Most people reach for the Nikkei 225 VI when they want Japanese equity volatility.
That is one number for the whole surface. JPX publishes implied volatility for every
listed strike, every business day, at no cost — in the daily settlement file
rbYYYYMMDD.csv.
You do not have to imply the volatility yourself, and you do not need an options vendor to see the skew.
On 4 September 2026 the September series had seven days to expiry with the index at 65,020.94, and the file carried 298 distinct strikes. Selected points:
| Strike | Distance from spot | Call IV | Put IV |
|---|---|---|---|
| 55,000 | −15.4% | — | 50.2% |
| 58,000 | −10.8% | 37.0% | 40.0% |
| 62,000 | −4.6% | 28.8% | 29.6% |
| 65,000 | at the money | 25.0% | 25.8% |
| 68,000 | +4.6% | 25.2% | 26.2% |
| 72,000 | +10.7% | 27.5% | 34.2% |
| 75,000 | +15.3% | 30.1% | — |
The shape is the familiar equity smile, and it is steeply one-sided. Fifteen percent below spot prices at roughly double the at-the-money level (50.2% against 25.0%), while the same distance above spot prices at 30.1%. Downside protection is expensive; upside is not. A single volatility index cannot show you that.
Read the whole column and the range looks absurd: 1.0% to 114.8% on the same day in the same expiry. Neither extreme is a market volatility.
Deep in-the-money options print 1.0%. It is a floor value, not a quote — those contracts barely trade and their settlement price is essentially intrinsic value, so no meaningful volatility can be implied from them. In the chain above, every call struck below 55,000 shows exactly 1.0% for this reason.
The rule that follows is simple and easy to get wrong: build each side of the smile from out-of-the-money contracts. Puts below spot, calls above spot. If you average calls and puts at every strike without filtering, the in-the-money placeholders drag the whole surface toward zero.
Filtering the in-the-money side is not enough. Far into the wings, the file stops producing distinct values and repeats a single number across dozens of strikes.
On 9 September 2026, in the September series, 73.21% appeared at 119 different put strikes and 38.46% at 64 different call strikes. Those are ceilings, not quotes. Plotted raw they turn the tails of the smile into two flat shelves, which makes the skew look milder than it is and invites nonsense if you fit a curve through them.
The test is mechanical: count how many strikes share each volatility. Adjacent strikes genuinely can round to the same value near the money — within 10% of spot the largest cluster that day was six strikes — so a threshold somewhere above that separates real quotes from the plateau. Drop the repeated values and the smile comes back.
Neither of these traps is documented anywhere in the file. You only find them by plotting the column and noticing the shape is wrong.
Volatility per strike is what makes gamma computable from public data alone. Gamma depends on strike, time and volatility; open interest supplies the size. With both, dealer gamma exposure can be estimated across the whole chain without a vendor feed. That is how our gamma exposure series is produced, and why the gamma flip level can be located on some days and not on others.
It is also a cross-check on the headline volatility index. When the Nikkei 225 VI rises but at-the-money implied volatility does not, the move is in the wings — someone is paying up for tail protection rather than for movement in general.
The settlement CSV is linked from the JPX settlement price page, published around 18:00 JST, and encoded in Shift_JIS rather than UTF-8. Each row carries the strike, settlement price, implied volatility, days to expiry, interest rate and underlying level. Where to get Nikkei 225 data lists the rest of the files.
→ Live open interest by strike · → Gamma exposure explained · → Contract specifications