Long Butterfly
Prerequisite strategies: you must have traded the long call, the covered call and at least one vertical — the bull call spread is the one this is built from — so that you have granted an option, been assigned once and priced a two-leg spread from a live chain. Clear the Level 2 gate first. Next: the iron butterfly, then the broken wing butterfly at Level 3.
Why this structure exists
Every other structure in this tier expresses a direction. A butterfly expresses a number. It is the cheapest listed way to say “I think this share is 2,100p in a month” and be paid for the precision rather than the direction, and the payoff ratio it offers — 3.41 : 1 below — is the market pricing how unlikely it thinks you are to be right about a specific number.
Mechanically it is two verticals joined at the middle strike: a 2,000/2,100 bull call spread bought, and a 2,100/2,200 bear call spread sold. The first pays you for the move up to your target; the second pays you for the move not continuing past it, and its credit almost funds the first. That is why the whole tent costs £220.00 where the vertical alone costs £355.00.
Why not just buy that vertical? At £355.00 you pay 35.5% of the width, break even at 2,036.06p and stay long everything above 2,100p. The butterfly hands the upside back and cuts the cost to 22.0% of the width and the breakeven to 2,023.12p. If your view is “up”, take the vertical: it wins across an entire half of the price line. If it is genuinely “up, to about here, and then stop” — a share into a level, a bid rumour with a plausible ceiling, a re-rating with an obvious target — the butterfly is the only structure that charges you for one opinion instead of two. The rest of this page is about the two ways it goes wrong: readers hold it for the maximum, which is paid at one price on one afternoon, and readers hold it into delivery, which on a UK contract is physical.
Construction
| Leg | Buy / Sell | Quantity | Strike rule | Expiry rule | Target delta | Price |
|---|---|---|---|---|---|---|
| Lower wing | BUY (debit) | 1 contract = 1,000 shares (ICE UK); 100 (US) | One strike interval below the body | 20–45 days; one expiry for all three strikes | 0.50–0.60 | 53.5p = −£535.00 |
| Body | SELL (credit) | 2 contracts, same expiry | Exactly at your target price | Identical to both wings | 0.20–0.30 | 18.0p × 2 = +£360.00 |
| Upper wing | BUY (debit) | 1 contract, same expiry | One interval above the body, equal spacing | Identical to the lower wing | 0.05–0.12 | 4.5p = −£45.00 |
| NET | Net debit | 4 contracts, 1 : 2 : 1 | 2,000 / 2,100 / 2,200, GSK modelled at 2,000p | 30 days | +0.120 | 22.0p = −£220.00 |
All calls or all puts, never mixed; one underlying, one expiry, equal spacing, middle quantity exactly twice the wings. Break the 1 : 2 : 1 ratio and you have built a ratio spread with undefined risk, which is Level 3 and you are not permitted to hold it. Four inequalities before the order goes in:
Formulas: max loss = debit × contract size + opening commission. Max profit = (wing width − debit) × contract size − both commissions. Breakevens = lower wing + debit + round-trip costs per share, and upper wing − debit − those costs. Modelled at entry: a 34.3% chance of finishing anywhere inside the profit zone, a 56.0% chance of the full £225.60 loss, and a 4.5% chance of finishing within 10p of the body — which is where the £768.80 lives.
The tent is an expiry object and the two dashed curves are the trade you actually own. With 30 days left the position can be worth at most +£55.61, and that only at 2,096p; with five days left, at most +£350.73, at 2,099p. The £768.80 peak requires GSK to be sitting on 2,100p at 16:30 on one specific Friday, and there is a 4.5% chance of it landing within 10p of that. Everything on this page follows from the gap between the solid line and the dashed ones.
Entry criteria
| Gate | Rule | Reason |
|---|---|---|
| IV rank / IV percentile | IVR 50 or above. Below 30, buy a debit vertical for the same view instead | A butterfly is short vega, so dear premium makes the tent cheap. Modelled, this structure costs 25.14p at 14% implied volatility and 16.08p at 36% — the payoff ratio moves from 2.87 : 1 to 4.98 : 1 for an identical picture |
| Days to expiry | 20–45, never 60+. This structure does not use the tier’s 21-day time stop | The tent is a fixed band; the distribution around it is not. The 30-day tent’s 153.76p zone is 1.22 SD wide, but the 60-day tent’s 161.76p zone covers only 0.91 SD, so the chance of finishing inside falls from 34.3% to 29.6%. The longer tent is cheaper — 18.0p against 22.0p — and that discount is exactly what it buys |
| Strike selection | Body at your written target, inside +1 SD; wings one equal interval either side; profit zone ≥ 1 SD wide | The body is the forecast. If you cannot name the number, this is the wrong structure |
| Cost discipline | Net debit ≤ 33% of one wing’s width | Fixes the payoff ratio at 3 : 1 or better before you look at a chart. £220.00 of £1,000 is 22.0% |
| Liquidity | Spread ≤ 10% of mid on each of the four contracts; open interest ≥ 100 at all three strikes; one combo order | 10% on each leg costs £94.00 round-trip: 42.7% of the debit and 8.4× the commission |
| Underlying | A liquid FTSE 100 name with a three-strike ladder around your target; the index itself if the chain is thin | ICE UK single-stock series are physically delivered over 1,000 shares; the index option is cash-settled at £10 a point |
| Event calendar | No results, capital markets day, index review or ex-dividend date inside the window | A butterfly bets that nothing happens except your number. GSK’s next ex-date is 12 November 2026, which is why this is the September series |
Do not enter if: IV rank is below 30 — the tent costs a third more, the ratio falls under 3 : 1, and you are short vega into the expansion a low IV rank makes likely; IV rank is high because of a binary event in the window, which is volatility telling you the share will not sit on your number; the debit exceeds a third of the wing width; any of the four contracts fails the liquidity screen, or the platform will not take the structure as one order; you are in a cash account, because the order is rejected before it reaches the exchange; or you cannot write down the price and the date first.
Debit or credit: the same tent, two structures
A long call butterfly and an iron butterfly at the same three strikes are the same picture, one bought and one sold. Put–call parity keeps them within a rounding of each other: the £220.00 debit and the £775.00 credit sum to 99.5p of the 100p width, the difference being the present-value discount plus the 0.5p tick. IV rank does not pick between them the way it picks between a bull call and a bull put spread. Execution and tax do.
| Long call butterfly (debit) — this page | Iron butterfly (credit) | |
|---|---|---|
| Legs on GSK at 2,000p | Buy 2,000 call 53.5p, sell 2 × 2,100 call 18.0p, buy 2,200 call 4.5p | Sell 2,100 call 18.0p and 2,100 put 111.0p, buy 2,200 call 4.5p and 2,000 put 47.0p |
| Cash at entry | Pay £220.00 | Receive £775.00 |
| Max profit / max loss | £768.80 / £225.60 | £763.80 / £236.20 |
| Breakevens | 2,023.12p and 2,176.88p | 2,023.62p and 2,176.38p |
| Buying power used | £220.00 (the debit) | £225.00 (width − credit) |
| Early-assignment exposure | Two short calls, in the money only above 2,100p | Two short legs, one of which is always in the money |
| Day-one taxable gain | £360.00 (two 2,100 calls granted) | £1,290.00 (2,100 call and put granted) — 166% of the credit |
For a UK reader the last two rows decide it. The iron butterfly always has one short leg in the money, so on an American-style physically delivered ICE series it is exposed to early assignment from day one, and it books £1,290.00 of chargeable gain on the day it is opened against a £763.80 maximum profit. The all-call version grants two out-of-the-money options for £360.00 and cannot be assigned until GSK is above 2,100p. Same tent, a third of the tax event: on a UK contract build the butterfly from calls above spot or puts below it, and reach for the iron version only where the underlying is cash-settled.
Greeks at entry and how they evolve
| Greek (net, per structure) | Entry: 30 DTE, 2,000p | 15 DTE, unchanged | 5 DTE, unchanged | +1 SD (2,126p) at 15 DTE | −1 SD (1,874p) at 15 DTE |
|---|---|---|---|---|---|
| Delta | +0.120 (120 shares) | +0.242 | +0.451 | −0.099 | +0.068 |
| Gamma | −0.00069 | −0.00026 | +0.00501 | −0.00317 | +0.00138 |
| Theta | +£1.60/day | +£0.17/day | −£14.25/day | +£9.76/day | −£3.34/day |
| Vega | −£5.01/pt | −£0.92/pt | +£6.04/pt | −£12.94/pt | +£4.37/pt |
Black–Scholes at 22% implied volatility, 4% rates and no dividend inside the window — GSK’s Q2 2026 shares went ex on 13 August 2026 and the Q3 ex-date is 12 November 2026, both outside the September series. Per 1,000-share contract. At the body the numbers are far larger: with GSK on 2,100p the structure marks 38.90p at 15 DTE with theta at +£10.85 a day and vega at −£14.71 a point, and 58.18p at 5 DTE with theta at +£35.38 a day.
Vega decides the entry; gamma decides the exit. Read the vega row across: the structure is short volatility everywhere near the body and turns long volatility only if GSK falls away. Vega changes sign between 1,955p and 1,960p — a 2.1% fall — and theta flips with it between 1,960p and 1,965p. That is where the position changes character. Above 1,960p you own a decaying, premium-selling structure that wants nothing to happen; below it, a long-volatility lottery ticket that needs a rally to survive. Readers who watch the share drift the wrong way rarely notice they are now in the opposite trade.
Then gamma. Mildly negative for the first fortnight — −0.00069 at entry, −0.00026 at 15 days — it turns violent at the body near expiry: −0.00369 with 15 days left, −0.01207 with five, −0.03298 with one. On the final day a 20p move away from the body costs £60.43 of mark, and delta swings from +0.585 at 2,080p to −0.598 at 2,120p across 40p. Holding a butterfly to expiry is therefore a gamma decision, not a patience one: the last £275 of the £768.80 is collected by standing on a knife edge for one afternoon, in a contract that then physically delivers.
GSK modelled at 2,000p, and you think the autumn takes it to 2,100p and stops
GSK plc traded either side of 2,000p on the LSE in mid-August 2026. The ICE Futures Europe GSK option is quoted in pence per share, one contract confers rights over 1,000 shares, it is American style and physically delivered, the tick is 0.5p (£5.00 a contract), and the September series stops trading at 16:30 London on Friday 18 September 2026. One penny of option price is £10 of contract value. Modelled at 2,000p spot, 22% implied volatility, 4% rates, 30 days, fills rounded to the tick.
The trade, one four-contract combo order on Wednesday 19 August 2026: buy 1 × GSK September 2,000 call at 53.5p, sell 2 × September 2,100 calls at 18.0p, buy 1 × September 2,200 call at 4.5p.
Branch A — GSK 2,100p on 3 September, 15 days left. Dead right, far too early.
Branch B — GSK still 2,100p on Monday 14 September, four days left. Same price, eleven days later.
Branch C — GSK 1,965p on 3 September, 15 days left. The share has gone nowhere.
Branch D — GSK 2,260p on 3 September. Right on direction, wrong on the ceiling: the branch a vertical wins and this structure loses.
Branch E — GSK pinned at 2,100p on 18 September and you let it settle. The branch with the stamp duty and the naked short in it.
Branch E is the whole argument. At 2,100p on the last afternoon the short calls are exactly at the money, so you must decide whether to exercise the 2,000 call before you know whether you have been assigned on two, one or neither. Assigned on both, you are short 1,000 shares over a weekend. On one, you hold 1,000 shares you did not want and paid £100 of stamp duty on. On neither, you have exercised into £21,000 of stock. No version of that can be planned for, which is what pin risk means.
On the FTSE 100 index instead — £10 per index point, European, cash-settled at the EDSP — the same tent carries no assignment and no SDRT, which is the venue to use if you intend to hold one late. On a US chain a contract is 100 shares and the gain is still computed in sterling on each disposal date: a $2.00 butterfly costs $200, £147.58 at GBP/USD 1.3552; close it for $6.00 with the rate at 1.4000 and the proceeds are £428.57. The dollar profit is 200%, but the chargeable gain is £280.99 against the £295.16 an unchanged rate would have given — £14.17 of currency, before the conversion spread, on a trade whose whole edge is a couple of hundred pounds.
This is a single, hand-picked, favourable scenario. It is not a typical or expected outcome. Every premium is modelled from Black–Scholes at the stated inputs and rounded to the exchange tick rather than taken from a live chain, and real ICE UK quotes on a three-strike structure are materially wider than the screen used here. Past performance, including any illustrative example shown here, is not a reliable indicator of future results.
Management and adjustment
| Trigger | Diagnosis | Action | Do NOT do this |
|---|---|---|---|
| Body reached, 15+ days left | Right, early: the mark is 39.0p, 20.7% of the maximum, because the tent has not formed | Hold to the target and let theta work — +£10.85 a day at the body, +£35.38 by 5 DTE | Take the £158.80 out of relief. You paid for the last fortnight; this is the fortnight |
| Stalls below the lower breakeven | The forecast has not happened and there is no time left for it | Close at the 15-day gate for the residual — £154.40 at 1,965p against £0 if it expires there | “Give it a chance”. A butterfly not near the body inside two weeks is a decaying option |
| Trades through the upper wing | Right on direction, wrong on the number; every further penny hurts | Close. At 2,260p with 15 days left it marks £114.40 and is falling towards £0 | Roll the tent up — at 2,100p, 15 DTE that means selling at £390.00 to pay £241.35 for a new trade plus £11.20 on eight contracts |
| IV expands after entry | You are short vega: −£14.71 a point at the body | Nothing. At 2,100p, 15 DTE, 22% to 30% cuts the mark from £390.00 to £295.00 — inside the loss already agreed | Buy a second one “now it is cheaper”. Tents are cheap in a spike because the share will not sit still |
| IV collapses after entry | The best case: at 2,100p, 15 DTE, 22% to 16% lifts the mark from £390.00 to £500.00 | Check the target — a vol collapse fires it days before the calendar suggests | Hold for the peak. The vega gift is realised on the close, nowhere else |
| Ex-dividend date in the window | Above 2,100p with the short calls’ extrinsic below the dividend, assignment is rational for the holder | Assume assignment on both bodies; close the business day before the ex-date | Leave it. You learn from the overnight statement, holding 2,000 delivered shares you did not fund |
| Assigned early on one body | Short 1,000 GSK; the 2,000 call covers it and the 2,200 wing is loose | Exercise the 2,000 call to deliver, or close the lot in one order. Either way it is over | Buy the shares in the market — 0.5% SDRT on the higher price, and you still hold a short call |
| Tempted by the upper wing | The 2,200 call looks like dead money once the share is near the body | Nothing. It is the only thing making this defined risk | Never sell it. At 2,100p, 15 DTE it raises £78.85 and turns the position into a 1×2 ratio spread with unlimited upside risk you cannot hold |
ROLL WHEN — in one case only: the share has arrived at the body with more than 21 days left, your revised target is a different number, and you would open the new butterfly as a fresh trade at today’s prices and IV rank. ROLL TO a new three-strike structure centred on that target, same expiry, as one eight-contract order — then rewrite the maximum loss, because it is now the new debit.
DO NOT ROLL a butterfly out in time and call it one trade. A roll is a close and an open with one ticket over them: you sell convexity you waited a month for and buy a flatter one. The October tent on the same strikes costs 18.0p against September’s 22.0p, and it is cheaper for a reason — September’s 153.76p zone is 1.22 standard deviations wide over 30 days, while October’s 161.76p zone covers only 0.91 over 60, so the chance of finishing inside drops from 34.3% to 29.6%. If you would not open that October butterfly on its own merits, do not acquire it by rolling.
THE CORRECT ACTION IS TO CLOSE, NOT ADJUST, in every case but that one: when the profit target fires, when the 15-day gate fires, when the share trades through either wing, when an ex-dividend date appears in the window, and always before the last trading day. A butterfly has no defence because there is nothing to defend with — all four contracts are load-bearing, and any adjustment that raises cash from one either doubles the risk or removes the definition. It has an exit, and the exit is the strategy.
Exit rules
The last £275 of the £768.80 needs GSK to sit on one number for one afternoon while you are short 0.033 of gamma. If all four rules are silent, do nothing and check the mark against the target tomorrow.
Margin and broker reality
A cash account cannot hold this trade, and that is where most UK first attempts die. A butterfly contains two granted options and a cash account has no mechanism to carry one: Interactive Brokers permits only limited purchase and sale of options in a Cash account, so the order is rejected in the preview rather than at the exchange. You need a Margin account, for which IBKR’s published minimum is USD 2,000 or equivalent. Hargreaves Lansdown, AJ Bell and Trading 212 offer no options at all, so this is an Interactive Brokers or Saxo trade.
What you do not need is more money or uncovered-option permission. Both short 2,100 calls are fully covered — one by the 2,000 call, one by the 2,200 call, same expiry and size — so the initial requirement is the net debit and nothing more: buying power falls by £220.00, maintenance is nil, and it cannot rise whatever GSK does. Two conditions attach. It must be recognised as a butterfly, which means a single four-contract combo order; leg it and you briefly hold two naked short calls, which the platform will refuse or margin at several thousand pounds. And it stops being true the moment a leg is assigned: an overnight assignment on both bodies turns £220.00 of requirement into a £21,000 short stock position and a margin call before you can act, which is why the delivery-avoidance exit is a rule and not a preference.
Then treat the ICE bid-ask as a margin-equivalent cost, because on a four-contract structure it dominates. Ten per cent of mid on each leg is £94.00 round-trip against £11.20 of commission — the market maker charges 8.4 times what the broker does, and 42.7% of the debit. One full 0.5p tick against you on each contract is £20.00, 9.1% of the debit, before the share moves a penny. UK single-stock butterflies are only worth building on the few ICE series with genuine three-strike depth; elsewhere the honest answer is that the structure cannot be built at retail size, and the choice is the cash-settled FTSE 100 index or a US chain with the currency exposure that brings.
close at 2× the debit, or by the Wednesday of expiry week, whichever comes first. The peak is a drawing, not a plan.Portfolio fit
One structure contributes a net delta of only +0.120 — 120 GSK shares, £2,396 of share-equivalent exposure on £225.60 of risk, or £10.62 per pound at risk. That sounds like leverage until you price the alternative: the 2,000/2,100 vertical carries +0.288 of delta on £357.80 of risk, £16.11. A butterfly is not the leveraged version of a spread, it is the less directional one, and its delta is unstable — +0.242 by 15 days, +0.451 by five, then through zero as it crosses the body. What it contributes is volatility: net vega is −£5.01 a point at entry and −£14.71 at the body, so butterflies belong alongside your iron condors, not filed as directional risk because they happen to be debits.
At the 2% rule a £225.60 maximum loss needs at least £11,280 of account. Six of them on £25,000 put £1,353.60 at risk (5.4% of capital) and use £1,320 of buying power (5.3%), well inside the 25% cap this tier works to — and that is the trap. Butterflies are cheap enough that buying-power discipline never bites, so the binding constraints are the two a margin screen does not show: the combined short vega, which turns six positions into one volatility sale; and the hit rate, because six bets at 34.3% make a book whose modal quarter is four small losses and two moderate wins. Size them on how many small losses in a row you can watch, not on what they cost.
What to trade instead
Simpler, from earlier in this tier: the bull call spread. It costs £135 more and caps lower, but it wins across an entire half of the price line instead of one 154p band, carries one short leg rather than two, and generates three CGT events instead of four. Take it whenever your view is a direction rather than a number — which is most of the time.
Alongside, at this tier: the iron butterfly is the same tent sold for a credit, which suits a cash-settled index far better than a delivered UK share; the iron condor widens the body into a plateau, trading the peak for a much larger profit zone; the long straddle is the opposite bet, that the share will not sit still.
More precise, from the tier above: the broken wing butterfly widens one wing so the structure can open for a credit with no risk on the near side — but the asymmetry is a skew trade, and you must understand why the wide wing is priced as it is before you own one.
Risk statement
Listed options are complex instruments and most retail directional positions lose money. This is educational material about mechanics, margin and UK tax treatment, not a recommendation to trade GSK or anything else, and it takes no account of your circumstances. Every premium here is modelled rather than quoted. If your trading becomes frequent enough to raise the investor-versus-trader question, that is one for a qualified adviser.