The Earnings Implied Move vs the Realised Move

The night before a company reports earnings, its at-the-money straddle — a call and a put at the same strike — is priced almost entirely for one thing: how much the stock will jump overnight. Back that price out into a percentage and you have the market's own forecast of the earnings-day move, the implied move. This is about comparing that forecast to what the stock actually does, systematically, across many earnings events, to see whether the forecast runs persistently too high or too low.

Reading the implied move off the straddle price

The at-the-money straddle price divided by the stock price gives a rough estimate of the expected move, because a straddle's value is dominated by the magnitude of the move, not its direction — you profit whether the stock goes up or down, as long as it goes far enough to cover the premium paid.

Worked example. A stock trades at $100 the afternoon before earnings. The at-the-money straddle ($100 call plus $100 put, same expiry, a few days out) is quoted at $6.00 combined. Dividing premium by stock price, $6.00 / 100 = 6\%$, is the market's implied overnight move. If the stock actually moves 4% after the report, the straddle buyer loses money even though the stock did move: it just moved less than priced in. If it moves 9%, the buyer profits well beyond the premium paid.

The systematic finding, and the trade built on it

Studies going back to Patell and Wolfson in 1979 find that implied volatility rises ahead of earnings in a way that roughly anticipates the extra uncertainty, but that on average, across many stocks and quarters, the realised move comes in somewhat below the implied move — consistent with a modest volatility risk premium specific to earnings, similar in spirit to the general variance risk premium but tied to a known, dated catalyst. The trade built on this finding is selling the straddle (or strangle) into earnings and buying it back, or letting it expire, the next morning — collecting the gap between what was priced in and what typically happens.

Worked example, continued. If the historical average realised move for a class of stocks running this trade is 4.5% against an average implied move of 6%, selling the straddle systematically captures that 1.5 percentage point gap on average, across many trades, in exchange for occasionally being on the wrong side of a much larger-than-implied move on any single name.

What erodes the edge

• The distribution is fat-tailed by nature. Earnings surprises are rare-but-large — a single quarter can move a stock 30–40% on a guidance shock, wiping out many quiet quarters of small gains at once.
• The trade is well known. Systematic earnings-straddle-selling has been traded for decades; the gap is compensation for bearing tail risk, not a free inefficiency, and it varies by name, sector, and regime.
• Diversification is weaker than it looks. Running the strategy across many single-name events reduces variance, but earnings season clusters reports into a few weeks, and market-wide shocks can push many names' realised moves up together.

In interviews

Show you can back out the implied move from a straddle price by hand (premium divided by spot), then state the empirical finding precisely: realised moves average somewhat below implied moves, which is a volatility risk premium specific to the earnings event, not a mispricing that violates any no-arbitrage bound. Finish with the risk framing — the edge is real on average and compensation for real tail risk, and sizing has to respect that the loss distribution is skewed the wrong way for anyone selling premium.