You are considering a fund that buys recent stock-market winners. A rising market sounds like good news for it. But what if yesterday's losers recover fastest?
In the March–May 2009 rebound, the past-winner basket in Daniel and Moskowitz's research gained 8%. The past-loser basket gained 163%.
Both rose. Yet a strategy betting on winners beating losers suffered. Before choosing a momentum fund, find out whether it only buys winners or also bets against losers. That choice changes what a rebound does to your money.
A rule for comparing past returns
GARP and quality looked for business evidence. Momentum starts with the market's return record.
Cross-sectional momentum ranks stocks by their past returns and favors the stronger group. Here, the comparison uses total returns, including reinvested dividends. The period measured is the momentum formation window, also called the lookback.
One research calendar comes from Ken French's data library. To choose stocks for January, rank the previous January–November returns and skip December. That gap helps keep very short-term price reversals out of the signal.
Rank is relative. Suppose one stock lost 10% and another lost 30%. The first ranks higher even though neither made money. A winner can be the stock that fell least.
The window and review dates are set in advance. Buying whatever went viral this morning has no such discipline.
What the evidence shows
Much of the research measures a winner-minus-loser portfolio: it owns past winners and takes short positions in past losers, bets that lose when those stocks rise. A long-only momentum fund owns stocks without those bets. It captures only one side of the comparison.
Jegadeesh and Titman's 1993 study examined US stocks from 1965–1989. Buying past winners and selling past losers produced positive average returns over holding periods of 3–12 months. The formation window looks backward to choose stocks; the holding period looks forward to measure what happens next.
One proposed explanation is underreaction: investors absorb news slowly, so prices keep adjusting. Later, return chasing can push the move further. Slow reactions and eager followers could both feed the same trend; the historical pattern alone does not prove its cause.
A process that can be repeated
A repeatable strategy fixes three choices before trading:
- Which stocks qualify. Set the market and rules for how easily the shares must trade.
- How to rank them. Use comparable total returns over the same window for every stock, using only data known at the decision date.
- What to own and when to review. Form a diversified basket with stated weights, then apply the selection rules and rebalance on a fixed schedule.
The rule can replace a stock that is still rising if other stocks rise faster. Replacing holdings is turnover, and it creates trading costs and record-keeping work.
Funds can add risk limits and buffers that keep a stock through small ranking changes, reducing unnecessary trades. There is no single momentum recipe. Two funds can rank different stocks as winners because they use different windows or adjust differently for risk.
Why rebounds can hurt
Daniel and Moskowitz's Momentum Crashes builds baskets from US common shares in the CRSP database, covering NYSE, Amex and Nasdaq. Each decile is one tenth of the ranked stocks. The researchers use the lookback calendar described earlier and give larger companies more weight.
They rebuild the baskets at every month-end. The bars show the gains across March–May 2009 as those holdings change: the former losers leave the winners far behind.
At those reported returns, separate $100 stakes become $108 in past winners and $263 in past losers. The return gap is 163 − 8 = 155 percentage points.
After a deep market decline, distressed former losers can rebound sharply. Bets against them become especially painful. A momentum crash is a severe loss for the strategy as the old ranking reverses. The market's recovery can be the strategy's bad news.
Owning only winners avoids losses on those short positions, but the fund still faces market risk and can lose money or lag the market.
What remains after trading
Zero commissions still leave spreads and other execution costs. Use a $200 paper portfolio just for the arithmetic: suppose one rebalance replaces $160 of holdings, and trading costs 0.25% of every dollar bought or sold.
That is $0.80 ÷ $200 = 0.4% of starting capital for this one rebalance. Replacing the same fraction of a $10,000 portfolio costs $40 at that rate. Replacing $160 means trading $320.
For a US resident's taxable account, a gain or loss on an asset held one year or less is generally short-term; net short-term gains are taxed as ordinary income.
Running the strategy yourself means keeping records and following the review schedule through reversals. If that routine is a poor fit, a fund can handle the trading. It delegates the work; you keep the risk.
Before choosing that fund, look for its return window, review schedule, cost treatment and approach to reversals. If those are missing, the decision can wait while you find its methodology. A performance chart cannot tell you what rule you are buying.
Factor investing widens that fund comparison: momentum is one of several traits a portfolio can systematically favor.
In short
- Momentum ranks stocks by past returns; a winner may simply have fallen less.
- Historical continuation is an average tendency that can reverse sharply.
- Owning winners and shorting losers has a different payoff from owning winners alone.
- Turnover, taxes and following the rule shape what reaches your account.