Hedging a UK portfolio with options: what it costs and what it misses

What an options hedge on a portfolio does, and what it does not

A put option on the FTSE 100 pays out when the index finishes below the strike, so a portfolio that moves with the index can be given a floor for a chosen period, at a price paid in advance. That is the whole idea, and each part of it carries a condition. The floor applies to the index, not to the portfolio. It lasts until one expiry date and then has to be bought again. It comes in contract sizes that rarely match the portfolio. And in the UK the tax treatment depends on where the portfolio and the option are held, which is often not the same account. This page works through each of those conditions in pounds, on two model portfolios.

Portfolio A is £100,000 in a fund that tracks the FTSE 100, so its beta to the index is 1.00 by construction. Portfolio B is £50,000 split £20,000 in the same FTSE 100 tracker and £30,000 in a FTSE 250 tracker. Over the 250 trading days to Monday 17 August 2026 the FTSE 250 had a beta of 0.80 to the FTSE 100 and a correlation of 0.72 (daily returns; price data: Yahoo Finance, checked 28 September 2026), so portfolio B's beta is (20,000 × 1.00 + 30,000 × 0.80) ÷ 50,000 = 0.88. Both portfolios are teaching examples, not a view on what anyone holds; the page describes mechanics and costs and does not say whether hedging suits anyone.

Model inputs. Monday 17 August 2026. FTSE 100 at 10,750, a model level (it closed at 10,720.3 that day; price data: Yahoo Finance). FTSE 100 options (ICE, ESX) at £10 a point, European, cash-settled on the EDSP, priced with Black-Scholes-Merton on the library's surface IV(K) = 14.0% − 0.40 × ln(K ÷ 10,750), sticky-strike; dividend yield 3.05% (FTSE Russell, 28 August 2026); Bank Rate 3.75%. Fills on the 0.5-point tick; commission £1.70 a contract (Interactive Brokers UK fixed rate, checked 26 September 2026). Portfolio values move with the price index; the dividends a tracker pays (about 3.05% a year) are left out of every comparison, and the option prices already allow for them. The FTSE options page sets out the contract, and the methods page the model sheet.

How many contracts: the hedge ratio

One ESX contract moves £10 for every point of the FTSE 100, so at 10,750 it carries £107,500 of index exposure. The number of contracts that matches a portfolio is its value, scaled by its beta, divided by that exposure:

Contracts = portfolio value × beta ÷ (index level × £10). Portfolio A: £100,000 × 1.00 ÷ £107,500 = 0.93. Portfolio B: £50,000 × 0.88 = £44,000 of beta-weighted exposure, ÷ £107,500 = 0.41.

Neither answer is a whole number, and listed options come in whole contracts. One ESX contract on portfolio A covers 1.075 times its exposure, a modest over-hedge. On portfolio B one contract covers 2.44 times, and that changes what the position is. The table sets out the instruments a UK investor can size with, and what each does to the two portfolios.

Each route has a catch. The Mini FTSE daily option sizes almost exactly but cannot hold a quarter's cover, and whether a given broker offers it could not be confirmed (checked 26 September 2026). The spread-bet option sizes to the penny, but it is an over-the-counter contract with the firm, its all-in quote on FTSE futures options was 4 to 8 points at IG (checked 28 September 2026), and its tax is different, as the tax section shows. A short future hedges point for point in both directions: it removes the upside as well as the downside, and it is a futures position with margin, not an option. The listed, CFD and spread-bet comparison sets those products side by side.

One contract on a £50,000 portfolio turns a hedge into a short position

Suppose portfolio B buys one December 10,250 put (the put priced in the next section, £1,776.70) and the FTSE 100 settles 20% lower at 8,600 on 18 December. If portfolio B moves in line with its beta it loses 50,000 × 0.88 × 20% = £8,800. The put pays (10,250 − 8,600) × £10 = £16,500. Portfolio and put together finish +£5,923.30: the "hedged" investor gains from the crash. Below the strike the put earns £10 a point while the portfolio loses about £4.09 a point, so the pair is a bet on a fall of £5.91 a point. A spread-bet put at £4.09 a point on the same premium in points would have cost £725.98 and left portfolio B at −£2,777.48, against −£8,800 unhedged. That is what a matched hedge looks like: a smaller loss, not a profit. Whether portfolio B really moves with its beta is a separate question, taken up under basis risk.

A put, a put spread or a collar on £100,000

Three ways to protect portfolio A to Friday 18 December 2026, 123 days away, each with one ESX contract. All three buy the December 10,250 put, 4.65% below the index. The put spread also sells the 9,250 put, so the cover stops 1,000 points lower in exchange for a smaller cost. The collar sells the 11,100 call instead, giving away the portfolio's gains above 11,100 (3.26% up) in exchange for a premium that pays for the put.

Read across the 8,600 row, a 20% fall. Unhedged, portfolio A is £20,000 down. The put holds the loss to £5,276.70; the put spread, whose cover ran out at 9,250, to £11,273.40; the collar to £3,448.40. Read across the 11,750 row, a 9.3% rise, and the order reverses for the collar: it keeps £2,853.93 against £9,302.33 unhedged. Because one contract covers £107,500 against a £100,000 portfolio, every line slopes by £0.70 a point where an option is in the money: the put and put-spread results improve slightly as the index falls further, and the collar's result drifts down above 11,100.

None of the three is better in general. The put costs the most and keeps the upside. The put spread costs £503.30 less and protects against a moderate fall, not a crash. The collar costs nothing up front (here it brings in £51.60) and pays for the floor with the gains above 11,100. Which trade-off fits is a question about the investor, not the market. The collar page works the same structure on a single share, and the bear put spread page the spread on its own. Open the collar in the strategy builder (the builder holds £107,500 of index, one contract's worth, so its pounds match a portfolio of that size rather than £100,000).

What protection costs per year

The price of a put covers one period. To compare periods, the table shows each put's cost as a share of the £107,500 one contract covers, and that share scaled to a year (cost × 365 ÷ days). The scaled figure is the cost of keeping the same protection in place for twelve months if it could be bought again at the same price each time, which is a comparison device, not a forecast of what future puts will cost.

Two patterns run through the table. At the money, short protection is expensive per year: the October 10,750 put costs 13.35% a year on this measure, the June 2027 put 5.60%, because an at-the-money option loses its time value fastest near expiry and a short-dated put has to be bought again more often. Further from the money the pattern flattens and then turns: the October 9,750 put costs only 1.67% a year on this measure, because in 60 days the index is unlikely to fall 9.3% and the price reflects that. Cheap per year is not the same as cheap for the protection given: that 9,750 put pays nothing on a fall of less than 9.3%.

The table also depends on the level of implied volatility. The model's at-the-money 14% sits near the bottom of the range the 60-day FTSE 100 Implied Volatility Index covered in the year to June 2026, which averaged 17.24. Three points higher, the December 10,250 put costs 241.13 points, £2,411.28 on the model, instead of £1,774.70 (177.47 points; £1,776.70 filled with commission): £636.58 more for the same cover. Protection bought after a fall usually costs more, because implied volatility on the FTSE tends to rise when the index drops; the implied volatility page explains why and what the skew adds.

Keeping cover in place: rolling each quarter

A hedge that is meant to run for a year is a series of puts. At each expiry the old put either pays out or lapses, and a new one is bought, usually at a strike reset to the same distance below the index. The table shows the December roll on three paths: the index 10% lower on 18 December, unchanged, or 10% higher. The new put is the March 2027 series, 91 days, with the strike about 4.65% below the index on the day, priced on the model's sticky-strike surface.

After a fall, the cash from the old put more than pays for the new one, but the new one costs £1,820 rather than £1,345, because a lower strike carries a higher implied volatility on the surface. After a rise, the old put lapses and the new cover is cheaper, £845, and sits higher, protecting part of the gain. That is the case for rolling: the floor follows the index. The cost of it shows over a full programme. At an unchanged index, the December put and then March and June puts cost £4,470.10 with commissions over the 305 days to 18 June 2027, £5,349.46 scaled to a year (4.98% of £107,500). One June 2027 10,250 put bought on 17 August costs £3,581.70, or £4,286.30 a year (3.99%), but its strike stays at 10,250 however far the index rises.

Each put in a programme is its own asset for tax: one that settles in the money is a disposal on its settlement day, one that lapses is an allowable loss in the tax year it lapses, so the results of a year of hedging net off within each tax year. The rolling page covers roll mechanics and costs in general, including a FTSE put spread rolled across 5 April.

Basis risk: when the portfolio is not the FTSE 100

A FTSE 100 put pays on the FTSE 100. Portfolio B's beta of 0.88 says how it tended to move with the index on ordinary days in one past year; it says nothing about a year in which the two indices go different ways. 2022 was such a year.

From 31 December 2021 to 30 December 2022 the FTSE 100 rose from 7,384.5 to 7,451.7, up 0.91%. The FTSE 250 fell from 23,480.8 to 18,853.0, down 19.71%. At the low on 12 October 2022 the FTSE 100 was 7.56% below where it started and the FTSE 250 29.26% below (price indices; price data: Yahoo Finance, checked 28 September 2026). Portfolio B held through 2022 would have finished −£5,730.66, a fall of 11.46%. Its beta predicted a small gain, about £400. A December 2022 FTSE 100 put struck about 5% below the start-of-year level, at 7,000, expired worthless: on the expiry day, Friday 16 December 2022, the index closed at 7,332.1, and portfolio B was down £6,392.52. The hedge worked exactly as written; it was a hedge on the wrong index.

Three sources of basis risk apply to most UK portfolios. Mid and small companies, as in 2022. Single shares: a put on the index pays nothing when one holding falls 30% on its own results while the rest of the market is steady. And dividends: the FTSE 100 is a price index, so a put on it ignores the dividends a tracker collects, which is why the model figures leave them out on both sides. Portfolios that hold overseas shares add currency, which the dollar example below measures.

One holding hedged with 100-share minis

A standard ICE stock option covers 1,000 shares, which is often more than a private investor holds of one company. On 22 large UK names ICE also lists a mini option over 100 shares, quoted in pence with a 0.25p tick; the basics page lists them. HSBC is one (code 8HC). Take a holding of 500 HSBC shares at 1,530p, a model level (HSBC closed at 1,531.0p on 17 August 2026), worth £7,650.

Model inputs. Monday 17 August 2026; HSBC 1,530p; IV 25%; Bank Rate 3.75%; no ex-dividend date before 16 October 2026 in the model (HSBC's third interim dividend goes ex in early November; the date was not confirmed). The October 1,450 put is American, priced on the binomial tree (200 and 201 steps averaged): model 26.04p, filled at 26.00p. Commission £1.70 a mini, the library's placeholder because no mini rate is published.

Each mini put costs 26.00p × 100 shares = £26.00. Five cover the 500 shares for £138.50 with commission, 1.81% of the holding for 60 days; the commission alone is 6.5% of the premium, a real cost on contracts this small. To 16 October the holding can lose at most (1,530 − 1,450) × 500 shares = £400 plus the £138.50: £538.50. One standard HSBC put would cost £261.40 with its £1.40 commission and cover 1,000 shares, twice the holding. Each mini put starts with a delta of about 26 share-equivalents, so the five offset about 132 of the 500 shares at the outset, more as HSBC falls towards 1,450p.

At expiry an in-the-money put can be sold or exercised, and for a hedge the two are different events. Exercising sells the 500 shares at 1,450p: a disposal of the shares on the exercise date, with the premium an incidental cost of that sale (TCGA 1992 s144(3)(b)), which crystallises whatever gain the shares carry. The writer who is assigned buys the shares and pays the 0.5% stamp duty reserve tax; the exercising holder pays none (who pays SDRT). Selling the put instead keeps the shares, and the put's result is its own gain or loss. The minis are listed on ICE; whether a broker offers them and quotes a two-way price has to be checked on its platform (could not be confirmed, 26 September 2026). The long put page works a protective put on a 1,000-share holding in full.

A US holding: the put protects dollars, not pounds

Model inputs. 100 shares of a hypothetical US company at $150 on Monday 17 August 2026 (the model share of Examples 14 and 15 on the tax worked examples page), IV 30%, US rate 3.625%, no dividend; converted at an illustrative $1.3559 per £1 on 17 August 2026 (ECB reference-rate cross); use your broker's rate. One October $140 put, 60 days, American, priced on the tree: model $2.91, filled at $2.90 on the $0.05 tick. Commission $1.00 an order (Interactive Brokers' minimum).

The holding is worth $15,000, £11,062.76. The put costs $290.00 a contract, £213.88. Now suppose that on 16 October the share is at $120, 20% lower, and the pound has risen to $1.40, an assumption for the example. In dollars the holding has lost $3,000 and the put is worth $2,000, so the hedged holding is down $1,292 with costs, about 8.6%. In pounds the picture is worse: the shares are now worth £8,571.43, down £2,491.33 or 22.52%, and the put sold for $2,000 brings £1,428.57. After its cost, the sterling result is −£1,278.09, 11.55% of the starting value. The put did its job in dollars; the currency move is unhedged.

For tax, the put is converted leg by leg, each on its own date (HMRC CG78310): proceeds £1,428.57 at $1.40 less the premium and both commissions converted at their own dates, £215.33, gives a gain of £1,213.24 in 2026/27. A net dollar figure is never converted in one step. The US options page covers access, hours and the W-8BEN, and the tax worked examples the two-rate rule.

Tax: the ISA and GIA asymmetry

Options cannot be held in any ISA, so a listed put protecting an ISA portfolio has to sit in a general investment account (GIA) (the ISA rule). The two accounts are then taxed on different bases: nothing in the ISA is ever taxed or relieved, while the put's result in the GIA is a chargeable gain or an allowable loss. The table puts numbers on it for portfolio A and the December 10,250 put, at an 18 December EDSP of 8,600 (a 20% fall) and of 11,825 (a 10% rise). A spread-bet put at £9.30 a point, struck at the same premium in points, is shown for comparison, because a spread bet is outside capital gains tax altogether (HMRC CG56105).

The asymmetry cuts both ways. If the market falls, an ISA investor with a listed put owes tax on the put's gain in the same year the portfolio lost more than the put made, and the ISA loss can never be set against it. If the market rises, the put's cost, and on a collar the call's settlement, become allowable losses that are useful only to someone with other gains to absorb them. For a portfolio held in a GIA the difference is timing: the put's gain is taxed in the year it settles, while the portfolio's fall produces a loss only when holdings are sold, and selling to realise that loss then buying back within 30 days is matched with the purchase instead (the 30-day rule). Buying the put is not itself a disposal of the portfolio; the option is a separate asset.

The spread bet sits the other way: its gain is untaxed and its loss unrelieved, which mirrors an ISA portfolio. That symmetry has costs of its own: the provider's all-in spread (at 4 to 8 points, half of it is £18.60 to £37.20 on a £9.30 stake at each end), the FCA rules for restricted speculative investments, and a counterparty that is the firm rather than an exchange. HMRC's manual on bets made in the course of a trade (BIM22019) concerns businesses hedging trading risks; it does not bring a private investor's hedging bet into tax. Tax figures assume the £3,000 annual exempt amount is used by other gains. The wrappers page covers each account, and the tax worked examples the bought-option rules in full.

Pages that put these ideas to work

Long put (a floor under 1,000 shares) · collar (the floor paid for with the upside) · bear put spread (cover between two strikes) · long strangle · jade lizard · FTSE 100 options · position sizing.

How these numbers are calculated