Short Straddle
A short straddle writes a call and a put at the same strike and expiry: here the FTSE 100 October 10,750 options, sold for 420.0 points, £4,200.00 on one £10-a-point contract. That credit is the most it can make, kept whole only at a settlement of exactly 10,750. The loss has no limit above the strike and grows until the index reaches zero below it. The writer gives up any view on direction: the straddle profits only if the index ends within 420.0 points of 10,750 on 16 October. It is designed for a period expected to be calmer than the options have priced. The FTSE 100 is a model underlying, not a view on it.
This page assumes the reader has worked through the short strangle, this page's mirror with the strikes moved apart, and the iron butterfly, the same straddle with wings bought. Where the two short-volatility pages share mechanics, this one quotes the strangle's figures and links rather than repeating them. Every number is modelled on the library's model sheet.
Writing the 10,750 call and put on Tuesday 1 September
The index options are ICE's FTSE 100 contract (ESX): £10 a point, European, so neither leg can be exercised early, and cash-settled on the exchange delivery settlement price, so nothing is delivered and no stamp duty or SDRT arises (FTSE 100 contracts; contract sizes). The FTSE 100 closed at 10,789.3 on 1 September 2026 (price data: Yahoo Finance); the library prices every FTSE example at 10,750 so that pages can be compared, and quotes each strike's implied volatility from one model surface. ESX strikes are listed at intervals of 25 to 200 points depending on expiry and distance from the money, so the live chain decides which strikes exist. IBKR lists the short straddle under its Options Level 4 permission, which needs a margin account (permissions and account types).
Because both options share one strike, one leg is always in the money at expiry unless the settlement price is exactly 10,750: the question is never whether a leg pays out, only how much.
Open this worked example in the strategy builder (both fills, 45 days; the builder solves each leg's volatility from its fill and gives probabilities at 14%, without skew).
Payoff: one peak at 10,750, and £10 a point off it in both directions
The table has one row where the whole credit survives, and it is a single number. Every point the settlement lands away from 10,750 costs £10, so the profit shrinks steadily from the peak and turns into a loss beyond 10,330 and 11,170. The credit is 0.79 of the model's one-standard-deviation move over 45 days, 528.4 points at 14%, so the breakevens lie inside the one-SD range of 10,234.3 to 11,291.6. At 9,910 and 11,590 the loss equals the credit. The two curves under the payoff show how little of the credit is earned early: with the index back at 10,750 on Friday 25 September and volatility unchanged, the position shows £1,325.24, and £2,874.76, 68% of the credit, still depends on the last three weeks.
Stress and margin: seven instant moves on 1 September
Each row moves the index instantly on the entry day. Each strike keeps its own implied volatility when the index moves (the library's sticky-strike convention, explained on the implied volatility page), and every row then adds the stated parallel change in volatility: a fall is assumed to lift implied volatility and a rise to lower it. The requirement column uses the Cboe strategy-based formula for a short put and call on a broad index: for each leg, its value plus 15% of the index less any out-of-the-money amount, with a floor of 10% of the index for a call and 10% of the exercise price for a put; then the larger leg plus the value of the other (uncovered margin). It is shown for illustration: a UK broker margins ICE options by its own method, and the figure in its order preview is the one that applies.
The initial requirement uses the premiums received: the call leg's 1,827.0 points (its 214.5 plus 15% of 10,750) is larger than the put leg's 1,818.0, so the requirement is that plus the put's 205.5, £20,325.00, of which the credit itself covers £4,200.00 and the account £16,125.00. From there the requirement moves with the marks. In every adverse row it rises while the mark falls: after the 20% fall it is £34,505.36, 70% above entry, against a mark of −£17,405.36, a loss of 4.1 times the credit, so an account that began with less than £51,910.72 would be below its requirement with one contract. How brokers act at that point, and the index's own record of one- and two-day falls, are on the Level 3 page (the margin spiral; gap history; stress-test method). Listed options carry no negative-balance protection (why).
The 1 SD rows teach the other half. At 10,234.3 the position is marked at −£2,113.00, more than twice its −£956.60 result if the index settled there, because 45 days of time value and two extra points of volatility sit on top of the put's intrinsic value. The up-move marks less than the down-move at the same distance only because volatility is assumed to fall on a rally.
Where a stop at one times the credit fires
A common convention for short straddles closes the position when the marked loss reaches the credit: £4,200.00 here, the straddle bought back at 840 points. Solving the model for that mark puts the stop well beyond the breakevens, not before them. On 1 September with volatility unchanged it fires at about 9,920 (−1.63 SD on the 45-day lognormal scale) or 11,550 (+1.46 SD); with volatility four points higher, at about 9,976 (−1.52 SD) or 11,478 (+1.33 SD). The breakevens sit at −0.81 and +0.78 SD. By Wednesday 16 September, with 30 days left and volatility unchanged, the lower level is about 9,908, next to 9,910, where the expiry loss itself equals the credit.
So on an at-the-money straddle, a one-times stop and the expiry payoff agree closely: once the index is 800 points away there is little time value left to separate the mark from the payoff. The stop does not protect the breakeven; it caps the loss at roughly the credit when the move arrives as a drift. In the model (lognormal, IV 14%, zero drift) the index touches 9,920 at some point before expiry with probability 10.6% and 11,550 with probability 13.9%, while it finishes below 10,330 with probability 21.6% and above 11,170 with 21.1% (the tables on this page use risk-neutral probabilities from the skew surface, which put the model probability of finishing between the breakevens at 57.8%). A gap goes straight through any stop: the 10% row above is marked at −£7,300.76 the moment it opens. The strangle page's stop menu shows how different the same convention is when the strikes are apart.
Why the gamma sits at 10,750: one expiry, three widths
The decision a straddle writer makes is how far apart to put the two strikes. The table prices three answers on the same chain and the same surface: the straddle, a 25-delta strangle and the 16-delta strangle worked on the short strangle page.
Three readings. First, the straddle carries 1.63 times the 16-delta strangle's gamma, and carries it at the price the index is at now; the strangle's gamma only grows as the index approaches one of its strikes. Second, per unit of gamma the three earn almost exactly the same time decay: each needs the index to move less than about 78.5 points a day for theta to cover the gamma (78.4 and 78.9 for the strangles), because at one level of implied volatility theta is the rent paid for gamma (theta as rent). Widening the strikes changes the size of the bet on movement, not its price. Third, the gap column barely changes with width: an instant 10% fall marks the 16-delta strangle at −£5,932.84 against the straddle's −£7,300.76. Matching the straddle's credit would take 4.6 of those strangles, which together would lose about £27,533.72 in the same fall.
Implied against realised: the straddle hedged each evening
Selling the straddle at 14% is selling a price for movement. The Greeks turn that into a daily figure: at entry the position earns £46.27 a day of theta and loses about half its gamma times the square of the day's move, so the two cancel at a move of about 78.5 points a day, roughly 10,750 × 14% ÷ √365 (78.8). For a writer who hedges the delta, a day-one move of 50 points is worth £27.79 and a move of 150 points −£121.31. The implied volatility page explains why implied volatility has on average sat above realised for the FTSE 100, and why an average hides the days that decide a short position (implied against realised).
The chart follows the same straddle, hedged at every close so that its delta is back to zero, along stylised 45-day paths. Each path moves the index the same distance every trading day, alternately down and up, with each move sized to the calendar days since the previous close, so the path realises exactly the stated volatility over the 45 days (33 moves across 34 trading days). The hedge is modelled at the index level each evening, before costs and before the small difference between a future's price and the index.
The two right-hand columns answer different questions. Left alone, the straddle is a bet on distance: the three alternating paths all end within about 100 points of 10,750 and all make money, however much the index churned on the way, while the steady slide, which barely moved on any one day, ends 900 points away at −£4,800.00. Hedged each evening, it becomes a bet on realised volatility: the quiet path earns £1,853.54, the path that realises exactly the 14% sold ends at −£47.95, the busy path ends at −£2,462.93, and the slide, with realised volatility of 4.3%, earns £944.76. The fourth path is the pattern that sits behind the long-run averages: the hedged position had built up £521.44 by the close before the fall and stood at −£3,187.27 a day later, because a 6% move is far larger than any hedge adjusted once a day can follow.
Hedging the delta with FTSE 100 futures
ICE lists a full-size and a mini future on the index. The FTSE 100 Index Future is £10 a point, with a 0.5-point tick worth £5, four quarterly expiries and cash settlement on the EDSP; the Mini FTSE 100 Index Future is £1 a point, with a 0.5-point tick worth £0.50 and two quarterly expiries (ICE product pages, checked 27 September 2026). The straddle's delta in pounds a point converts straight into contracts: divide by 10 for the full-size future, or read it as a number of minis. Both are quarterly, so an October straddle is hedged with the September future until it expires on 18 September and with December after that; a future's price includes the carry between now and its expiry, so the hedge ratio is close to, but not exactly, the index delta.
The full-size future is too coarse for one straddle until the index has moved about two standard deviations; the mini can match the delta to within half a pound a point. Hedging does not change what a gap costs, because the index is already past the hedge when it opens, and it adds a futures position with its own margin, commissions and computations for tax. What it does change is the question the position answers, as the path table shows.
Portfolio margin: more straddles per pound, the same loss per straddle
IBKR offers portfolio margin as an optional upgrade for accounts with options approval and at least USD 110,000 of net liquidation value; if the account falls below USD 100,000, Reg T margin applies instead. It is not what permits a short straddle: that is the Options Level 4 permission. A risk-based requirement sets the margin to the worst modelled loss across a range of moves instead of a fixed percentage of the index. US portfolio-margin rules test a high-capitalisation, broad-based US index option position across moves from 8% down to 6% up (Cboe's portfolio-margin rule, SEC Release 34-50886, 2004). How IBKR's risk-based margin treats an ICE FTSE option is IBKR's own calculation; the US range is used below only to show the method.
Across that range, with volatility unchanged, this straddle's worst loss is at the 8% fall: −£4,469.72 (at the 6% rise, −£2,922.34; with volatility four points higher, −£4,859.01). Against the strategy-based £20,323.94 that is less than a quarter as much. The difference is capacity, not risk. USD 110,000 is about £81,127 at the model rate of $1.356 per £1. At that size the strategy-based requirement fits 3 straddles and the risk-based illustration up to 18. An instant 10% fall costs −£7,300.76 a contract under either method: £21,902.28 on 3 contracts, 27% of the account, or £131,413.68 on 18, 162% of it. The sizing framework is on the position sizing page.
Worked example: four ways the October straddle can end
Branch A: the 25% target, Monday 21 September. Twenty days in, the index is back at 10,750 and volatility is unchanged. The call is worth 159.34 and the put 154.20; bought back at 159.5 and 154.0, the straddle costs 313.5 points, £3,135.00, below the 315.0-point level at which 25% of the credit is banked. The worked plan closes: 25.4% of the credit, £1,058.20 after four commissions, £998.20 after crossing the illustrative spread both ways. Holding to expiry instead would have paid £4,196.60 with a settlement exactly at 10,750, £2,696.60 at 10,900, and −£803.40 at 11,250.
Branch B: tested at 10,300, Wednesday 16 September. The index has fallen 450 points with 30 days left and volatility two points higher. The put is worth 489.32 and the call 46.61; the straddle is marked at −£1,159.33, its delta has become +£6.29 a point (long the index without having bought it) and the requirement is £20,809.33. The one-times stop is not close. The convention this plan uses is to roll the untested side toward the money: buy back the 10,750 call at 46.5 (IV 16.00%) and write the 10,300 call at 211.0 (model 211.00, IV 17.71%), a further 164.5 points, £1,645.00. The call strike is now below the put strike: the position is an inverted strangle, and between 10,300 and 10,750 both legs will settle in the money for a combined 450 points. The arithmetic that decides whether that is acceptable is simple: total credit collected, 584.5 points (£5,845.00), against the inversion width, 450 points (£4,500). Because the credit is larger, every settlement from 10,300 to 10,750 keeps £1,345.00, and the breakevens move to 10,165.5 and 10,884.5. Had the credit been smaller than the width, no settlement price could have made the inverted position profitable.
The roll halves the delta, to +£3.00 a point, and raises the gamma from −£0.113 to −£0.132, because the new call sits at the money. It pays for a narrower but flat-topped profit zone by giving up the straddle's recovery case: if the index had climbed back to 10,750, the unchanged straddle would have kept £4,200.00 and the inverted one keeps £1,345.00. The mechanics of rolling, and when closing is the simpler choice, are on the rolling page.
Branch C: a 10% gap, Tuesday 8 September. The index opens at 9,675 with 38 days left and volatility ten points higher. The straddle is marked at −£7,079.59, 1.69 times the credit in loss, and the requirement is £25,792.09. The one-times stop has been passed before any order could fill, so the worked plan closes at the opening prices: the call at 32.0 (model 32.09) and the put at 1,096.0 (model 1,095.87), 1,128.0 points or £11,280.00, a realised −£7,086.80 after commissions and −£7,146.80 after the spread. Holding would have left the result to the settlement: −£6,553.40 if the index stayed at 9,675, −£3.40 if it recovered to the 10,330 breakeven, and more loss below 9,675.
Branch D: held to settlement, EDSP 10,870 on 16 October. The put expires worthless and the call settles in cash at 120 points, £1,200.00. The result is £4,200.00 less £1,200.00 less the opening commission: £2,996.60. There is no closing trade, no delivery and no SDRT. This branch ignores the 21-day time stop to show the settlement itself.
The exits in this plan, and what each one trades away
This page's worked plan takes profit at 25% of the credit rather than the 50% the library uses for most credit positions (teaching conventions). The reason is the shape of an at-the-money straddle's decay. With the index and volatility unchanged, its value falls to 75% of the credit on Monday 21 September, with 25 days left, but only reaches 50% on Monday 5 October, with 11 days left, well inside the last three weeks, where its gamma has grown from −£0.150 to −£0.305. On a straddle, a 50% target would mostly fire after a 21-day time stop, not before it.
The time stop itself sits on Friday 25 September. By then £1,325.24 is marked as profit and £2,874.76 is left to earn; gamma has risen to −£0.221 and theta to £68.18 a day, and a 100-point fall costs £87.27, against £41.88 at entry (a 100-point rise costs more in each case, because the position starts slightly short the index). By Friday 9 October, a week before expiry, the same 100 points cost £175.49. Holding on keeps the £2,874.76 if the index stays near 10,750; the time stop exchanges that for stepping off the steepest part of the gamma curve. Neither number says which is better, and the balance shifts with the size of the position. Being flat before trading in the expiring series stops shortly after 10:15 on 16 October also avoids leaving the result to the EDSP auction (how the EDSP is set).
Greeks from 1 September to expiry week
Read along the first three columns, the index never moves and the position grows more dangerous as it grows more profitable: gamma more than doubles between 1 September and 9 October, and theta rises with it. The move that theta pays for barely changes: about 78.5 points a day at entry, 78.6 at 21 days and 78.7 at 7 days. What changes is the size of both sides of that exchange. Read across the last two columns, one standard deviation turns a delta-neutral position into a directional one: its delta equals −0.73 full-size futures after a rise and +0.59 after a fall (position Greeks).
The same trade on BP: American legs and the 12 November ex-date
On a UK share the straddle becomes a different instrument. BP has only the standard ICE contract, 1,000 shares, American style and settled by delivering shares; it is not one of the 22 names with a 100-share mini. Take the BP November 530 straddle written on Monday 17 August 2026 for expiry on Friday 20 November (95 days), with BP at the library's illustrative 530p (it closed at 519.6p that day), IV 26%, and BP's third-quarter ex-dividend date of 12 November inside the options' life, modelled with an assumed 6.39p dividend; BP's third-quarter results on 30 October also fall inside it (event premium). Priced on the binomial tree, the call is worth 29.34p and the put 28.48p. The call's European value is 27.13p: the 2.20p difference is what the right to exercise before the ex-date is worth to its holder. The put's European value, 28.32p, is almost the same as its American one, because a put holder gains from the ex-date and has little reason to exercise before it. Filled at 29.25p and 28.50p, the straddle collects 57.75p, £577.50, with breakevens at 472.25p and 587.75p. BP is a model underlying; this is not a view on BP.
The two legs face opposite dividend logic. A call holder exercises early to collect the dividend when it is worth more than the call's remaining time value: an options intermediary with stamp duty relief gains from about 542p, and a private holder, who would pay 2.65p a share of SDRT on the 530p strike, from about 551p. Because the holder on the other side is often an intermediary, the lower figure is where assignment of the written call becomes likely; assigned on the evening of 11 November, the writer would be short 1,000 BP shares the next morning without the dividend (early exercise before an ex-date; assigned without the shares). A put holder gains from the drop in the share price on the ex-date, so has every reason to wait: at 460p on 11 November the put is worth 76.33p, 6.33p more than exercising it. Once the dividend has gone, a deep put is worth exercising: on Friday 13 November, with 7 days left, the model values the put at exactly its intrinsic value at about 491p and below (at 480p, 50.00p, with 0.00p of time value), because the interest on 530p for a week (0.38p a share) is worth more than the option's remaining time value (early put assignment). An assigned put writer buys 1,000 shares at 530p and, as the buyer, pays £26.50 of SDRT (who pays SDRT).
Held to expiry, one BP leg will almost certainly finish in the money, so unless the position is closed first its writer ends up either buying 1,000 shares or short 1,000 of them (pin risk and automatic exercise). That, and a credit of £577.50 against the FTSE version's £4,200.00, is why the index is this page's main example.
UK tax: two grants, and a settlement that can move a year
Each written option is one disposal, dated when it is written (TCGA 1992 s144(1)). A buy-back is folded into that grant as a cost (s148). A leg settled in cash against the writer is different: grant and settlement become one transaction, dated at settlement (s144A(2); CG12321). Figures assume the £3,000 annual exempt amount is used by other gains; 18% or 24% depends on the basic rate band left.
The last row is the straddle's own trap. Written in March with the same premiums and settled after 5 April, the lapsing leg's gain stays in 2026/27, costing £369.59 at 18% or £492.79 at 24%, while the settled leg, which paid out £7,500, becomes a 2027/28 loss that cannot be carried back. Overall the position made −£3,303.40. Buying the call back before settlement, even after 5 April, would have folded its cost into the March grant (s148) and kept the whole result in 2026/27 (positions across 5 April; counting computations). No option can be held in an ISA, and we found no SIPP administrator that permits them (wrappers, checked 26 September 2026).
What commission and the spread cost, and where the contract trades
On a £4,200.00 credit the costs of trading are small; on the strangle page the same pounds are a much larger share. ESX options trade from 08:00 to 16:50 London time. Some spread-betting and CFD firms quote their own options on the index; the three routes differ in tax as well as in counterparty (listed options, spread bets and CFDs). Brokers that offer ICE options are mapped on the broker page.