Max Pain in Options: Computed From a Real Chain, Checked Against the Close
Max pain is the price at which the options open on one expiration date would be worth the least to their holders if the underlying settled there. It is computed from open interest at every strike, and the theory attached to it says prices drift toward that level as expiration nears.
Max pain turns a whole option chain into a single price, which makes it easy to quote and hard to test.
The arithmetic behind it is simple enough to do by hand, and this page does it on real Cboe option chains saved after the close on Friday 25 September 2026. It then checks the answer against where ten underlyings actually finished that day, and says plainly what one day of data can and cannot show.
How it works
Start with open interest at every strike. Open interest is the number of contracts still open, counted by the Options Clearing Corporation from the opening and closing trades the exchanges report. The OCC updates it once a day, so the figure a trader sees during a session describes positions as of the previous night.
Pretend the underlying settles at one strike. Every call option with a strike below that price would be worth the difference, times 100 shares per contract, times its open interest. Every put option with a strike above it would be worth its own difference in the same way. Everything else expires worthless.
Add the two sums, then repeat at every strike. The strike that produces the smallest total is max pain. The name comes from the idea that this is where the holders of options, as a group, would be hurt most, and the sellers of those options helped most.
Where the theory comes from
The theory is a separate claim. Some traders argue that market makers, whom the theory assumes to be net sellers of options, nudge the price toward max pain as expiration approaches.
A more careful version points to hedging: dealers adjusting stock hedges near large strikes can dampen moves, which is the mechanism behind gamma exposure. Neither version is established by the calculation itself. The number describes where contracts sit; it does not show that anyone is steering the price.
Why it spread anyway. Max pain can be computed from any chain that shows open interest, it changes every morning, and it produces a single price, which makes it easy to quote. None of that is evidence. A figure that is always available tends to get remembered on the days it lands close and forgotten on the days it does not.
What would count as evidence. A fair test would compute max pain from open interest before many expirations, then compare the gap to the close with a plain alternative such as the previous day’s close or the nearest round strike. If max pain did no better than those, the pull would be an illusion of round numbers and heavy strikes.
A worked example
SPY, expiration Friday 25 September 2026. The Cboe chain saved at 03:44 UTC on 26 September, after that day’s close, still listed the 25 September contracts with the open interest they carried into the session: 418,508 calls and 767,841 puts across 175 strikes.
One strike, done by hand. Suppose SPY settles at $770. The 765 call is then worth $5 a share, or $500 a contract, and 8,730 of them were open, so that strike alone pays $500 x 8,730 = $4,365,000. The 772 put is worth $2 a share, $200 a contract, on 4,449 contracts: $889,800.
Do this for every strike below $770 on the call side and every strike above it on the put side, and the totals are $63,761,800 to call holders and $2,954,500 to put holders, or $66,716,300 in all.
Now slide the price. At $765 the total falls to $52,487,800: calls $33,726,700, puts $18,761,100. At $766 it is $51,972,900, the smallest of any listed strike, so $766 was max pain for that expiration. One dollar higher, at $767, the total is back up to $52,979,600.
At $760 it is $84,181,600 and at $780 it is $233,373,200, because heavy call open interest sat just above the market.
Where SPY actually finished. SPY closed at $771.35, up from $767.18 the day before. That is $5.35, or 0.69%, above max pain. At the $771 strike, the nearest one, holders would have collected $75,661,900, about $23.7 million more than at max pain. The day’s move took SPY further from max pain, not closer.
The original data
The sample. Cboe’s delayed quote chains for ten heavily traded underlyings, all saved after the 25 September 2026 close, each with open interest for the contracts expiring that day. Closing prices for 24 and 25 September come from Yahoo Finance and match the closes in Cboe’s own files.
Max pain was computed at listed strikes only, exactly as described above. Every figure is in the max pain table, and SPY’s full curve is in the SPY payout file.
How far each close landed from max pain. The median gap across the ten was 1.76% of the closing price. AMZN finished 0.13% from its $250 max pain, IWM 0.37%, TSLA 0.57% and SPY 0.69%, so 4 of 10 closed within 1%. The widest gaps were AMD at 8.42% (close $630.63, max pain $577.50), MSFT at 4.10% and META at 3.88%.
QQQ, NVDA and AAPL sat between 1.28% and 2.51%.
Which way the day went. Comparing each close with the previous day’s, 4 of the 10 ended nearer to max pain and 6 ended further away. The median absolute move on the day was 0.50%, so most were ordinary sessions, although MSFT and META each moved more than 3%.
The closest cases prove the least. AMZN closed at $249.67 against a max pain of $250, which is also the nearest strike and a round number. A price sitting near a round strike with heavy open interest looks the same whether max pain pulled it there or it simply ended the week near a common level.
What one expiration can and cannot show
What cannot be measured here. A real test needs strike-level open interest for many past expirations, taken before each one, set against the closing price that settled it. Cboe’s free delayed chain shows only the current day, and the saved files hold only this one expiration.
So this page does not claim max pain works or fails over time. It shows one Friday: the gap was under 1% for four underlyings and over 3% for three.
The forward numbers are not results. The same chains put SPY’s max pain at $765 for both the 2 October and 16 October 2026 expirations. Open interest will change every night until then, so those figures describe positions on 25 September, not a forecast.
When it fails
Open interest is stale by design. It is counted overnight, so a large position opened or closed on expiration day itself is invisible in the calculation until it no longer matters.
Open interest has no side. A call held by a hedge fund against a short stock position and a call sold by a market maker look identical in the count. The theory assumes dealers are short the options, and the data cannot say whether that is true at any strike.
It ignores the stock market. An earnings report, a rate decision or a sector selloff moves prices far more than any hedging flow tied to one expiration. On 25 September 2026 AMD finished 8.42% above its max pain with a day’s move of only 0.22%: it never came near the level.
Many expirations overlap. SPY lists contracts expiring on every weekday; the 25 September chain had one for each session from 28 September to 9 October. Each has its own max pain, and they can point in different directions on the same afternoon, so a single figure hides competing pulls.
The cost of trusting it is concrete. A short option position opened on the belief that price will finish at max pain carries the full risk of a move away from it, which is what six of these ten underlyings did on an ordinary day.
Related
The open interest page covers the count this number is built from and why it only updates overnight. Options expiry explains what happens to contracts at the close on their last day and why the final week behaves differently. And gamma exposure describes the dealer-hedging mechanism that gives the pinning idea its most defensible form.
I read max pain as a map of where open contracts sit, never as a price target. Check how far away it is and how much open interest is behind it before paying it any attention, and ignore it entirely on a day with news.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.