Uncovered (Naked) Short Call
An uncovered short call writes a call option without owning the shares and without buying a higher call. The worked example writes BP's October 580 call for 7.00p, £70.00 on one 1,000-share ICE contract. The most it can make is that premium, £68.60 after the opening commission, if BP finishes at or below 580p. The most it can lose has no limit: every penny above 587.00p costs £10, and a 40% rise by expiry would cost 22.2 times the premium. Against the bear call spread, it gives up a ceiling on that loss. It is built for a view that a share will not rise far, in an account that can meet a requirement that rises with the price.
This page assumes the reader knows the covered call (the same call with the shares behind it) and the bear call spread (the same call with a higher call bought). The short straddle and short strangle carry the same uncapped call, plus a written put. BP stands in as a model underlying, priced at an illustrative 530p against its actual close of 519.6p on 17 August 2026 (price data: Yahoo Finance); nothing here is a view on BP. Every figure is modelled: inputs and method.
One BP October 580 call, written without the shares
The model puts a 19.8% probability on BP finishing above 580p and 16.8% above the 587.00p breakeven (risk-neutral, lognormal, IV 26%), and a 39.4% probability that it trades at 580p at some point before expiry. Those numbers describe how often the call is a problem, not how large the problem is; the tables below do that.
BP options trade on ICE Futures Europe in the standard 1,000-share size only (BP is not one of the 22 UK names with a 100-share mini; contract sizes). They are American, so the call can be exercised against the writer on any business day, and exercise delivers shares. IBKR lists a short naked call at its Options Level 4, the highest of its four levels, and it needs a margin account: IBKR's cash accounts accept covered calls and fully funded written puts but not an uncovered call (account types and permissions). IBKR's portfolio margin is an optional upgrade for accounts with at least USD 110,000 of net liquidation value; the ordinary strategy-based requirement below applies without it.
Open this worked example in the strategy builder (the fill, no dividend in the October life).
Payoff: £68.60 at best, and no floor
The profit is a thin shelf: £70.00 anywhere at or below 580p. The chart's dashed and dotted lines sit below that shelf everywhere, because before expiry the call still has time value; buying it back early always costs more than the expiry line suggests. The comparison line is the 580/600 bear call spread, which gives up £40.00 of the premium and stops losing at 600p.
The tail in 10% steps
A chart has to stop somewhere; the loss does not. The table follows BP up in steps and sets three things against each other: the uncovered call's result at expiry, the same view held as the 580/600 bear call spread, and the requirement a broker would ask for if BP jumped to that price on the first day.
Each 10% rise beyond the strike adds £530 of loss, and the spread column shows what a ceiling is worth: from a 20% rise on, the spread's loss is fixed at £172.80 while the uncovered loss keeps growing. The last column matters before expiry. A requirement that is £630.00 at 530p is £3,141.54 at 742p and £6,955.64 at 1,060p, so the cash an account needs to hold the position grows at the same time as the loss, which is the mechanism behind forced closures (the margin spiral).
Stress and margin: the requirement rises with the share price
The requirement uses the Cboe and FINRA strategy-based formula for an uncovered equity call: the option's value plus 20% of the share price, less any amount by which the call is out of the money, with a floor of the value plus 10% of the share price. At entry that is £630.00 (the floor would be £600.00), of which £560.00 is cash beyond the premium received. A UK broker sets its own requirement for ICE options, and the preview on its order ticket is the number that counts (uncovered margin formulas).
Volatility is stepped up on the way up in this table, the opposite of an index, because on a single share the upside is where a bid or a squeeze sits; on a cash bid it collapses to 10%, since the price is then pinned near the offer. The down rows are worth at most +£68.43. The up rows show the requirement rising 242% two standard deviations higher and 5.0 times on the bid, while the marked loss reaches 22.7 times the premium. It would take 23.1 expiries at the full £68.60 to earn back one such morning. A single contract written from an account below £4,726.69 would be under its requirement on the bid morning. At ten contracts the premium after commission is £686.00, the bid loss −£15,870.43 and the requirement £31,396.43: size multiplies the premium and the tail by the same number. Level 3 explains the stress method and lists the historical gaps that make such rows plausible (gap history).
BP from 17 August to 16 October 2026
Branch A: half the premium on Tuesday 8 September
BP has slipped to 512p with 38 days left, implied volatility 25%. The call is worth 1.29p; bought back on the tick at 1.25p, inside the 3.50p that marks half the premium, so the worked plan's half-credit convention closes it.
Holding to expiry would have added at most £13.90, the 1.25p buy-back and its commission (+£68.60 on a lapse against +£54.70). From that day the model put a 6.2% probability on BP finishing above 580p and 4.6% above the breakeven (risk-neutral, lognormal, IV 25%): the last £13.90 came with the whole of the tail still attached.
Branch B: tested on Tuesday 1 September, and the stop
BP has risen to 575p with 45 days left and implied volatility has risen to 28%. The call is worth 21.45p and would be bought back at 21.50p: past the 21.00p at which the worked plan's two-times-premium stop sits (a £140.00 marked loss). The position is short about 503 share-equivalents, and the requirement has risen 109%, to £1,314.46.
The roll pays nothing: moving the strike 20p further away and a month further out costs more than it collects, and the new call would carry BP's 30 October results and 12 November ex-date. Under a credit-only rolling habit there is no roll here; the rolling page sets out when that habit traps (close or roll).
Branch C: a hypothetical bid
The same hypothetical bid for the model underlying that the bear call spread page follows through the Takeover Code timetable (takeover-morning table): no offer for BP exists or is implied. On Tuesday 1 September a possible-offer announcement lifts BP 20% to 636p and implied volatility to 40%; on Tuesday 29 September, the 28th day, either a firm cash offer arrives at 742p with volatility at 10%, or the bidder walks away and BP returns to 520p with volatility back at 26%.
Branch D: held to 16 October at 742p and assigned
If the call is left open into expiry with BP at 742p, it is exercised and the writer must deliver 1,000 BP shares at 580p, £5,800.00, without owning any. Buying them in the market costs £7,420.00 plus £37.10 of SDRT, because the writer is now the buyer of shares; the holder who exercised pays SDRT of £29.00 on the strike consideration (who pays SDRT). With the premium, the opening commission and the commission on assignment, the result is −£1,589.90, against −£1,552.80 for buying the call back at its intrinsic value on the last day: the delivery route costs the stamp duty more. If the shares are not bought at once, the broker borrows them for the account at a fee or buys them in, and any dividend paid while the short is open is owed to the lender (assigned without the shares).
Greeks: quiet in the middle, violent at the strike
With BP unchanged the call fades: by 9 October it is short only 7 share-equivalents and £69.58 of the £70.00 premium is in hand. That quiet column is the common path and the misleading one. Put BP on the 580p strike instead and gamma grows as expiry approaches: each 10p rise adds 75.0 share-equivalents to the short with 45 days left, 110.1 with 21 and 190.9 with 7, while the writer is short about 515 share-equivalents. The last week at the strike pays 2.4 times the theta of the 45-day point (+£6.24 a day against +£2.63) for 2.5 times the gamma. Delta is not the probability of assignment either (delta is not a probability).
Squeezes and bids: Volkswagen 2008, GameStop 2021, Hargreaves Lansdown 2024
The tail table is arithmetic; these three episodes show the jumps are not hypothetical. In each, the share price jumped within days or weeks, and a call written on it before the jump would have lost many times its premium. The causes differed, which is the point: a disclosed holding, a wave of buying, a cash bid.
Where a UK reader can look. Three public sources show some of this risk before a trade, though none shows all of it:
None of these measures a squeeze in advance, and the Volkswagen holding was cash-settled options, not shares. The bear call spread page times a hypothetical bid through the Takeover Code's deadlines in more detail.
BP's 12 November ex-date and the December 580 call
The October call expires before BP's third-quarter results on Friday 30 October and its ex-dividend date on Thursday 12 November (bp financial calendar 2026). A December version carries both. Priced on 17 August with the modelled 6.39p dividend (the second-quarter 8.66 US cents at the model exchange rate; BP sets the actual figure with its results), the December 580 call is worth 14.29p, of which 0.22p is the value of being able to exercise before the ex-date; without the dividend it would be 16.08p.
A holder with stamp duty relief (an options intermediary) gains by exercising from about 634p; a private holder, who also pays 0.5% SDRT on the 580p strike, only from about 664p (early exercise before an ex-date). The holder on the other side of a written call is often an intermediary, so the lower figure is the one that matters to the writer. For an uncovered writer the consequence is sharper than for a spread writer: assigned on the evening of 11 November, the account is short 1,000 BP shares on the ex-date and owes the lender the dividend, £63.90, on top of buying or borrowing the shares. The bear call spread page works the same test on the December 560 call.
A FTSE 100 call instead: what the index removes and what it keeps
The index version removes the company-specific jumps and the delivery, and keeps the unlimited loss: a 10% rise in the FTSE 100 by expiry would cost £4,806.70 on one contract, because each point is worth £10 on a far larger position. FTSE 100 contract details are on the FTSE 100 options page.
Closing, stopping or rolling the call: the conventions in pounds
Each habit trades one outcome in pounds for another, and none is advice; where each comes from, and what evidence exists for it, is on the methods page. A stop is a decision made at a price, not a guarantee of that price: Branch C opens straight through it. The library's sizing framework measures a position like this by its stress rows rather than a maximum loss, since it has none (sizing framework).
UK tax: a buy-back and an assignment land in different years
Writing the call is a disposal on its grant date, 17 August 2026: +£68.60 in 2026/27 if it lapses. A buy-back is folded into the grant (TCGA 1992 s148), so the gain falls or turns into a loss in the grant's own year. An assignment replaces the grant computation: the premium joins the proceeds of the shares delivered (s144(2)(a); HMRC CG12313), and any tax already charged on the grant is set off (CG12317), here a short sale matched with the shares bought to close it (assignment table). Figures take the £3,000 annual exempt amount as used elsewhere; 18% or 24% turns on the basic rate band left.
The trap is the date, not the size. Closed in the market, the loss belongs to 2026/27 and can be set against that year's gains. Left to be assigned, almost the same loss moves to 2027/28, where it cannot be carried back against a 2026/27 gain, only forward (across 5 April); the disposal date on exercise is the exercise date, not the settlement date (TCGA 1992 s28(2)). Options cannot sit in an ISA, and we could not find a UK SIPP administrator that accepts them (checked 26 September 2026): wrappers compared.
Costs, and what the premium has to cover
The costs are small against the premium and irrelevant against the tail; the real cost comparison is with the wing. The 600 call costs 4.00p, 57% of the premium, and caps the loss at £172.80. The bear call spread page prices that choice for the 560 call.