Factor Momentum in Pan-European Large Caps

Ehsani and Linnainmaa, and Gupta and Kelly, show that the factors themselves trend. A factor with a positive trailing year tends to have a positive next month. That is a different claim from stock momentum: the security is the factor portfolio, and the signal is the factor's own return.

Ken French's European factors are a natural dataset for this, and this site already uses Ken French for the US factor lab. Here the factors are built from the Euronext panel instead, so the timing result refers to the same Paris, Amsterdam, Brussels, and Milan names as the betting-against-beta page. Two factors are enough to see the mechanism. Momentum is the 12-1 winner minus loser. Size is below-median trailing dollar volume minus above-median.

Each month the momentum signal is the sum of monthly returns from $t-12$ through $t-2$. The long leg is the top 30 percent, the short leg the bottom 30 percent, equal-weighted, using only names with a realised return in month $t$. Size uses the three-month average of month-end dollar volume (unadjusted close times share volume). Small minus big is the equal-weight return of names at or below the median minus names above it.

The timing rule looks at each factor's own trailing twelve monthly returns. If the sum is positive, next month's weight on that factor is one. If it is negative, the weight is zero. The timed book averages the two timed factors. The untimed book always holds both. Cash months earn zero rather than the opposite factor, which is the conservative version of the rule: it does not require a short in the factor portfolio.

Dollar volume is a size proxy, not market capitalisation. On this panel it ranks liquid large caps against less liquid large caps. It is not the Fama–French SMB breakpoint.

The wealth chart shows the timed book, the always-on book, and each factor alone. The bars are annualised Sharpes of those four series. The test is the Newey–West gap between timed and untimed. A factor can have a positive Sharpe and still be a poor timing instrument if its trailing year does not line up with the next month.

Two factors are not a five-factor model. Value, profitability, and investment are absent because the study does not pull financial statements. Adding them would change both the untimed portfolio and the thing being timed.

Euro, sterling is not in this panel, and Swiss names are not Euronext listings. The result is not a pan-European factor catalogue.

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.