Equal-weight rank-rotation tests for a disclosed value book

Read the book as a holding map

A long public institutional series from 2001 through 2016 can be rebuilt into a one-year-hold evaluation of a concentrated value-investing book. Value-investing, in this archive sense, means reading that book as a sector-and-name allocation map rather than as a short-term timing signal.

Those reports become public about 45 days after the report date. Because of that thirteen-f-lag, the evaluation delayed entry by two months and did not assume fills on the filing day. After the delay, the reconstructed names or sector baskets were kept for a one-year-hold and then rescored.

Score the same book two ways

Equal-weighting scores the disclosed book by averaging each holding’s one-year return as if every name were the same unit of analysis. Capital size does not change that count.

A capital-weighted-average scores the same book by multiplying each holding’s one-year return by its portfolio weight and dividing by the sum of those weights. Large sleeves then dominate the measured path.

Both scores use the same holdings and the same one-year-hold. One reading counts names as equal units. The other reading follows where the capital sat.

Park each sleeve beside its sector basket

Over that 15-year window the reconstructed book sat mainly in three sectors, with average weights near 38% finance, 36% consumer staples, and 6% information technology.

Sector context used nine long-running sector ETFs and omitted the then-new real-estate sleeve. Each stock group could then be compared with its matching industry basket rather than with a single market index.

The study’s premise was sector-first value-investing: locate where capital sat, then ask whether individual names or the matching sector ETFs carried the measured yearly path. A sector-sleeve is the portion of the book mapped to that basket, so a single name can be judged against its industry tide.

Rotate the cheapest and the lightest names

Rank-rotation rebuilds the book from a short ranked slice, either the lowest entry price or the lowest position weight, and compares that slice with the full-name set over the same holding period.

One test took the five and fifteen lowest-priced names, equal-weighted them for one year, and asked whether entry-price rank changed the path versus the full-name book.

A parallel rank-rotation of the five and fifteen lowest-weighted names asked whether light position size, not only cheap price, produced a different equal-weighted path than the full book.