One-Month Reversal on Hang Seng Constituents

The US momentum studies on this site skip the most recent month on purpose. That skipped month is the reversal. Jegadeesh (1990) documented that last month's losers outperform last month's winners at the next rebalance, and the premium is large enough, before costs, that a 12-1 momentum sort is built to avoid it.

This page measures the reversal on its own, on Hong Kong listings. The panel is the Hang Seng large-cap list used by the carry study, so both Hong Kong pages share the same adjusted-close file. The book is long the bottom quintile of last month's return and short the top quintile.

Month-end adjusted closes produce a simple return $R_{i,t-1}$. Names at or below the 20th percentile are the long leg. Names at or above the 80th percentile are the short leg. Inside a leg the weight is equal, and each leg is scaled to one half so the book is dollar-neutral:

$$r_t^{\mathrm{rev}}=\frac{1}{2}{\bar{r}}_t(\mathrm{losers})-\frac{1}{2}{\bar{r}}_t(\mathrm{winners})-0.0020\Vert w_t-w_{t-1}{\Vert }_1$$

Twenty basis points is harsher than the carry study because a quintile reversal replaces most of its names every month. The Hang Seng Index is plotted beside the book. The statistical test is on the reversal mean itself.

The wealth chart has three lines: the cost-adjusted reversal book, the equal-weight of the names that had a return that month, and the Hang Seng Index. A reversal book can compound positively in a falling market because it is long-short. The annual table is the place to see whether that happened in a given year.

A one-month sort on Hong Kong names picks up a large bid-ask bounce. Part of a gross reversal is the bounce, which is why the cost is not optional in the headline number.

Quintiles of this Hang Seng panel are a handful of names. One earnings gap moves the leg. The Newey–West lag of three months is there to keep a single cluster of gaps from looking like a precise mean.

Yahoo Finance adjusted closes are not a point-in-time index membership file. The constituent lists are a liquid large-cap panel, fixed in the repository, not the official index as it stood on each rebalance date.

The risk-free rate is set to zero in every Sharpe ratio. A local bill rate would lower the ratios and would not, by itself, change a long-short spread.

Costs are a flat one-way haircut on weight turnover. They do not include borrow fees, lot-size rules, auction imbalance, stamp duty, or the Shanghai and Hong Kong short-sale bans that bind a live short book.

This is a research record. It is not a recommendation to trade the sleeve.