Hedging: Insurance Has a Price
Hedging is taking a second position whose gains offset losses in the first. It reduces variance in both directions rather than removing risk, and it always costs something. Because the two legs are rarely the same instrument, the difference between them is the exposure you kept.
How it works
A hedge is a second position taken to reduce the risk in the first. Its gains are meant to arrive when the original loses, so the two partly cancel.
It is insurance, and insurance has a price. The offset cancels gains as readily as losses, paid for in premium, financing, or the bid-ask spread on both legs.
A producer hedges a real exposure they already have. A farmer with a crop in the ground, or the mill that must buy it, uses futures to fix a price in advance. The exposure exists whether they act or not.
The financial version copies its shape. A shareholder buys protection with options, writes a covered call, or offsets an index with a short. Legitimate, and never free.
And opposing trades in one account are not a hedge. A long and a short in the same instrument is a closed position in costume, paying two spreads and, on some platforms, financing besides.
The risk you kept
The two legs never move exactly together. The hedge is rarely the same instrument as the exposure: an index against a single share, one delivery month against another, a fund against a basket.
That difference is the risk you kept. It is called basis risk, the part nobody mentions: a familiar exposure swapped for a smaller, stranger one governed by correlation.
Often the honest hedge is a smaller position. Two opposing trades cost two sets of fees to achieve what halving the size achieves for free, and the halving has no basis risk in it.
Hedging wins only when the exposure cannot be sold. A concentrated holding, a tax position, an illiquid asset, a business exposure — those earn a second leg.
And two positions pay two sets of costs. Entry and exit on both legs, financing overnight on a short selling leg, and premium if the protection was bought.
In practice
A thin hedging instrument is a worse hedge. Volume decides what leaving costs, and a hedge you cannot lift at a fair price is protection on paper.
The longer the hedge runs, the more it costs. Premium decays, financing accrues nightly, and an inverse ETF — an exchange-traded fund built to move against an index — drifts from its benchmark as the days add up.
And a gap moves the two legs by different amounts. An opening gap is where basis risk shows itself: the share reprices overnight, the index barely does, and the mismatch is realised at once.
A hedge is not a stop and does not replace one. A stop loss removes the exposure; a hedge keeps it and adds a second, which is why risk management still begins with risk per trade.
Every round trip costs 2% of a bar. On this site’s shared 576-bar history a round trip costs 0.0098 price units: 2% of a median bar’s range and 45% of the smallest.
The word travels further than the mechanism. In research/broker-coverage.json, a scan of 31,760
videos, 31 titles use “hedge” at a median of 48,469 views; the nine covering correlation manage 1,162.
Sizing the second leg
A hedge is sized by money at risk, not by counting shares. Matching share for share or contract for contract assumes the two instruments move one for one, which is the assumption basis risk breaks.
The honest calculation converts both sides into currency. Take the value of the exposure, multiply by its sensitivity to whatever you are hedging with, and buy that much offset — a beta-adjusted size when an index is standing in for a share.
Bought protection is sized the same way, through delta. A put option does not move one for one with the share; it moves by its delta, so the contract count follows from the delta, not from the holding.
And the size does not stay right on its own. Delta shifts as the price moves and sensitivity drifts with the market, so a hedge set once and left alone stops matching what it was meant to offset.
What hedging is not
It is not the removal of risk. It swaps a large exposure for a smaller, stranger one.
It is not free. Premium, financing and two spreads are paid whether the hedge was needed or not.
It is not a long and a short in the same instrument. That is a closed position paying twice.
And it is not a substitute for sizing. A hedge on an oversized position leaves it oversized.
When it fails
In a range the hedge just costs money. A trading range delivers neither the fall the protection was bought for nor the rise it would have cancelled, and the bill arrives anyway.
The failure that matters is basis. The exposure moves and the hedge does not follow, so a book described as protected takes a loss nobody had budgeted for.
A second is buying it after the move. Protection is priced off expected movement, so implied volatility is dearest at the moment hedging feels most obviously right.
Movement itself is not steady. On this site’s shared history the 14-bar average true range has a median of 0.5994 but runs from 0.2823 at the tenth percentile to 0.7954 at the ninetieth, a ratio of 2.82.
A third is leaving it on. A hedge is a decision about a period, not a state to live in, and a forgotten one quietly cancels the returns it was bought to defend.
And a fourth is hedging what could simply be sold. Where the position can be reduced, reducing it is cheaper and has no second leg to go wrong.
The original data
A hedge held permanently is an annual cost, and fee drag shows what that does. In
research/series-measurements.json, from site/measure_series.py, a charge compounded alone over
thirty years — no return assumed — removes 1.5% of the pot at 5 basis points (one hundredth of a
percentage point each), 5.8% at 20, 20.2% at 75 and 36.5% at 150.
And being underwater is the ordinary condition of holding anything. On the same 576-bar history 95% of bars sit below a prior peak — median shortfall 1.36%, deepest 3.76%, longest stretch 73 bars. Hedging each of those means paying continuously, so price the alternative of simply holding less first, and hedge only when the exposure genuinely cannot be reduced.
Related
Risk management sets the size before anything is hedged, which is usually where the problem started. Correlation is what basis risk is made of, and it is measured on the past. And options are the instrument most protection is built from, with a price attached to every one.
I hedged for a while because it felt more sophisticated than admitting the position was too big. What I actually bought was a second thing to watch, a second set of costs, and a gap between the two legs that moved on its own. Trimming the position would have done the same job for nothing. Now I only hedge what I genuinely cannot sell.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.