Why Does Theta Accelerate Near Expiration?

Somewhere around 14 days to expiry, a trader holding a long ATM call notices the bleed isn't what it was. The position had been losing a few cents a day for weeks. Manageable. Then suddenly the rate doubles, triples. The option didn't just lose time value. It started hemorrhaging it. This acceleration isn't a quirk or a market microstructure effect. It falls directly out of Black-Scholes and the mathematics of uncertainty. Every options trader encounters it, but understanding *why* it happens changes how you think about when to hold options versus when to trade out of them. It also reframes the entire relationship between gamma and theta as two sides of the same coin.

In Black-Scholes, option price depends on volatility through the term σ√T, where σ is annualized vol and T is time in years. This square-root relationship is the source of the nonlinear decay. Consider: moving from 90 DTE to 60 DTE removes 30 days from a 90-day window, one-third of the total time. But √(60/365) is about 0.405, while √(90/365) is about 0.496. The drop from 0.496 to 0.405 is roughly 18%, not 33%. The first 30 days of the option's life carry disproportionately little time value. Now move from 30 DTE to 0 DTE. √(30/365) is 0.286; at expiration it hits zero. That drop of 0.286, representing the same 30 calendar days, is larger in absolute terms than the 90-to-60 comparison above. The final 30 days carry far more time value per day than the first 30 days did. The implication: theta is roughly proportional to 1/√T. As T approaches zero, theta approaches infinity for ATM options. The last few weeks before expiration are where time decay stops being a nuisance and starts being a force.

Take a stock trading at $100. An at-the-money call with 90 days to expiry, 25% implied vol, and current rates is priced at roughly $5.80. Theta is approximately -$0.032 per day. About 3.2 cents. Roll to the same setup with only 5 days to expiry. That ATM call is priced at about $1.40. Theta is approximately -$0.22 per day, or 22 cents. Same strike, same vol, but the daily bleed is 7x higher. Now picture a cheaper option: an ATM call priced at exactly $2.00 with 90 DTE has theta around -$0.014. If vol and the underlying haven't moved and you've held it through to 5 DTE, the option is worth maybe $0.50, decaying at roughly $0.08 per day. That's 40% of remaining value evaporating per week. Long-dated options give buyers room to be wrong about timing. With 90 days, a delayed move costs relatively little per day. With 10 days, every morning the stock sits still costs real money. This asymmetry is one reason professional traders often roll positions well before final expiration rather than holding through the steepest part of the curve.

The rise of weekly options and zero-days-to-expiry trading has made theta acceleration more than an academic concept. A weekly ATM option listed on Monday morning starts life deep in the acceleration zone. There is no slow-decay phase. By Wednesday, theta is punishing. By Friday morning, whatever extrinsic value remains is vanishing by the hour. This is why 0DTE options are almost exclusively a seller's game in terms of theta. Buyers need violent, immediate moves to overcome the decay. Sellers collect premium that erodes at a rate long-dated options never reach. The math is identical to what drives the 90-to-5-DTE example above, just compressed into a single trading session. For context: a weekly ATM option on SPY might have theta of -$0.15 at listing, but -$0.40 or more by Thursday afternoon. A monthly with 30 DTE on the same underlying might sit at -$0.05 for days without much change. The weekly lives its entire life in the steep part of the 1/√T curve.

The √T relationship is most dramatic for at-the-money options. For deep out-of-the-money options, the dynamics look different. An OTM option with very little time left has almost no time value to begin with. A call struck 10% out of the money with 5 DTE and 25% vol is essentially worthless, maybe $0.02. Its theta is tiny in dollar terms because there's nothing left to decay. An OTM option with 90 days still has meaningful time value because there's a real probability of the stock reaching the strike. As expiration approaches, that probability collapses, taking the value with it. But the collapse happens earlier in the option's life relative to ATM options. OTM options lose their value more gradually across the whole lifespan rather than in a final-weeks rush. Much of the decay occurs between 60 and 30 DTE, and by the time expiration week arrives, there is often little left to lose. Practically, this means the "accelerating theta" story is primarily an ATM phenomenon. Sellers running short straddles and strangles feel it most acutely in the final month. Buyers holding OTM lottery tickets often find that time erosion has already done most of its damage well before expiration arrives.

Most of the Greeks describe shared reality. Delta is delta. Vega is vega. Both sides of a trade see the same number, even if they interpret it differently. Theta is where the conversation splits. A seller with 14 DTE looks at accelerating decay and sees a tailwind, premium collapsing in their favor faster every day. A buyer with the same position sees a countdown, each session more expensive than the last. They are reading the same √T curve from opposite ends, and the final two weeks are where those readings diverge most sharply. Neither side is wrong. The math doesn't take sides. But the experience of holding a short option into accelerating decay feels nothing like holding a long one through the same window. This is where sellers and buyers stop talking about time the same way.