In 1993, finance professors Narasimhan Jegadeesh and Sheridan Titman published a finding that should have vanished within years. Stocks that had performed well over the past three to twelve months, they showed, tended to keep performing well. Losers kept losing. Thirty years later, the pattern remains stubbornly intact.

This creates an uncomfortable puzzle for anyone who believes markets are efficient. If a profitable strategy is widely known, sophisticated traders should exploit it until it disappears. Yet momentum persists—not just in US equities, but across international markets, currencies, commodities, and bonds. It's the cockroach of market anomalies.

The persistence of momentum forces us to ask deeper questions about how markets actually process information. Is this pattern evidence of investor irrationality that somehow resists correction? Or does it reflect something more fundamental about the structure of financial markets themselves?

The academic evidence for momentum is remarkably consistent. Studies spanning over a century of US data, dozens of international markets, and multiple asset classes all point to the same conclusion: past winners outperform past losers over horizons of three to twelve months. The effect typically generates annual excess returns of 6-12%, making it one of the most economically significant patterns in finance.

What makes momentum particularly compelling is its robustness to different measurement approaches. Whether you rank securities by returns over the past six months or twelve months, whether you hold positions for three months or nine, the pattern emerges. It appears in small stocks and large stocks, growth companies and value companies, developed markets and emerging markets.

The pattern even predates modern financial markets. Researchers examining commodity prices from the 1800s and equity prices from the early 1900s find similar momentum effects. This isn't a statistical artifact of data mining—it's a feature of how prices have always behaved when human beings set them.

Critics initially suggested momentum was simply compensation for risk. Perhaps recent winners become riskier, and the higher returns just reflect that elevated risk. But extensive testing has failed to identify any traditional risk factor that explains momentum returns. The premium remains after controlling for market exposure, size, value, and other known factors. Something else is driving this pattern.

Takeaway

Momentum's persistence across time, geography, and asset classes suggests it reflects something fundamental about how markets process information—not a temporary inefficiency waiting to be arbitraged away.

If momentum is real and profitable, why don't sophisticated investors trade it away? The answer lies in understanding the practical constraints that limit arbitrage. Knowing about an anomaly and profiting from it consistently are very different things.

Career risk looms largest. Momentum strategies can underperform for extended periods—sometimes years. A portfolio manager who trails their benchmark for three consecutive years may not have a fourth year to prove themselves right. When momentum crashed spectacularly in 2009, many momentum-based funds faced massive redemptions precisely when sticking with the strategy would have been most profitable.

Capital constraints compound the problem. Momentum strategies require frequent rebalancing, generating transaction costs that eat into returns. They also involve shorting recent losers, which creates additional costs and risks. Borrowing fees for hard-to-short stocks can exceed the expected return from the position. During market stress, short positions become even more expensive or impossible to maintain.

There's also profound timing uncertainty. Momentum works on average over long periods, but any individual momentum bet might fail. The strategy is vulnerable to sudden reversals—what practitioners call 'momentum crashes.' These tend to occur after sharp market declines, exactly when investors can least afford to absorb losses. The 1930s, 2009, and 2020 all saw violent momentum reversals that wiped out years of accumulated gains in weeks.

Takeaway

Arbitrage has limits. Career risk, capital constraints, and timing uncertainty create gaps between theoretical profits and real-world implementation—allowing patterns like momentum to persist.

For investors considering momentum, the first decision is whether to build positions directly or access the strategy through funds. Direct implementation offers lower fees but requires regular rebalancing, disciplined execution, and comfort with periods of underperformance. Most individual investors underestimate how difficult it is to stick with a strategy that's currently losing money.

If implementing directly, research suggests using formation periods of six to twelve months—the lookback window for measuring past performance. Holding periods of three to six months balance capturing the momentum effect against transaction costs. Excluding the most recent month's returns is important, as very short-term reversals can contaminate the signal.

Risk management is essential because momentum crashes, while rare, can be devastating. Some practitioners reduce momentum exposure when recent market volatility spikes, as crashes often follow turbulent markets. Others cap position sizes or diversify across multiple momentum timeframes. The goal is surviving the inevitable bad periods to capture long-term returns.

Perhaps most importantly, momentum works best as one component of a broader approach. Combining momentum with value strategies has historically reduced drawdowns, as the two factors tend to move independently—sometimes even inversely. When momentum struggles, value often thrives, and vice versa. This diversification benefit may matter more than optimizing any single factor's implementation.

Takeaway

Implementing momentum successfully requires managing the psychological and practical challenges of strategy drawdowns—the edge comes not from knowing the pattern exists, but from having the discipline to stick with it when it hurts.

Momentum's persistence reveals something important about financial markets: they are not purely information-processing machines. They are ecosystems shaped by human psychology, institutional constraints, and structural frictions that prevent quick corrections of even well-known patterns.

This doesn't mean momentum is free money. The strategy demands patience through painful drawdowns, discipline in execution, and realistic expectations about returns. Many investors who intellectually understand momentum still fail to profit from it because they abandon the approach during inevitable rough patches.

The deeper lesson may be this: market anomalies persist not because investors are ignorant of them, but because exploiting them requires bearing risks that many cannot or will not accept. Knowledge alone doesn't create alpha. Willingness to endure discomfort does.