The Value Line Timeliness Anomaly

Value Line was, and still is, a research publisher that ranks around 1,700 stocks every week from 1 (expected to outperform over the next 6–12 months) to 5 (expected to underperform), using a fixed, published formula combining earnings momentum, price momentum, and earnings surprises. Because the formula was mechanical and public — anyone could read Value Line's methodology and replicate the ranking — standard finance theory predicted it shouldn't be able to beat the market by much, if at all, once it became widely known.

It did anyway. Academic studies from the 1970s through the 1990s repeatedly found that stocks ranked 1 by Value Line outperformed stocks ranked 5 by several percentage points a year, and the gap persisted for years after the methodology itself was public and the newsletter had tens of thousands of subscribers.

The competing explanations

Researchers offered three broad explanations, and the honest answer is probably some mix of all three. First, known-factor overlap: much of Value Line's edge could be replicated by combining well-documented factors like earnings momentum and price momentum, meaning the ranking wasn't a truly independent anomaly, just a bundled version of effects academics separately understood; later research (notably Choi, 2000) showed a large fraction of the return spread could indeed be explained this way. Second, implementation frictions: many rank-1 stocks were small or mid-cap and had wide bid-ask spreads, so the paper return from buying rank-1 and shorting rank-5 stocks shrank substantially once realistic trading costs were subtracted. Third, and least satisfying: some residual predictive power remained unexplained even after adjusting for known factors and costs, which is the part that kept the "enigma" label attached to the anomaly for so long.

Worked example

Suppose over a given year, a portfolio of Value Line rank-1 stocks returns 14% while the broad market returns 10%, and rank-5 stocks return 5%. A long-rank-1/short-rank-5 portfolio earns roughly a 9-point spread before costs. If academic factor models attribute 6 points of that spread to already-known earnings and price momentum exposures, only about 3 points of "unexplained" edge remains — still statistically notable across many years of data, but a much smaller mystery than the headline 9-point spread suggested. That shrinking-on-closer-inspection pattern is typical of how quant researchers have picked apart the anomaly piece by piece.

What this means in practice

The Value Line case is a standard teaching example for two lessons at once: mechanical, fully disclosed ranking systems can persist longer than theory predicts, especially when real-world trading costs limit how much competing capital can actually exploit them; and a large "anomaly" often turns out to be a bundle of smaller, already-known effects once researchers control carefully for factor exposure.