Conservatism Bias and Earnings Surprises

Investors often update their beliefs too slowly when earnings arrive, a pattern known as conservatism bias earnings surprises. This sluggish re-evaluation opens a window where stock prices drift predictably upward (after positive surprises) or downward (after negative ones) over the weeks following an announcement—a phenomenon called post-earnings drift.

What Is Conservatism Bias?

Conservatism bias is a cognitive tendency to cling to prior beliefs even as new evidence contradicts or reinforces them. Rather than updating beliefs sharply in response to surprising new information, people anchor to their original view and adjust only partially. This leaves a gap between the stock’s initial price reaction and where it should ultimately settle once the market fully processes the news.

When a company beats earnings expectations, the stock might jump 2–3% on the announcement day. But if investors are exhibiting conservatism bias, they treat the beat as a one-off anomaly rather than a signal that their earnings forecasts were systematically too low. Over the following three to six weeks, as evidence accumulates that the beat reflects genuine operational strength, the stock drifts higher. The same logic works in reverse for negative surprises.

Post-Earnings Drift as the Visible Pattern

Post-earnings drift (PED) is the observable market manifestation of conservatism bias. After an earnings announcement, the abnormal return doesn’t end on day one; instead, the stock continues to drift in the direction of the surprise. Academic research has documented this pattern across decades and markets.

A positive earnings surprise might produce a 1% jump on announcement day and a further 2–4% drift over the following month. This lag seems to contradict the efficient market hypothesis, which suggests prices should incorporate new information instantly. Yet the drift persists because conservatism bias is widespread: individual investors and even some institutions don’t update their forecasts sharply enough, allowing sophisticated traders to exploit the mismatch.

Why Investors Underreact

Several mechanisms explain why conservatism bias produces predictable underreaction to earnings.

Anchoring on old expectations. Analysts and investors publish earnings forecasts weeks or months in advance. When actual results arrive, the psychologically salient comparison is to that prior prediction. A beat of 5% might feel modest compared to historical volatility, even if it signals a genuine shift in the company’s trajectory. This anchoring slows the repricing.

Difficulty in revising mental models. Updating an earnings forecast requires revising the underlying story: product demand, competitive position, cost structure, growth rate. Many investors hold a single coherent narrative about a company (e.g., “mature slow-growth name” or “turnaround candidate”). A single quarter that contradicts that narrative often produces reluctance to abandon it wholesale. Instead, the earnings beat is mentally filed as “noise” or “temporary,” and belief updating lags.

Information decay. As days pass after earnings, the news becomes less salient. New headlines accumulate. For passive investors or those monitoring dozens of positions, the earnings surprise fades into background. But the stock price continues to drift as the market collectively recognizes what it missed. Conservatism bias creates a kind of temporal friction—belief updating is stretched across weeks rather than compressed into hours.

Evidence and Predictability

The academic literature, beginning with studies in the 1980s and 1990s, has documented PED as a robust phenomenon. Stocks that beat earnings expectations by one standard deviation tend to outperform over the 60 days following announcement. Stocks that miss by the same margin tend to underperform. The drift magnitude varies, but the direction is consistent.

What makes this noteworthy: if conservatism bias universally caused underreaction, the drift should offer a trading profit for anyone patient enough to hold after earnings. Yet the drift persists even as the evidence has become well-publicized. This suggests the bias is partly structural (real costs to gathering information and revising models) and partly behavioral (genuine cognitive resistance to belief change, not pure ignorance).

The Role of Forecast Revisions

Analyst earnings revisions accelerate the market’s repricing. When a company beats earnings and the street revises next-quarter or next-year forecasts upward, the post-earnings drift accelerates. Conservatism bias can operate at the analyst level too: after one surprise, analysts are still reluctant to raise forecasts dramatically, expecting reversion to the old trend. This further delays price discovery.

Over multiple quarters—if a company continues to beat—conservatism bias erodes. The market eventually accepts the new reality. But the temporary mismatch between price and justified value, lasting weeks to months, is where the pattern plays out.

Limitations and Alternative Views

Not all post-earnings drift reflects conservatism bias alone. Some may reflect risk considerations: a company that beats earnings might signal higher volatility or competitive intensity, raising required returns. Some drift may reflect market microstructure (limited arbitrage capital prevents instant repricing). And some drift appears to have shrunk in recent decades as information spreads faster and index investing has grown.

Moreover, the drift can operate asymmetrically: negative surprises sometimes produce sharper initial selloffs (possibly driven by loss aversion) than positive surprises produce gains, complicating the conservatism story.

Practical Implications

Understanding conservatism bias and post-earnings drift matters for different participants differently. Long-term investors can be aware that initial earnings surprises may not be fully priced; further drift creates an opportunity cost if they anchor too much to the announcement-day price. Short-term traders have historically found trading on earnings surprises (if they can execute quickly and cheaply) worthwhile. Analysts should check whether they’re anchoring too much to their prior forecast or the consensus, allowing genuine changes in fundamentals to slip past unrevised.

See also