Dividend Risk In Options

A stock that pays a $1 dividend doesn't just hand shareholders $1 — its price also drops by roughly $1 on the ex-dividend date, because the company just sent $1 per share out the door and is worth that much less. This isn't random noise the market has to react to; it's scheduled and knowable in advance, announced weeks ahead. Options pricing has to bake that scheduled drop in correctly, and two very different failure modes show up when it doesn't.

Withdrawing cash from a joint account before splitting it

Imagine two people own a savings account together and have agreed to split whatever's in it on a fixed future date. If one of them withdraws $1,000 from the account the week before the split, the remaining balance to be divided is $1,000 smaller — that's not a surprise or a market shock, it's arithmetic everyone should have accounted for in advance. A dividend does the same thing to a stock: the company withdraws cash from itself and hands it to shareholders, so the remaining company (and hence the stock price) is worth that much less the moment the dividend is paid out. An option that doesn't build the expected withdrawal into its pricing is pricing the wrong company.

The formula

$$C = C_{BS}(S - PV(D), K, \sigma, r, T), \qquad S_{ex} = S_{cum} - D$$

In plain English: to price an option correctly on a dividend-paying stock, don't plug the current stock price $S$ straight into Black-Scholes. Subtract the present value of dividends expected before expiry, $PV(D)$, first — because the option holder never gets the dividend, only the stock price after it's been paid out, so the option should be priced off a stock price that already reflects the coming departure of that cash. On the ex-dividend date itself, the stock's price mechanically drops from its "cum-dividend" price to its "ex-dividend" price by (approximately) the dividend amount $D$. A wrong estimate of $D$, or a wrong estimate of when it happens, feeds a wrong stock price into the pricing formula and produces a mispriced option on both sides of the market.

Worked example 1: pricing a call correctly around a dividend

A stock trades at $100 and will pay a $2 dividend in 30 days, before a 60-day option expires. Risk-free rate 4%. Present value of the dividend: $2 \times e^{-0.04 \times 30/365} = 2 \times 0.9967 = 1.9934$, i.e. $1.9934. The dividend-adjusted stock price to feed into Black-Scholes is $100 - 1.9934 = 98.0066$, i.e. $98.0066, not $100. Pricing a $100-strike call using $100 instead of $98.01 overstates the call's delta-relevant stock price by about $2 — for an at-the-money option with delta near 0.5, that alone overprices the call by roughly 0.5 \times 2 = \1.00, a substantial error for what might be a \5 option. Puts move the opposite way: using the wrong (too-high) stock price would underprice the put by a similar amount, since puts gain value as the stock falls.

Worked example 2: early exercise triggered by a dividend

An investor holds an American call, deep in the money, on a stock about to pay a large dividend. Stock at $100, call struck at $80 (16 days to expiry), and a $3 dividend goes ex-dividend tomorrow. The call's intrinsic value today is $100 - 80 = 20$, i.e. $20; its remaining time value, given so little time left and how deep in the money it is, is small — say $0.30. If the holder exercises today (just before the ex-date), they capture the stock at $100 cum-dividend and are entitled to tomorrow's $3 dividend as the new owner: net position $100 stock plus $3 dividend right, minus the $80 paid, worth $23 minus the $80 exercise cost nets to a locked-in $20, plus $3 of dividend rights = effectively $23 of value captured. If instead they hold the call through the ex-date, the stock drops to roughly $97, and the call is now worth about $97 - 80 = 17$, i.e. $17, intrinsic plus the same $0.30 time value, or $17.30 — clearly worse than the $23 available by exercising early and collecting the dividend outright. This is exactly why deep in-the-money American calls are exercised early, almost always the day before a large dividend, and almost never otherwise.

Picture this explorer's smooth paths with a single sharp, known-in-advance notch cut into each one at the ex-dividend date — dividend risk is really the risk of getting the size or timing of that notch wrong, not risk about the notch existing at all.

What this means in practice

Every options pricing system on a dividend-paying underlying needs a dividend schedule, not just a current stock price and volatility — a wrong dividend forecast (a company cutting or raising its dividend unexpectedly) moves option prices immediately, independent of any move in the stock itself. Early-exercise risk from dividends is also why option holders and short sellers of American calls both watch the ex-dividend calendar closely: being short an in-the-money call on a stock about to pay a large dividend means a real chance of assignment.