Iron Condor
An iron condor writes an out-of-the-money put and an out-of-the-money call, and buys a further put and a further call as wings. The worked example writes the FTSE 100 October 10,400 put and 11,100 call and buys the 10,150 put and 11,350 call for a net credit of 91.0 points, £910.00 at £10 a point. The most it can lose is one wing's width less the credit, £1,590.00, or £1,596.80 with the opening commission. What it gives up is everything outside the 700 points between the short strikes: a move either way costs money, and a quiet market pays no more than the credit. It is built for an index that stays inside a range until the options expire.
This page assumes the reader has met its two halves, the bull put spread and the bear call spread. Every figure is modelled, not quoted: inputs and method.
Four FTSE 100 October options, 700 points apart at the middle
Two things about the construction matter more than the strike numbers. The wings cost £695.00, 43.3% of the £1,605.00 the two short options bring in on their own; that is the price of turning an open-ended short strangle into a position whose worst case is known on the first day. And the two wings are the same width, so only one side can finish in the money and the worst case is one width, not two. The short strikes sit the same distance from the index, 350 points, but not at the same delta: the site's FTSE surface gives lower strikes higher volatility, so the 10,400 put is priced at 15.32% and the 11,100 call at 12.72%.
These are ICE FTSE 100 index options (ESX): £10 a point, European, so they can only be exercised at expiry, and settled in cash against the exchange delivery settlement price, so no shares change hands and no stamp duty arises (FTSE 100 contracts; every contract size is on the basics page). A spread needs a margin account. IBKR lists a short iron condor at its Options Level 3 (account types and permissions), and a broker holds one width, £2,500.00, of which the credit covers £910.00, so £1,590.00 of the account's own money for the whole life of the position (spread margin).
Open this worked example in the strategy builder. The builder page's own worked condor uses the same four strikes at model value, with each volatility rounded to two decimals, so its premiums differ from this page's by a few hundredths of a point; this page fills each leg on the tick, and the builder then solves each leg's volatility from the fill.
Where the condor pays on 16 October, and on the way there
The flat top runs from 10,400 to 11,100, and the position makes money anywhere between 10,309 and 11,191: a band 882 points wide, 8.2% of the index. The one-standard-deviation range the model gives for 16 October, 10,234 to 11,292, is wider than the profitable band at both ends: a move of about one standard deviation either way already costs money. The dotted entry-day line shows how little an early move registers: on the first day a fall to 10,250 would mark at −£316.14, not the £1,590.00 it would cost at expiry, because 45 days of time value still sit in all four legs.
Read across the middle row and the credit arrives slowly, then quickly: with the index unchanged the position shows £430.18 with 21 days left and the rest in the last three weeks. Read down any column and the shape of the risk changes. At 10,250 the mark is −£316.14 on the first day, −£321.86 with 21 days left and −£400.87 with a week left: time helps the position while the index sits inside the short strikes and hurts it once the index sits outside them.
Probability against payoff: six condors on one chain
Every condor on a chain is a choice of where to put the short strikes and how wide to make the wings. Move the short strikes out and the model probability of keeping money rises; the credit falls faster. The table prices six condors on the same 45-day chain, each with 250-point wings, and the chart plots the three columns that carry the trade-off. Probabilities are model probabilities (risk-neutral, from the FTSE skew surface), not forecasts.
From the tightest condor to the widest, the model probability of a profit climbs from 43.2% to 81.5%, and the credit per £100 at risk collapses from £151.26 to £15.74. The last column is the point of the table: once each result is weighted by its model probability, every row comes to within £2 of zero, with the credit assumed to earn Bank Rate until 16 October; what is left is the fills rounding to the tick. A high probability of profit is bought with a large loss when the minority outcome arrives, and a large credit is bought with a low probability of keeping it. Neither end of the menu is better on the model that prices it.
A common US convention writes both short strikes near 16 delta, which on this chain means about 10,200 and 11,250. The nearest row of the menu is the ±550 condor (deltas −0.16 and 0.13): a model probability of profit of 75.3% for £485.00 of credit against £2,015.00 at risk. This page's condor writes nearer, at about 25 delta, for more credit and a narrower band. Why a writer might prefer either, and what evidence there is for delta conventions, is on the methods page. Delta is also not the chance of finishing in the money: the 10,400 put has a delta of −0.254 but a model probability of 22.6% of expiring in the money on the surface (delta is not a probability).
The wings are the second choice. Holding the short strikes at 10,400 and 11,100 and changing only the wings:
Widening the wings from 100 to 250 points raises the credit by 104.5% but the maximum loss by 186.5%. Wider wings make the maximum loss less likely and larger; the margin a broker holds grows with it, because for a condor the broker holds the width and the part of it beyond the credit is the maximum loss. Width is therefore the size decision, and the credit follows from it. The version with no wings at all is the short strangle, a Level 3 structure.
What 14% implied volatility paid, and what 10% or 18% would have
The model sheet prices the FTSE 100 at 14% implied volatility at the money. For scale, FTSE Russell's 30-day FTSE 100 implied volatility index (IVI) averaged 13.55 in 2025, 17.06 in the year to June 2026 and 18.96 since 2000 (FTSE 100 IVI, on the implied-volatility page). So this example sits near last year's average and below the long-run one. The same four strikes on the same day, with every strike's volatility moved up or down:
Four points of volatility either way move the credit from £510.00 to £1,205.00, and the maximum loss the other way, from £1,990.00 to £1,295.00, because the width is fixed. The condor's vega at entry is −£86.22 per volatility point: a writer who opens in a richer market is paid more for the same 250 points of risk, and one who opens in a quiet market is paid less. Whether implied volatility on the FTSE 100 has tended to exceed the volatility that followed is covered, with FTSE Russell's data since 2000, on the implied-volatility page; the library's IV-rank bands are explained there too (IV rank).
Skew changes where the credit comes from. On a flat 14% chain the put spread would be worth 45.57 points and the call spread 47.23; on the surface the put spread is worth 42.95 and the call spread 48.24. The surface makes both puts dearer than at 14%, and the bought 10,150 put gains more than the written 10,400 put because it sits further down the slope, so the put spread is worth less. On the call side both calls are cheaper than at 14%, the bought 11,350 call most of all, so the call spread is worth more. What skew prices is a fatter tail below the index: the model probability of the maximum loss on the put side is 12.7% and on the call side 11.9%, even though the call side's strikes are no further away (skew).
Scheduled events between 1 September and 16 October
Any 45-day window on the FTSE 100 contains scheduled news. US inflation figures come out every month and the Bank of England's Monetary Policy Committee meets eight times a year, so a writer cannot choose a window without them; what can be known is which events fall inside, and what a move of a given size would cost on the day.
A one-day move of one standard deviation at 14% is about 95 points. At entry the condor barely notices it: an instant 95-point fall changes the mark by +£16.97. With seven days left and the index at the 10,400 short put, the same fall changes the mark by −£370.71. The market prices scheduled news as extra implied volatility on the expiries that contain it (event premium); on this condor each volatility point is worth about £86.22 at entry.
The October condor from 1 September to 16 October 2026
Branch A: the index stands still, and two conventions disagree by one trading day
The index is still at 10,750 on Friday 25 September with 21 days left, and nothing else has changed. The condor marks 47.98 points: 47% of the credit has been earned, short of the half that the library's profit convention looks for (45.50 points). The library's 21-day convention closes here anyway, because the last three weeks are where the gamma sits: on this day the position earns £24.91 a day and carries −£0.085 of gamma per 10 points, against −£0.044 at entry.
Had the writer waited one trading day for the half-credit mark, it would have come on Monday 28 September at 40.18 points, a buy-back at 40.0 for £496.40. Holding to 16 October with the index unchanged would have kept £903.20. From 25 September the model gave the position a 78.3% probability of ending in profit and a 9.3% probability of the maximum loss. Closing on 25 September gives up as much as £486.80 of further profit in exchange for three fewer weeks in the gamma; the methods page says where each comes from.
Branch B: 400 points lower with 21 days left
By Friday 25 September the index has fallen 400 points to 10,350, just below the short put, and every strike's implied volatility is one point higher. The condor marks 101.10 points, a result of −£101.01 before costs, and a position that opened almost flat is now long £2.12 for every index point, because the short put has come into play while the call side has almost no value left. Its theta has fallen to £5.03 a day and its vega to −£20.98.
The close fixes a small loss; holding keeps a 46.5% model probability of the full £903.20 against 29.0% of losing £1,596.80. Rolling the untested call spread down, rolling the put spread out, or converting the position to an iron butterfly are all priced, on a 60-day FTSE condor tested the same way, on the rolling page; each changes where the pounds land, and none repairs the position on the model.
Branch C: a 10% fall overnight
On Tuesday 8 September, 38 days before expiry, the index opens 10% lower at 9,675 and every strike's volatility is ten points higher. Both put strikes are now in the money. The condor marks 189.31 points, a result of −£983.10 before costs, 62% of the maximum; the long 10,150 put is doing its job. Closing at the open, at 189.5 points, fixes −£998.60 after commissions; holding leaves the loss between that and £1,596.80 depending on the EDSP.
The same move on the short strangle with the same short strikes shows why the wings exist. The strangle collects £1,605.00 rather than £910.00, but after the gap it marks −£6,495.59, 6.6 times the condor's loss. Its margin also moves against it. On the Cboe formula for broad-based index options, used here only as a yardstick (a US rule; ICE Clear Europe and each UK broker set their own requirements), the strangle needs £14,230.00 at entry (£12,625.00 beyond its £1,605.00 credit) and £22,613.10 after the gap, while the condor's requirement stays at its £2,500.00 width, £1,590.00 of it beyond the credit. An account of £20,000 holding the strangle would be left with £13,504.41 of equity against a £22,613.10 requirement, and a broker can close positions to restore it (the margin spiral; history of such days: gap history).
Branch D: Thursday's close against Friday's EDSP
On Thursday 15 October the index closes at 10,342, 58 points below the short put, with one day left. The condor marks 69.36 points; it can be bought back for 69.5, a result of +£201.40 after all eight commissions. The writer holds instead. On Friday morning the EDSP auction sets 10,300: the put spread settles 100 points against the writer, the other three options expire worthless, and the result is −£96.80, with no closing commission. A 42-point difference between Thursday's close and Friday's settlement moved the result by −£298.20. On Thursday afternoon the position's delta was £7.41 a point and its gamma −£0.339 per 10 points: in the last day almost all of the risk is the settlement price itself.
The worked plan's conventions on this condor, in pounds
None of these is a rule, and each moves money between outcomes rather than creating it. Size is the lever the conventions do not touch: on the library's sizing ladder, which caps a spread's maximum loss at 2% of an account, one of these condors fits an account of £79,840 (sizing framework).
Greeks from entry to expiry week
The entry column is close to neutral: a delta of −£0.39 a point, theta of £12.89 a day paid for with −£86.22 of vega. The third column is the one that matters for a writer who holds into expiry week with the index on a short strike: theta has risen to £25.43 a day, but gamma has doubled from the entry level to −£0.088 per 10 points, so a 95-point fall that barely registered at entry now changes the mark by −£370.71 (see above). The one-standard-deviation columns show the condor's delta turning against the move in either direction, which is what being short volatility means. Units and position Greeks are explained on the Greeks page.
Settlement on the EDSP, not the last close
An ESX option settles against the exchange delivery settlement price, set by an intraday auction on the morning of the third Friday; trading in the expiring series stops after about 10:15 London time, and the cash moves the next business day (how the EDSP is set). Three consequences follow for a condor. It cannot be assigned early: the options are European, so nothing happens before expiry however deep a short strike goes. It never delivers shares, so there is no stamp duty and no weekend with an unwanted position. And the last price that matters is not Thursday's close but a Friday-morning auction, which Branch D shows can differ from it by enough to turn a profit into a loss. The expiry-day timetable in UK time, including US options for comparison, is on the assignment page.
The same shape on a single UK share behaves differently. On BP (1,000 shares, American, physically settled; 60 days from 17 August 2026 at the 530p model level, IV 26%), the builder's 470/500/560/590 condor fills at 3.00p, 9.00p, 12.00p and 5.25p for 12.75p, £127.50, with a maximum loss of £178.10. Either short option can be assigned early, an assigned 500 put makes the writer a buyer of 1,000 shares with £25.00 of stamp duty to pay (who pays SDRT), and a spread with one leg assigned has to be unwound in the market (one leg assigned). A bid-ask of 10% of each leg's price, half paid each way, costs £29.25, 22.9% of the credit.
UK tax: four computations, and how the end decides the tax year
Each written option is one disposal dated when it was written, adjusted by any buy-back (s148) or, if it settles in cash against the writer, redated to the settlement (s144A(2)); each wing is a disposal when sold, when it lapses or when it settles (the rules). A closed or expired condor is therefore four computations. Figures assume the £3,000 annual exempt amount is used by other gains; tax is at 18% or 24%.
On the 2026/27 SA108, Branch D goes in "Other property, assets and gains": 4 disposals in box 14, proceeds of £1,605.00 in box 15 (the two premiums received), costs of £1,701.80 in box 16 (the settlement paid, the two wings and the commissions), a gain of £693.30 in box 17 and losses of £790.10 in box 19 (SA108 boxes).
The trap particular to a cash-settled condor comes when it is opened before 5 April and ends after it. Take the same condor opened on Tuesday 2 March 2027 for the 16 April 2027 expiry (also 45 days), with the index and every volatility assumed the same as on 1 September so that the prices are identical, and suppose the EDSP is 10,100, below the lower wing:
Left to settle, the position ends at −£1,596.80 in total, yet 2026/27 shows only the winning call's £693.30 gain: £124.79 of tax at 18% or £166.39 at 24%, while the loss arrives in 2027/28 and cannot be carried back. It is relief deferred, and lost only if never used. Bought back on the Thursday, the put's loss joins its March grant and sits in 2026/27 against the call's gain. The totals differ a little (−£1,596.80 against −£1,450.20) because the bought put still had time value to sell on the Thursday; the tax years differ because of how the written put ends. HMRC gives no worked example of a cash-settled written option across 5 April; the reading follows from s144A(2) and is worked for a single FTSE put in Example 5 on the tax page. The rules: written options, bought options, counting computations, across 5 April. Options cannot be held in an ISA (wrappers).
Costs in pounds, and the smaller FTSE contracts
Crossing the bid-ask costs 3.3 times the round-trip commission; the quoted width of a four-leg FTSE package is the figure to check before the commission. IBKR UK also offers a tiered rate for UK index options, £0.60 a contract plus exchange and clearing fees (checked 26 September 2026). Two smaller or shorter FTSE contracts are listed on ICE, and each changes the structure rather than just its size:
What systematic condor writing has returned: Cboe's CNDR index
Cboe publishes a benchmark that writes an iron condor every month on the S&P 500: the Cboe S&P 500 Iron Condor Index (CNDR) sells a one-month SPX put and call near 20 delta, buys a put and a call near 5 delta, holds each position to its monthly settlement and keeps a Treasury bill account of ten times the maximum loss, so that one month's worst case is about a tenth of the index. Its value includes the interest on those Treasury bills and assumes trades at the mid-price, with no commission or bid-ask. Cboe launched the index on 3 August 2015; values before that date are back-tested by Cboe.
Since launch the index has returned 0.98% a year including its Treasury bill interest, spent more than seven years below its September 2018 peak, and gained in about two months out of three. The back-test before launch looks far better, which is the usual gap between a rule fitted to history and the same rule run forward. The index is a US benchmark with a different underlying, a monthly roll and no costs, so it is evidence about the strategy, not about this page's condor; it does not show systematic writing to be a reliable income.