected for that date. The length of the line segments in the exhibit indicates the length of time that factors remained in the model. The bottom of the exhibit shows, for example, that up-to-down revisions was in the model from 1982 until 1996 but has not been in the model since then. P/E was in the model for a few years in the 1990s, where earnings was analyst forecast; see the label PREDICTED_EP in the middle of the exhibit. P/E where earnings represents historical earnings has not yet been in the model.
Exhibit 15.6 shows, in broad terms, that some structural changes occurred in the persistence of factors in the model. The horizontal line segments on the right side of the exhibit are somewhat shorter and more dispersed among the pool of contenders than those in the middle of the exhibit. This pattern suggests that in recent years factors need to be changed more frequently for the process of alpha repair to be effective. That does not necessarily mean that the ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access