Implied Volatility: What the FX Options Market Tells Spot Traders

Every day, currency options worth enormous amounts change hands, and their prices hide one number: implied volatility (IV), the volatility the market expects. A spot trader who never touches an option can still use it. You can turn it into the range the market expects, check whether a profit target is realistic, and see which side the market is more worried about. This article explains the mechanism, works through the numbers step by step, and is honest about where it goes wrong.

Options and IV in one paragraph

An option is a right (not an obligation) to buy or sell a currency pair at an agreed price, called the strike, by an expiry date. The right to buy is a call, the right to sell is a put. The buyer pays a premium up front. The more the market expects price to swing, the more that right is worth and the more expensive the premium.

Option prices come from a model. FX uses the Garman–Kohlhagen model, which is Black–Scholes adapted for two interest rates. Almost every input is known (spot, strike, time to expiry, interest rates). The one nobody knows is future volatility. So from the price an option actually trades at, you can work backwards to how much volatility the market has built in. That number is IV, shown as a percentage per year.

The FX market even quotes options in IV directly rather than in money, for example “EURUSD 1-week 7.0”, which means one-week options are priced at a volatility of 7% per year. If you’ve read Risk-on / Risk-off, you know VIX, which is the same idea for the S&P 500 stock index.

Turning IV into the market’s expected range

IV is an annual number, so you need to scale it down to the period you care about. Volatility grows with the square root of time, not in a straight line.

One-standard-deviation move (±1σ) ≈ Spot × IV × √(days ÷ 365)

Example: EURUSD is at 1.0850 and 1-week IV is 7%.

1. √(7 ÷ 365) = √0.019178 = 0.138485
2. 0.07 × 0.138485 = 0.009694 (about 0.97% of the price)
3. 1.0850 × 0.009694 = 0.010518, or about ±105 pips
4. ±1σ range = 1.0850 − 0.0105 to 1.0850 + 0.0105 = 1.0745 to 1.0955
5. ±2σ range = ±210 pips, which is 1.0640 to 1.1060

Using the lognormal distribution the model actually assumes gives 1.07453 to 1.09557, practically the same. Counting in trading days instead (5 out of 252 per year) gives about ±107 pips, again only a small difference. Other periods at the same 7% IV:

Period	±1σ	±1σ range from 1.0850
1 day	±40 pips	1.0810 to 1.0890
1 week	±105 pips	1.0745 to 1.0955
30 days	±218 pips	1.0632 to 1.1068

Under a normal distribution, price has about a 68% chance of closing the period inside ±1σ, and about 95% inside ±2σ. Notice that 30 days is only about twice as wide as 1 week, not four times, because the range grows with the square root of time.

Touching along the way is not the same as closing beyond

This is where a spot trader gains the most. The 68% range is about price at the close of the period, but our Stop loss and Take profit can be hit along the way.

Using a normal distribution with the same example (no trend, 1 week, 7% IV):

• Chance price closes above 1.0955 (beyond +1σ) = 1 − N(1) = 15.9%
• Chance price touches 1.0955 at some point during the week ≈ 2 × 15.9% = 31.7%
• Chance price touches either edge of the ±1σ range = 62.9%
• At ±2σ: closing beyond one side is 2.3%, touching one side is 4.6%, and touching either edge is 9.1%

The factor of two comes from the reflection principle of a random walk. Every path that touches the level then has a half chance of ending below it and a half chance of ending above. We checked this by simulating 20,000 paths of 1,000 steps each. They closed beyond +1σ 16.2% of the time, touched +1σ 31.2%, and touched either edge 61.6%, close to the formulas. The simulated figures are slightly lower because the price is checked at steps rather than continuously.

What it means in practice: if you place a Stop loss 105 pips away and hold for a week when IV is 7%, there’s roughly a 1 in 3 chance price reaches your Stop loss on that side, even if your direction turns out right by the end of the week.

Another figure you can compare with ATR: the expected High minus Low of a random walk is about 1.6 times the σ of that period (√(8 ÷ π) = 1.596). At 7% IV the daily σ is 40 pips, so the expected daily High minus Low is about 63 pips. If the pair’s actual daily ATR is far above this, realised volatility is running hotter than the options market expects, or the other way round. This ignores gaps and fat tails, so treat it as a rough comparison.

IV tells you the size of the waves, not their direction, and the chance of being touched along the way is always about twice the chance of closing beyond.

Risk reversals: which side the market pays up to protect

If everyone feared both directions equally, the IV of options far from the current price would be about the same on the call side and the put side. In real markets they usually aren’t. The FX market measures this with three standard numbers.

• ATM: the IV of options with a strike around the current price
• 25-delta risk reversal (RR) = IV of the 25-delta call minus IV of the 25-delta put. A 25-delta option has a strike a fair distance from the current price (out of the money), one on each side
• 25-delta butterfly (BF) ≈ the average of those two IVs minus ATM. It shows how much more the market charges for big moves in both directions than for the middle

Calls and puts here are on the first currency of the pair. For EURUSD that’s a euro call (EURUSD up) and a euro put (EURUSD down).

Example (made-up numbers): ATM 7.00%, RR −0.80, BF 0.25. You can split these back into the two side IVs with the standard approximation.

1. 25-delta call IV = ATM + BF + RR ÷ 2 = 7.00 + 0.25 − 0.40 = 6.85%
2. 25-delta put IV = ATM + BF − RR ÷ 2 = 7.00 + 0.25 + 0.40 = 7.65%
3. Check: RR = 6.85 − 7.65 = −0.80, and BF = (6.85 + 7.65) ÷ 2 − 7.00 = 0.25, matching what we started with

A negative RR means euro puts are more expensive. The market is paying more to protect against a sharp EURUSD fall than against a sharp rise. A positive RR means the opposite. For USDJPY the figure is quoted on dollar calls. When markets are scared, the USDJPY risk reversal usually goes more negative, because people rush to buy protection against a fast yen rally (USDJPY falling), which fits the yen’s role as a safe haven.

How to read it carefully:

• RR is the price of insurance, not a poll. The expensive side may come from exporters or funds that must hedge as part of their job. It doesn’t mean most people believe price will go that way
• Watch changes and compare with its own history. Each pair has its own normal RR level. A fast move deeper says more than a single reading
• RR tends to follow price. After a sharp fall, RR going more negative is normal, so the number reflects what just happened as much as what’s ahead
• Pair it with positioning data. An extreme RR together with extreme fund positioning in the COT report on the same side says the market is crowded one way. It doesn’t time a reversal

A positive butterfly is itself evidence that the market doesn’t fully trust the normal distribution. It prices big moves in both directions higher than ATM alone would suggest. That matches fat tails and black swans, where 3–4σ moves happen far more often than a textbook normal distribution expects.

Big option expiries near spot

Some days, financial news reports large option expiries, for example “EURUSD 1.0850, 1.5 billion euros, expiring today”. Most FX options expire at the New York cut, 10:00 New York time, which is 21:00 Thai time (UTC+7) while the US is on daylight saving time and 22:00 for the rest of the year. The other common one is the Tokyo cut at 15:00 Tokyo time (13:00 Thai time).

The mechanism people talk about: banks that hold or have sold options keep trading spot to hedge (delta hedging). If the banks own the options, hedging means selling when price rises and buying when it falls, which can hold price near the strike until expiry (pinning). If the banks have sold the options, hedging means buying into rises and selling into falls, which can speed moves up instead.

The problem is that the published lists show only the strike and the size. They don’t say who holds which side, and they don’t cover the whole market, because FX trades away from a central exchange. So you can’t know whether the effect will be pinning or acceleration. What can be said carefully is that large options with a strike very close to spot may make price behave unusually around the cut, and that effect usually fades after expiry. Treat it as a reason to be careful at that time, not a reason to trade.

Gold has its own index

Gold has the GVZ (Cboe Gold ETF Volatility Index), calculated from options on a large gold ETF using the same method as VIX. It shows the volatility the market expects over the next 30 days, as a percentage per year. The same range formula works. For example, a GVZ of 18 means a ±1σ range over 30 days of about 18% × √(30 ÷ 365) = 5.2% of the price (counting calendar days, as in the example above). If you trade gold, see the XAUUSD gold trading guide.

Using it as a spot trader

1. Check the target before you enter. Work out the ±1σ and ±2σ range for the period you plan to hold. If your Take profit is beyond 2σ within that time, the options market puts the chance of closing there very low (about 2% under a normal distribution). That’s not an ordinary target
2. Check the Stop loss against normal swings. A Stop loss closer than the daily σ sits in the middle of the day’s ordinary noise. Remember that the chance of being touched is about twice the chance of closing beyond
3. Adjust lot size to volatility. When IV jumps, a sensible Stop loss distance gets wider, so cut the lot size to keep the money at risk per trade the same. Use the lot calculator and read Position size from Stop loss
4. Watch IV for periods that cover big news. Short-dated IV that covers a rate decision or major release is usually higher than normal. It signals the market expects a jump, not a direction
5. Read RR as context. An RR deepening quickly shows worry about one side is growing. Use it when deciding how much risk to take, not as an entry point

Limits you need to know

• IV doesn’t forecast direction, and it isn’t an accurate forecast of size either. It’s the price the market agreed on at that moment
• IV can be wrong both ways. Many studies find that on average IV sits a little above the volatility that actually follows, known as the volatility risk premium, because insurance buyers pay extra. It’s clearest in stock indices. But in a crisis, realised volatility can blow far past IV
• The normal distribution understates the tails. The 68% and 95% figures and the factor of two are approximations from a world with no fat tails, no gaps and no trend. In real markets, moves beyond 3σ happen far more often than the normal distribution says
• The numbers change all the time. Monday’s 1-week IV is no good on Thursday. Recalculate with the latest value and the days that remain
• The data is hard to get. Most FX IV data sits in paid systems, free sources are often delayed or show only some expiries, and published expiry lists don’t cover the whole market

Common mistakes

• Reading high IV as a bearish signal. IV gives size, not direction. Look at RR to see which side is more expensive, and remember RR is still only the price of insurance
• Using annual IV as is. 7% per year doesn’t mean price will move 7% this week. Always multiply by √(days ÷ 365) first
• Treating the 68% range as the chance of not being stopped out. That range is about the closing price. The chance of a touch along the way is about twice as high
• Treating the ±2σ band as a wall. Price can break out of the range and run much further than the normal distribution suggests. Your position size has to survive a move beyond it
• Planning trades around option expiry lists. Public data can’t tell you which side the banks hold, so the effect may be the opposite of what you expect
• Using stale numbers. IV changes daily and jumps around big news. Last week’s figure can be well off