Implied Volatility for Beginners: What IV Rank Tells You Before an Earnings Report

Every input to an option's theoretical price is observable except one. Strike, expiration, spot price, the risk-free rate, the dividend schedule — all published. Implied volatility is the leftover: the volatility number you have to feed the pricing model to make its output equal the price the option is actually trading at.

That distinction matters. IV is not a forecast someone wrote down. It is a residual, solved for backwards out of a market price. Which means when you say "IV is high," you are saying "options are expensive relative to what this model would charge at lower volatility" — and nothing at all about which direction the stock is going.

Earnings season is where beginners meet this number for the first time, usually because a screener flagged a stock with an IV rank of 92 and it looked like a signal. Most of the time it is not one, and the reason is structural.

What implied volatility actually measures

IV is quoted as an annualized standard deviation, in percent. An IV of 40% says the options market is pricing a roughly one-standard-deviation move of 40% over a year, under the model's assumptions: lognormal returns, continuous trading, no jumps.

To get a usable number, scale it to your horizon. Multiply by the square root of the time fraction:

• Stock at $100, IV of 40%, 7 calendar days until expiration
• 0.40 × √(7 / 365) ≈ 0.0554
• One-sigma expected move ≈ $5.54 over that week

So the market is pricing roughly a two-thirds chance the stock finishes that week inside $94.46 to $105.54. That is the entire practical content of an IV quote.

Two caveats before you lean on it. First, the lognormal assumption underprices tail moves, and earnings are exactly the event that produces tails. Second, there is no single IV for a stock — every strike and every expiration has its own, which is what skew and term structure describe. When a data provider shows you one IV number, it has picked a convention (often a 30-day constant-maturity interpolation), and different vendors pick differently.

IV rank and IV percentile are not the same number

Both answer "is today's IV high for this stock?" using a lookback window, typically 52 weeks. They answer it differently.

IV rank is a position within the range:

IV rank = (current IV − 52w low) / (52w high − 52w low) × 100

It uses exactly three numbers. One panic day a year ago can set the high and permanently compress every reading since.

IV percentile is the share of trading days in the lookback where IV closed below today's level. It uses the whole distribution.

A worked case where they disagree hard. Suppose a stock's IV over the past year ranged from 20 to 80, but it spent most days near 30 — the 80 came from a single week of takeover speculation. Today IV is 35.

Same stock, same day, same IV. Opposite conclusions. IV rank is the more common default on retail platforms because it is trivial to compute, and it is the more fragile of the two.

Before you compare IV rank across two tickers, check that your platform computes them the same way. Lookback length (52 weeks vs 12 months vs 252 trading days), whether it uses a 30-day constant-maturity IV or the front-month ATM IV, and whether it samples closes or intraday extremes all move the output. A rank of 60 on one broker and 45 on another for the same symbol is usually a convention difference, not a data error.

Why IV rank spikes before earnings, and what to do with that

An earnings report is a scheduled event with a known date and an unknown outcome. Any option that spans that date has to price the jump risk, so front-month IV climbs into the print. After the release, the uncertainty resolves, IV falls back toward its baseline, and the option loses value even if the stock moved. That drop is the volatility crush, and it is the expected behavior of the pricing model, not a market inefficiency.

The consequence: a high IV rank a few days before earnings tells you almost nothing, because nearly every liquid single-name stock has a high IV rank a few days before earnings. The screener is finding a calendar, not an edge.

Three checks make the number informative:

1. Measure the term structure inversion. Compare ATM IV in the expiration immediately after earnings to the next expiration out. In normal conditions IV rises with time to expiry. Into a print, the front expiration inverts above the back one. The size of that gap isolates how much of the premium is event premium rather than baseline volatility.

2. Back out the implied move. The quick approximation: take the ATM straddle price (call + put at the strike nearest spot) in the first expiration after earnings and divide by the stock price. A $6.00 straddle on a $100 stock implies roughly a 6% move. It is a rough estimate — it slightly overstates the one-sigma move — but it is fast and it is what the market is charging.

3. Compare implied to realized, quarter by quarter. Pull the last 8 to 12 earnings dates for the ticker and record the implied move going in and the actual close-to-close move coming out. This is the only step that produces information the screener did not already have. A stock whose implied move has consistently exceeded its realized move has historically paid option sellers. That is a description of the past, not a prediction, and a single regime change (a lawsuit, a guidance reset, an activist stake) breaks the pattern.

Buying a straddle into earnings because "the stock always moves" ignores the break-even. You need a move larger than the implied move to profit, and you are simultaneously long a vega position that is about to get marked down. Selling premium into the crush has the mirror problem: defined profit, undefined or large loss if the move exceeds what you collected. Weekly options around earnings also carry the widest bid-ask spreads of the cycle — model your fills at the mid and then check what the actual spread costs you on entry and exit.

The quarter-by-quarter comparison is the part people skip, because it requires keeping records across quarters rather than reading a live screen. A plain table with ticker, date, implied move, realized move, and a one-line note on what drove the surprise is enough. What you want after four quarters is not a rule — it is calibration on how often your read was wrong.

What IV rank will not tell you

• Direction. IV is symmetric by construction. Skew hints at where demand for protection sits, but rank itself has no directional content.
• That mean reversion is coming. "High IV rank" is often read as "IV will fall." Sometimes IV is high because the situation genuinely changed — a pending regulatory decision, a going-concern question, a merger vote — and the elevated level is correct until the event resolves.
• What is in the same expiration. A macro print, an index rebalance, or a product launch inside the same window is priced into the option too. The earnings-only implied move you calculated may be contaminated.
• Whether the lookback is honest. A ticker with fewer than 12 months of trading history, or one that had a single volatility event dominating the range, produces a rank that is arithmetically valid and practically meaningless.

IV rank is a compression of a year of data into one integer. Used as a filter to decide what deserves a closer look, it is fine. Used as a trigger, it mostly detects the earnings calendar. The work that separates the two is the implied-versus-realized history, and there is no screener shortcut for it.

This is educational material about how options pricing conventions work, not investment advice, and none of it accounts for your position sizing, tax situation, or risk tolerance.

Originally published at pickuma.com. Subscribe to the RSS or follow @pickuma.bsky.social for new reviews.