Backspread
Prerequisite strategies: the broken wing butterfly, the long strangle (what unfinanced convexity costs) and the bull put spread (what a written strike does to your margin). Clear the Level 3 gate and position sizing first.
Why this structure exists
Convexity is expensive: a long put or a long strangle is long gamma and long vega, and theta bills you daily for holding it. A backspread makes the market fund most of that bill. Sell one option near the money, buy two further out in the same expiry, and the fat short premium covers nearly all of the two cheaper longs — £151 below, where the outright put costs £965.
The price is geometric rather than monetary. Between the strikes you sit on the wrong side of a written option, and the deepest point of that trough falls exactly on your long strike. The payoff is a plateau on one side, a cliff into the trough, then an unbounded run beyond it: a large cheap move, bought by accepting a small expensive drift.
The nearest simpler alternative is the long strangle — same character, no written leg, no margin, no trough. Why not just do that? Because it costs six times more, so it needs the move bigger and sooner. Take the backspread only when you can name, out loud, the level you must not be sitting at on expiry morning.
Construction
| Leg | Buy / Sell | Quantity | Strike rule | Expiry rule | Target delta | Price |
|---|---|---|---|---|---|---|
| Short leg | SELL (credit) | 1 contract | 1–2% out of the money | 45–75 days, monthly | 0.35–0.45 | 178.5 pts = +£1,785 |
| Long legs | BUY (debit) | 2 contracts | Width ≥ 2.5% of the index below the short | Same expiry | 0.20–0.28 each | 96.5 pts = −£1,930 |
| NET | Net debit | 1×2 ratio | 10,600 / 10,300 | 60 days, 16 Oct 2026 | −0.08 | £145 + £6 costs |
Max loss = width × multiplier + net debit + costs. Breakeven = long strike − (max loss ÷ multiplier). Max profit = (breakeven − settlement) × multiplier, bounded only by the index reaching zero. The mark cannot be worse than the expiry maximum either, because a European vertical spread can never exceed its width: over 1,000–20,000 points, 1–60 days and vol shifts of −5 to +40, the worst mark modelled was −£2,694.
The gap between the lines is what time and long vega are worth, and it is widest across the trough: at 60 days the screen shows a small loss where expiry shows −£3,151. You trade the solid line and watch the dashed one.
Entry criteria
| Gate | Rule | Reason |
|---|---|---|
| Implied volatility | IV rank below 30 | Net long £100.73 of vega a point |
| Skew | Net debit ≤ 8% of the width in cash (£240 here) | Steep put skew makes the longs dear; break the rule and the skew has priced your move |
| Term structure | Front month not inverted against the next | Backwardation means the event is priced |
| Days to expiry | 45–75, monthly series only | Below 45 the trough beats the move |
| Strikes | Short 0.35–0.45 delta; longs 0.20–0.28; width ≥ 2.5% of the index | Narrowing the width multiplies the debit: 10,600/10,500 costs £1,135 |
| Liquidity | Quote ≤ 3 index points a leg; open interest ≥ 250 | £45 in and £45 out against a £145 debit |
| Underlying | FTSE 100 index options — European, cash settled | No assignment, no delivery, no SDRT |
| Event calendar | MPC and CPI inside the window are a reason to be there; nothing within two days of expiry | The expiring series stops trading shortly after 10:15 on the third Friday |
Do not enter if: IV rank is above 40; the debit exceeds 8% of the width; the term structure is inverted; you cannot fund £6,151 against one structure; or you cannot name the level you must not be sitting at in 60 days.
Greeks at entry and how they evolve
| Greek | At entry (60 DTE, 10,750) | 30 DTE, unchanged | 7 DTE, unchanged | +1 SD (11,360) at 30 DTE | −1 SD (10,140) at 30 DTE |
|---|---|---|---|---|---|
| Delta | −£0.82/pt | +£0.32/pt | +£1.91/pt | +£0.19/pt | −£3.99/pt |
| Gamma | +£0.30 | +£0.19 | −£0.96 | −£0.07 | +£1.11 |
| Theta | −£14.49/day | −£11.59/day | +£29.24/day | +£1.70/day | −£40.24/day |
| Vega | +£100.73/pt | +£36.20/pt | −£29.69/pt | −£8.80/pt | +£152.13/pt |
Modelled at 14% at-the-money implied volatility with a downside skew, 4% rates, 3.5% dividend yield. Gamma is the change in position delta, in pounds per index point, for a 100-point rise in the index; a positive figure means long gamma.
Vega decides this trade early; gamma decides it late. At entry you own £100.73 a volatility point against a £151 outlay, so it can profit on a volatility spike with the index still. Now read the 7-day column: with the index above the short strike, gamma has gone negative, theta positive, vega negative. The longs are dust and the written option is alive — the structure has become a short put wearing a backspread's name. That flip is what the time stop exists for.
FTSE 100 at 10,750, and you want a cheap hedge for the autumn
The FTSE 100 closed at 10,750.11 on Friday 14 August 2026; this models 10,750. The ICE Futures Europe FTSE 100 option is £10 per index point, European style and cash settled at the EDSP struck in the LSE intra-day auction, tick 0.5 points (£5). The October series stops trading shortly after 10:15 on Friday 16 October 2026, 60 days away.
The trade: sell 1 × October 10,600 put at 178.5, buy 2 × October 10,300 put at 96.5.
Branch A — 9,900 with 21 days left, volatility 6 points higher: the structure marks at +£1,978. ACTION: the +£1,575 target has fired. Close.
Branch B — 10,450 with 21 days left. The mark is only −£655, which is the problem: it looks survivable.
Branch C — 10,800 on 16 October. All three puts settle worthless at the EDSP. You lose £151, and the £1,785 received in August is still a chargeable gain for 2026/27.
Branch D — the same trade written for a credit. Move the longs down to 10,200 at 78.0 and you collect 22.5 points, £219 after costs. The risk is different, not merely cheaper: above 10,600 you now make £219 instead of losing £151, but the width widens to 400 points, so the trough deepens to −£3,781 and the breakeven falls to 9,821.9 — an 8.6% fall, not 7.1%. A credit backspread pays you to be wrong in the flat case and charges more for being wrong in the middle. The tax point is identical either way.
Credit or debit is a different trade, not a cheaper one. Put the two side by side at expiry. Above 10,600 — the single most likely outcome — the debit version loses £151 and the credit version makes £219, a £370 swing in favour of the credit, which is exactly why the credit shape is the one that gets oversold. Now read the other two columns. The credit version's trough is £630 deeper, and its breakeven sits 163 points further down the chart, so the move that pays it has to be bigger and it has fewer days to arrive. You have not bought a better structure; you have sold 163 points of the thing you came for in exchange for being paid to be wrong in the flat case. Choose the debit version when you want the move and the credit version when you are being paid to wait for it — and size both off the trough, £3,151 or £3,781, never off the ticket price.
This is a single, hand-picked, favourable scenario. It is not a typical or expected outcome. Premiums are modelled from a Black-Scholes surface at 14% at-the-money volatility with a downside skew, not taken from a live chain, and real FTSE 100 quotes away from the front month are wider. Past performance is not a reliable indicator of future results.
Management and adjustment
| Trigger | Diagnosis | Action | Do NOT do this |
|---|---|---|---|
| Short strike tested, 10,560 at 40 DTE (−£198) | Working as designed; the trough is ahead | Nothing. Diarise the 21-day check | Do not bank a small profit the moment it ticks green |
| Inside 10,250–10,600 with ≤21 DTE (−£655 at 10,450) | In the cone, no time to leave it | CLOSE. The case where the right action is to close, not adjust | Do not roll the short strike down: at 10,470 with 21 days left, moving the 10,600 short to 10,400 costs £999 of debit and cuts the trough from £3,151 to £2,150 — £1,001 of tail bought for £999 |
| Through the long strike, 10,150 at 30 DTE, IV +3 (+£787) | Thesis confirmed; delta −£3.92 a point | Sell one long, run a 1×1, or hold for the target | Do not buy a third long |
| IV up ≥ 4 points, index unchanged, 45 DTE (+£120) | The vega gift: paid without being right | Take it if IV rank is now above 50 | Do not add vega at the higher level |
| IV down ≥ 3 points, index unchanged (−£394) | The reason you were here has gone | Close | Do not wait for volatility to come back |
| Rally to 11,000 with 14 DTE (−£253) | A write-off: the two longs are worth £36 and it still costs about £100 to close | Close anyway; release the £3,000 of margin | Do not leave it on "because it is only £151" |
| Overnight gap below the long strike (9,900 at 21 DTE, IV +6) | Max-profit territory; delta −£6.22 a point | Close, or sell one long and hold the 1×1 | Do not carry the 1×2 into an EDSP auction you cannot trade through |
ROLL WHEN: only if the thesis is intact with more than 30 days left, and then roll the whole structure to the next monthly as two closes and two opens — a new trade with a new £3,151 on the ticket. ROLL TO: the same width and delta band, never a narrower width to feel safer. DO NOT ROLL: never pay a net debit to defend a losing structure — row two shows £999 buying £1,001. The tier-wide rule that a credit position is never rolled for a net debit applies literally to Branch D. Where defence stops: inside the cone with under 21 days, every adjustment costs what it saves. Closing is the adjustment.
Exit rules
If all four are silent, do nothing.
Margin and broker reality
A margin account is mandatory — a cash account cannot hold a written index put — and spread permission is the minimum. Send the 1×2 as one combination order; if the platform rejects the ratio, do not leg in during the session.
Initial and maintenance. Interactive Brokers decomposes this into a short put spread (long one 10,300, short one 10,600) plus one extra long put paid in full. Its rule-based methodology charges a short put spread the strike differential times the multiplier and does not net off the credit received — the rule that makes an SPX 3000/3100 spread a $10,000 requirement. Here 300 × £10 = £3,000 initial and £3,000 maintenance, plus £965 for the second long.
After a 2 SD adverse move the requirement is still £3,000, because strike differentials do not move when the index does. That is the benefit of defined risk, and the last page in this tier where the sentence is true: the ratio spread is this trade run backwards, and its requirement moves with everything.
Your equity moves, and that is what liquidates you. IBKR liquidates when net liquidation value falls below maintenance. Funded with £4,000, that arrives with seven days left and the index near 10,535 — a 2.0% fall — at a mark loss of £1,014. Fund it with £6,151, the requirement plus the full £3,151, and nothing can force you out.
Portfolio margin buys less than you think. IBKR requires USD 110,000 of net liquidation value to upgrade an existing account, and restricts margin-increasing trades below USD 100,000. A risk-based model prices this from the worst revaluation in its scan range: on a ±6% index scan with a ±4.5-point volatility shift, about £435 at 30 days, £1,478 at seven and £2,511 with one day left. It lends rope early and takes it back in the week you need it. The chain away from front-month round strikes also trades wide, so the bid-ask, not the commission, is the real cost.
Stress test
| Scenario | Index | Mark at 30 DTE | Held to expiry | Requirement | Delta |
|---|---|---|---|---|---|
| −20% gap | 8,600 (IV +18) | +£13,946 | +£13,849 | £3,000 | −£9.43/pt |
| −2 SD | 9,530 (IV +7) | +£5,117 | +£4,549 | £3,000 | −£7.97/pt |
| −1 SD | 10,140 (IV +3) | +£828 | −£1,551 | £3,000 | −£4.01/pt |
| Unchanged | 10,750 | −£417 | −£151 | £3,000 | +£0.32/pt |
| +1 SD | 11,360 (IV −1) | −£183 | −£151 | £3,000 | +£0.17/pt |
| +2 SD | 11,970 (IV −2) | −£151 | −£151 | £3,000 | £0.00/pt |
One standard deviation over the 60 days is 610 points, 5.68%. Volatility responses are modelled, not observed.
The row to fear is not the gap. It is −1 SD. An ordinary 5.7% sell-off shows +£828 at 30 days and −£1,551 if you sit in it to expiry, because long vega flatters the mark and expiry does not care about vega. That £2,379 gap is where backspread traders lose money: they read the mark as confirmation and hold.
On a 1987 or March 2020-scale gap the structure returns 4.4 times its maximum loss with no margin call, because the requirement never moved. The way it has actually hurt people is the opposite: a slow 3–5% drift that parks the index in the trough while the volatility expansion meant to rescue it never arrives.
(width × multiplier) + net debit ≤ 2% of account equity. At £3,151 that means £157,550 — the honest answer to whether most UK retail accounts should hold a FTSE 100 backspread at all.What to trade instead
Simpler, from the tier below: an outright long put or a long strangle — same long-gamma, long-vega exposure, no written leg, no margin, no trough, no grant-date tax event. The trade-off is price: the 10,300 put alone costs £965 against £151. With a target rather than a tail view, a bear put spread says it for less.
More precise, from inside this tier: the broken wing butterfly buys the same skew asymmetry with the wide wing where you want it rather than where the ratio forces it. The ratio spread is this page run backwards — sell two, buy one — converting a defined trough into an undefined tail.
Risk statement
Listed options are complex instruments, and this tier assumes a margin account, uncovered-option permission and a written risk process. This is educational material about mechanics and UK tax treatment, not a recommendation to trade the FTSE 100 or anything else, and it takes no account of your circumstances. Every premium here is modelled, not quoted.