Bull Put Spread
A bull put spread writes one put and buys a cheaper put at a lower strike with the same expiry. The worked example writes the HSBC October 1,450 put and buys the 1,400 put for a net credit of 12.00p, £120.00 on one 1,000-share ICE contract. The most it can lose is the 50p gap less the credit, £380.00, or £382.80 with the opening commission; a delivery of shares costs more. What it gives up is every penny above the credit: a rally to 1,700p still pays £120.00. It is built for the view that a share holds above a level, not that it rises. HSBC is a model underlying here; this is not a view on HSBC.
This page assumes the reader has met the cash-secured put and the long put. Its mirror above the market is the bear call spread, worked on BP. Every figure is modelled, not quoted: inputs and method.
Two HSBC October puts, 50p apart
HSBC options trade on ICE Futures Europe in two sizes: the standard contract over 1,000 shares (code HSB), used throughout this page, and a mini over 100 shares (8HC), costed below. Both are American style, so they can be exercised on any business day, and both deliver shares. ICE sets strike prices itself and publishes no interval table, so 1,450p and 1,400p are plausible strikes rather than a quoted chain. Contract sizes for every UK and US product are on the options basics page.
Two structural facts decide what this position is. The long put must expire no earlier than the short one; otherwise the short put stands alone for the gap. And a spread needs a margin account: IBKR lists a short put spread at its Options Level 3, and in a cash account a written put must be backed by the whole strike value, £14,500 here (account types and permissions). In a margin account the broker holds the width, £500, of which £120.00 is the credit already received (spread margin).
Open this worked example in the strategy builder. The builder solves each leg's volatility from the two fills and draws the curves with European values, so its lines sit a fraction of a penny from this page's American ones.
Profit and loss at 16 October, and on the way there
The whole result lives in 50p of HSBC, 3.3% of the entry price. Above 1,450p both puts expire worthless and the credit is kept. Below 1,400p the loss stops growing, because each penny the short put gains the long put gains too. In between, the short put is in the money and the long put is not, and that is the zone in which a spread left to expire becomes 1,000 HSBC shares (assigned between the strikes). The dotted line shows why an early move barely registers: on the entry day a fall to 1,300p would show a model loss of £295.74, not £380.00, because 60 days of time value still sit in both legs.
Choosing the width and the short strike on one HSBC chain
Two choices set the trade, and they answer different questions. The short strike decides how likely the credit is to be kept. The width decides how much is lost when it is not. The first table holds the short put at 1,450p and moves the long put down; the second holds the width at 50p and moves both strikes. Same inputs as the worked example; fills on the tick.
Widening from 50p to 150p lifts the credit from 12.00p to 23.25p but more than triples the loss, and the model probability of losing all of it falls from 18.8% to 5.3%. There is no free lunch in either direction: the credit per pound at risk falls as the chance of the worst case falls. The long put is the price of turning a £14,240 obligation into a £380.00 one; the cash-secured put page covers the version without it.
Moving the short strike from 1,400p up to 1,500p raises the credit from 7.25p to 18.00p and the chance of the maximum loss from 11% to 29%. The library's worked plan writes the put about 5% out of the money at a delta near −0.26; that is a teaching choice, and the methods page says where such conventions come from and what evidence exists for them (none that shows an edge).
Volatility is the third input. The model sheet sets no 12-month implied-volatility range for HSBC, so this page quotes no IV rank (IV rank explained). Priced at 20% instead of 25%, the same spread would be worth 9.20p; at 30%, 14.22p. That is the −£5.05 of vega per point at work: a richer market pays a larger credit for the same 50p of risk, and a quieter one a smaller credit.
Credit or debit: parity on the 1,450 and 1,400 strikes
The same view can be bought instead of sold: buy the 1,400 call and write the 1,450 call. On these inputs that bull call spread costs 37.80p. Put-call parity ties the two together. With European values the put spread is worth 11.89p, and 11.89p plus 37.80p is 49.69p: the 50p width discounted at 3.75% for 60 days, a gap of 0.31p, which is interest. The American puts add 0.15p of early-exercise value, which is why the put spread's model value is 12.04p. Even the volatility exposure matches: vega is −£5.05 per volatility point on the put spread and −£4.95 on the call spread.
So at the same strikes the choice is not about direction or risk. It is about cash flow (receive £120.00, or pay £378.01 at model value), which leg can be exercised early, and which side of the money each strike sits. On early exercise the two differ in this window. The written 1,450 put can be assigned early once it is so deep in the money that the interest on the strike until expiry outweighs its remaining time value: that possibility is the 0.15p above, and Branch D shows the case. The written 1,450 call in the debit version is worth exercising early only just before an ex-date, and none falls before 16 October. Parity itself is on the Greeks and parity page; the debit version has its own page.
What a 73% model probability of profit is worth: the expected-value table
On the model, the spread makes money 73.3% of the time and keeps the whole credit 70.5% of the time. Those figures are model probabilities (risk-neutral, lognormal, IV 25%), not forecasts, and on their own they mislead. The table splits every possible finishing price into four zones and weights the result in each zone by its probability.
Read across the rows and the high hit rate stops looking like an edge. Winning outcomes average +£117.75 and occur 73.3% of the time; losing outcomes average −£322.01 and occur 26.7% of the time. Weighted together they come to +£0.34 a contract, which is zero apart from interest, the early-exercise value of the American puts and the rounding of the fills: the model that sets the price is the model that sets the odds, so the premium is fair by construction. After the £120.00 credit meets real costs (the £2.80 opening commission and £10.00 of half-spread on a 1.00p quote) the expected result is −£12.46. With no drift instead of the risk-neutral one, the probability of profit reads 71.3%; it changes the headline, not the arithmetic.
That leaves one place an edge could come from: the market's implied volatility turning out higher than the volatility HSBC actually delivers. The FTSE 100 record on that gap since 2000 is summarised on the implied-volatility page. The short put's delta of −0.26 is also not its chance of finishing in the money, which is 29.5% here (delta is not a probability).
The HSBC trade from 17 August to 16 October 2026
Branch A: the half-credit convention fires on Thursday 10 September
HSBC has risen to 1,560p with 36 days left and implied volatility has eased to 24%. The model values the puts at 9.14p and 3.38p; on the tick the spread buys back for 9.25p less 3.50p, 5.75p, inside the 6.00p that marks half the credit. A writer following the library's half-credit convention closes both legs as one order.
Holding on would have added up to £60.30 more if HSBC stayed at or above 1,450p to 16 October, a model probability of 83.7% from that day (risk-neutral, lognormal, IV 24%), and lost up to £382.80 with a probability of 7.4%. The convention swaps the last £60.30 for 36 fewer days of exposure.
Branch B: HSBC at 1,460p with 21 days left, and the November roll
By Friday 25 September HSBC has fallen to 1,460p, 10p above the short strike, and implied volatility has risen to 27%. The model puts are worth 31.54p and 13.70p, so the spread marks 17.75p on the tick (31.50p less 13.75p); the position is long about 196 share-equivalents. The library's 21-day convention closes or rolls here.
The roll's price rests on two stated assumptions: November priced at the same 27% (HSBC's third-quarter results on Tuesday 27 October fall inside the November life, and a results date usually adds event premium), and a dividend of US$0.10, 7.38p at the model exchange rate, going ex on Thursday 5 November. The dividend alone adds 0.90p to the November spread. The roll lowers the maximum loss, but only by carrying the position through results and the ex-date for 35 more days (56 days in all from 25 September to 20 November). The mechanics and the tax of a roll are on the rolling page. Holding the October spread instead would have kept £117.20 if HSBC finished at or above 1,450p (model probability 54.3%), ended at −£132.80 at 1,425p before the costs of delivery (−£206.70 once assigned, as Branch C shows), and lost £382.80 or more below 1,400p (probability 25.8%).
Branch C: HSBC at 1,425p on 16 October, assigned between the strikes
The short put finishes 25p in the money and is exercised; the long put finishes out of the money and lapses. The writer buys 1,000 HSBC shares at 1,450p. The spread's expiry result, −£130.00, becomes −£206.70 once the assignment's costs arrive: £76.70 of opening and assignment commission and stamp duty.
Branch D: HSBC at 1,380p on 16 October, close or deliver
Both puts finish in the money. Closed before 16:30 at the full 50p width, the result is −£385.60 after four commissions (−£405.60 with the half-spread). Left to expire, the short put is assigned (1,000 shares bought at 1,450p) and the long put is exercised (the same shares sold at 1,400p): −£458.10. The difference is the £72.50 of SDRT on the purchase, 19.1% of the £380.00 maximum loss, which the textbook formula leaves out. Defined risk is only defined when the spread is closed or settles in cash (defined only at expiry).
Assignment can also arrive before expiry. On Tuesday 6 October, with 10 days left, HSBC at 1,300p and implied volatility at 30%, the 1,450 put is worth exactly its 150.00p of intrinsic value (0.00p of time value), while ten days' interest on 1,450p is 1.49p a share: a holder gains by exercising at once. The long 1,400 put still carries 1.22p of time value, so selling it beats exercising it. The routes out of a spread with one leg assigned are worked on the assignment page, and why deep in-the-money puts go early is explained there too.
Greeks from entry to expiry week
Two columns carry the lesson. In expiry week with HSBC sitting on the short strike, the spread earns +£6.19 a day but carries −31.8 of gamma: a further 3% fall from 1,450p costs £164.08 with seven days left against £60.66 with 60 days left, both at 25% volatility. And after a one-standard-deviation fall on the first day the position sits almost on the long strike and every sign has turned: gamma +0.2, vega +£1.06, theta −£0.36 a day. Near its maximum loss the spread is long volatility, because only a large move can still help it. Units and position Greeks are explained on the Greeks page.
Assigned between the strikes: the spread, the cash-secured put and the wheel
Many put writers would be content to own the shares at the strike. A spread changes when that happens. The table compares this spread with writing the same 1,450 put on its own for the same 26.00p, at four expiry prices. Results include the opening and assignment commissions and the £72.50 of SDRT wherever shares are bought; shares still held are marked at the expiry price.
The spread delivers shares only inside the 50p window between the strikes. Below 1,400p its long put sells them straight back out, and the SDRT on the purchase is still paid. The cash-secured put delivers at any price below 1,450p, keeps the extra 14.00p the spread spent on protection, and ties up £14,500 instead of £500. A writer who wants HSBC at about 1,432p and would then write calls against it is running the wheel, which starts from the cash-secured put; the spread is the version for someone who wants the premium and a floor, not the shares.
Whichever structure, assignment follows the ICE timetable. A holder can exercise an American HSBC option until 18:30 London time on any business day; ICE Clear Europe allocates exercises among writers at random; the writer learns the next morning; and shares settle two business days after exercise. At expiry, in-the-money options are exercised automatically. The hour-by-hour expiry day in UK time is on the assignment page.
Put skew, and who buys the puts
On HSBC's model chain, priced at one volatility for every strike, the put spread pays more than a call spread at the same delta. A call credit spread written at the same delta, 1,650/1,700 with a short-call delta of 0.26, would collect 9.42p; this put spread, short delta −0.26, collects 12.04p. The reason is the gap between delta and probability: a put's chance of finishing in the money is larger than its delta suggests and a call's is smaller.
Real chains are not flat. The one fitted surface in this library, the FTSE 100 at 10,750 with 60 days left, prices lower strikes at higher volatility, and that changes the comparison.
Skew makes the written put dear, but it makes the bought put dearer still, because the long leg sits further down the slope. Per point of width the put spread then collects less than the call spread. What skew prices is the fatter left tail: the surface gives large falls more probability and moderate falls less, so the written put is paid for tail risk that the bought put then hands back (skew, explained). The usual explanation for that shape is who is buying: holders of UK share portfolios buy index puts as protection, and writers charge for carrying crash risk. So the buyer on the other side of a written index put may well be a hedger of that kind rather than someone betting on a fall. For single UK shares we could not find a free public source of ICE skew data (checked 27 September 2026); HSBC's legs here are priced flat at 25%, so a live chain with a steeper price on the 1,400 put would show a smaller credit than 12.00p.
The same view on the index settles in cash. A FTSE 100 October 10,550/10,500 put spread on the model surface (IVs 14.75% and 14.94%; model values 161.01 and 145.82 points) fills for 15.0 points, £150.00, against a maximum loss of £353.40 with commission. It is European, so it cannot be assigned early, and it delivers no shares, so no SDRT. The contracts and the settlement price are on the FTSE 100 options page.
HSBC's early-November dividend and the later expiries
The October series ends before two HSBC dates: third-quarter results on Tuesday 27 October and a third interim dividend whose record date HSBC's timetable puts in "early November", with the ex-date one business day earlier. The November and December series span both. The table prices the same 1,450/1,400 spread on each, entered on 17 August, with the dividend modelled at US$0.10 (7.38p) going ex on Thursday 5 November, an assumed date: HSBC has not yet announced it, and its 2025 third interim went ex on Thursday 6 November 2025 (HSBC announcement, 28 October 2025).
A dividend takes value out of the share on the ex-date, so every put across it is worth more. The 1,450 put gains 2.01p in November (38.03p without the dividend) and 2.19p in December (46.15p without it). The spread gains much less, 0.56p in each month on the unrounded model values (14.33p to 14.89p in November; 15.49p to 16.04p in December), because the long put gains too. For assignment the effect runs the other way from calls: the holder of an in-the-money put does better to wait until the share has gone ex and fallen, so early assignment of a short put is less likely just before an ex-date and more likely after it (early put assignment). The October series collects less but earns +£0.91 a day against +£0.36 for December, and it is over before results.
Managing the position: each convention in pounds
The branches above apply the library's teaching conventions. None is a rule, and each has a price. The table sets what each one did in this example against what holding would have given.
Size is the other lever. The library's sizing ladder caps a spread at 2% of an account in maximum loss, so on £50,000 two standard contracts (£382.80 each) fit and three do not; the method is on the sizing page. A common habit is to roll only for a net credit; a debit roll adds its cost to the maximum loss, and the rolling page works both.
UK tax: two computations, and the stamp duty the put writer pays
Writing the 1,450 put is a disposal on 17 August 2026, the grant; buying it back is folded into that grant (TCGA 1992 s148); a lapse leaves it standing. The long put is a separate asset until it is sold, lapses or is exercised. Figures assume the £3,000 annual exempt amount is used by other gains; tax at 18% or 24% depends on the unused basic rate band.
The trap particular to this structure is assignment: it makes the put writer the buyer of 1,000 shares, so the 0.5% SDRT, £72.50, is the writer's (who pays SDRT), and it joins the shares' cost rather than counting as a loss now. Timing is the other. Opened in March 2027 with both puts lapsing in April, the £258.60 grant gain would fall in 2026/27 (£46.55 at 18%, £62.06 at 24%) and the £141.40 lapse loss in 2027/28: relief deferred, and lost only if never used (Example 13 on the tax page). The rules: written options, counting computations, SA108 boxes; options cannot sit in an ISA (wrappers).
Costs in pounds, and the 100-share mini
Commission and spread together take £25.60, 21.3% of the credit, on one standard contract closed before expiry. The mini makes the position ten times smaller, but the commission does not shrink: at the £1.70 fixed rate used as a placeholder (no mini rate is published), ten minis cost £68.00 in commission against £5.60 for one standard contract. The mini is listed on ICE; whether a given broker offers it and shows a two-way price could not be confirmed (checked 26 September 2026). Brokers, fees and access are compared on the broker page.