Hedging in Practice: What It Can Protect — and What It Costs

Hedging has the best reputation of any word in trading — it sounds like what careful professionals do. Sometimes it is. A hedge is a deliberate offsetting position that reduces an exposure you want to reduce, for a period you can name, at a cost you have added up. That last clause is where the reputation outruns the practice: every hedge has a carrying cost, paid in spread, financing, and forgone profit, and a large share of retail “hedging” is something else wearing the word — a refusal to close a losing trade. This sheet works one legitimate hedge end to end, prices it honestly, and then dissects the popular counterfeit.

What a hedge is — and is not

Strip the mystique and ▲hedging is bookkeeping: you hold an ▲exposure you want to keep for the long run, something threatens it in the short run, and you open an opposing position so that, for a while, moves in one are partly cancelled by the other. The corporate version is unglamorous — an exporter expecting dollars in ninety days sells them forward today, exchanging an uncertain rate for a certain one. Note what the exporter buys: not profit, but certainty. That is all any honest hedge ever buys.

Three tests separate a real hedge from a rationalization, and they work as a checklist before you open one. Purpose: which specific risk, over which specific period? Cost: what will the hedge charge in spread, financing, and forgone gains — in numbers? Exit: what event or date removes it? A position that fails any of the three is not a hedge; it is a second opinion you are paying to hold.

The instruments: direct, proxy, and basis risk

A retail CFD account offers two hedge shapes. The direct hedge is ▲going short the same pair you are long (where the broker permits holding both): mechanically exact, since each pip cancels pip for pip across the legs. The proxy hedge offsets through a different but related instrument — short EUR/GBP against a long GBP/USD position, or a position in one dollar pair against another — leaning on ▲correlation to do the cancelling.

The proxy adds a risk the direct hedge does not have: basis risk, the gap between how the two instruments actually move. Correlation is a fair-weather statistic — the previous sheet showed it drifting in calm markets and snapping tight in stressed ones — and a proxy hedge can underperform exactly when the threat it was bought against arrives. The honest rule: hedge directly when you can; accept a proxy only when you have priced the chance that it cancels less than you hoped.

Availability differs by venue, and it shapes behaviour. Some regulators and platforms do not allow holding both directions of the same pair — a new opposing order simply offsets the existing position — which pushes traders in those venues toward proxy hedges and their basis risk by default. Where direct hedging is allowed, brokers differ again in how they margin the offsetting legs. None of this changes the arithmetic of this sheet; it changes which version of the arithmetic applies to you, which makes it one more line to check in the account terms rather than assume.

A legitimate case, fully costed

The setup: you are long 1 lot of EUR/USD from 1.0780, the price now stands at 1.0850, and your thesis is measured in weeks. In two days a central-bank decision lands — exactly the kind of event that can gap through a stop-loss. You want to keep the position but shrink its sensitivity through the event, so you sell 0.5 lots at 1.0850, planning to lift the hedge once the dust settles, about three nights later. The position's net sensitivity drops from $10 to $5 per pip.

Watch the symmetry do its work through the event. If the decision knocks EUR/USD down 60 pips, the unhedged position loses $600; hedged, $300. If it lifts the pair 60 pips, the unhedged position gains $600; hedged, $300. The hedge halved both outcomes — it did not improve the average result by a cent, and no hedge can. What it bought is a narrower range of outcomes through one event, and that purchase has a bill:

Roughly eight dollars in hard costs, plus half of whatever the market would have given you, to halve three days of event risk on a position you intend to keep. Reasonable — and now comparable. The alternative was closing the full lot and re-entering after the event: one extra spread (≈ $9 at 0.9 pips on a full lot), zero exposure through the decision, and the risk of re-entering 30 pips worse. Neither choice dominates; the point is that both can be priced, and a decision made in numbers. The ▲swap line deserves respect at longer horizons: a hedge held for a month at these rates costs about $33 in financing alone before anything else — carrying costs compound quietly while protection feels free.

Be precise about what even this clean, direct hedge does not do. It does not protect the 70 pips already earned — those stay exposed on the unhedged half. It does not remove gap risk; it halves it. And it leaves a decision for after the event: lift the hedge too early and the protection was theatre, too late and the carrying cost keeps running against a thesis that has already resumed. A hedge is not a decision avoided. It is two decisions, scheduled.

The self-deception hedge

Now the counterfeit. You are long 1 lot of EUR/USD from 1.0850 and the price sits at 1.0790 — sixty pips and $600 against you. Closing makes the loss real, and that hurts. So instead you sell 1 lot at 1.0790 and call it a hedge. Look at what you now hold: the long loses a dollar per pip up exactly as the short gains it, and vice versa. Your net result is frozen at −$600 wherever the price goes. Economically, you closed the trade — you just didn't tell yourself.

Except closing would have been cheaper. Both legs now pay swap — say a combined −$6.30 a night, which is −$63 over ten nights and −$189 over a month, billed against a position that can no longer earn anything. And where closing leaves one decision already made, the locked pair leaves two still to make: unwind which leg, when? Most traders who lock a loss open eventually unwind the wrong leg at the wrong moment, turning one frozen −$600 into two live mistakes. The three-test checklist catches this instantly. Purpose? None — the net exposure is zero, so nothing is being protected. Cost? Accruing nightly. Exit? Unplanned. It is not a hedge; it is a fee paid daily to avoid feeling a loss that has already happened.

If the pattern sounds familiar, the psychology region of this survey gives it a name: loss aversion — the preference for an open uncertainty over a closed certainty of the same size. The preference is human and nearly universal. The market simply charges rent for it, nightly, on the swap line of the ledger.

The decision rule

One question sorts every case on this sheet: if you held no position right now, would you open this one? Apply it to the original trade, not the hedge. In the event-risk case the answer was yes — the thesis was alive, only the next three days were the problem — and a priced, time-boxed hedge was a defensible tool. In the locked-loss case the answer was no; the trader no longer believed in the long and was only avoiding the act of closing it. When the answer is no, the instrument you want is not a hedge. It is the close button, followed, if the thesis truly reverses, by an honest position the other way — sized from scratch by the same arithmetic as any other trade.

The rule generalises beyond hedging, which is why it earns the last word on this sheet: positions are held by choice, every day they are open, and a choice you would not make fresh is a choice already running against you.