Lesson 12 of 25 · 7 min
The term structure of volatility
Each expiry quotes an average over its whole life, so the difference between two of them prices a stretch of time neither one covers.
Implied volatility varies by expiry as well as by strike, and the variation by expiry has a structure worth extracting. Take an illustrative set of at-the-money figures: one month 18.2 percent, three months 20.1 percent, six months 21.4 percent, twelve months 22.3 percent.
The first thing to notice is that these are not four forecasts of four different dates. Each one is an average over its own horizon. The three-month figure contains the first month inside it, so the two numbers overlap and cannot be read as independent statements.
Extracting a statement about the stretch in between uses the additivity from lesson 3. Variance multiplied by time adds, so the three-month total minus the one-month total leaves the total for months two and three. Compute it: 0.182 squared times one twelfth is 0.0027603, and 0.201 squared times a quarter is 0.0101003. The difference is 0.0073399, spread over two months, giving 0.0440395 in annual terms. Its square root is 20.99 percent.
forward variance = (far volatility squared × far time − near volatility squared × near time) ÷ (far time − near time)
- forward variance
- The variance the market is pricing for the stretch between the two expiries. Its square root is the forward volatility.
- far volatility
- The implied volatility of the longer expiry, as a decimal.
- far time
- The longer expiry's time to maturity in years.
- near volatility
- The implied volatility of the shorter expiry, as a decimal.
- near time
- The shorter expiry's time to maturity in years.
Doing the same for the later segments gives 22.63 percent between three and six months and 23.17 percent between six and twelve. So the curve is not merely rising; it is pricing successively higher volatility for successively later stretches, which is what an upward slope means once the overlap is removed.
For collateral the relevant end of this curve is the front, because a lender's assumed exit is measured in sessions rather than months. An upward-sloping structure puts the smallest figure on the surface closest to the lender's horizon, which means a haircut built on the near-dated point is the smallest one the surface supports. It is also the point that moves the most, which is lesson 13.
Key terms
- Term structure of volatility
- Implied volatility plotted against time to expiry. Normally upward sloping, and each point is an average over its own horizon rather than a forecast of its date.
- Forward volatility
- The volatility priced for the stretch between two expiries, backed out by subtracting one expiry's variance-times-time from the other's.
- Overlap
- The fact that a longer expiry's horizon contains a shorter one's, which is why two quoted figures cannot be compared as if they described different periods.
Worked example
Removing the overlap from a rising curve
Illustrative at-the-money implied volatilities: one month 18.2 percent, three months 20.1 percent, six months 21.4 percent, twelve months 22.3 percent.
- One-month variance times time
- 0.0027603
- Three-month variance times time
- 0.0101003
- Difference, over two months
- 0.0073399
- Annualised forward variance
- 0.0440395
- Forward volatility, months two and three
- 20.99%
- Forward volatility for the later segments
- 22.63% and 23.17%
0.182 squared ÷ 12.
0.201 squared × 0.25.
0.0073399 × 6.
The quoted curve climbed from 18.2 to 22.3 percent. Stripped of overlap, it is pricing 20.99, 22.63 and 23.17 percent for three consecutive stretches.
Check yourself
One-month implied volatility is 18.2 percent and three-month is 20.1 percent. What is the market pricing for the second and third months alone?
Saved in this browser only. No account, nothing sent anywhere.
Educational only. Nothing in this lesson is investment, legal or tax advice, or a recommendation to enter into any transaction. Figures are illustrative and computed from the inputs stated in the lesson.