petralLEARNDesk ↗
CHAPTER 09LESSON 043 MIN READ

Vol term structure and the cone

How implied volatility varies with expiry, and how to tell whether a level is unusual.

WHY IT MATTERS

Compare short and long expiries, then compare each with its own history. A high level for one tenor may be ordinary for another.

Implied volatility varies with time to expiry, and the shape of that variation is the volatility term structure. It usually slopes upward in calm markets and inverts in stressed ones, and the reason is the mean-reverting nature of volatility itself.

The logic is worth working through because it explains the shape rather than just describing it. If current volatility is low and volatility mean-reverts to a higher long-run level, then a long-dated option covers a period in which volatility is expected to rise, so its implied volatility should exceed the near-dated one. When current volatility is very high, the same reasoning runs backwards: the market expects the storm to pass, so near-dated implied volatility exceeds long-dated. An inverted vol term structure is therefore a statement that the current disturbance is expected to be temporary.

There is a specific interaction with the forward curve that has no equivalent in equities. Each crude expiry references a different underlying contract, and those contracts have their own spread dynamics. So a crude vol term structure is not purely about time horizon; it is partly about which delivery month each option is on, and the near contracts are genuinely more volatile than deferred ones because they carry the prompt physical constraint.

The vol cone is the tool for answering whether a current level is unusual. For each horizon, plot the historical distribution of realised volatility at that horizon, then mark today's value on it. The result shows immediately whether 30 percent at 30 days is a high reading or an ordinary one for this market, which is a question no single number can answer.

percentile rank = share of historical observations at this horizon below the current value
horizon
the window length, e.g. 10, 30, 60, 90 days
historical observations
overlapping realised vol windows from the archive
Overlapping windows are correlated, so the effective sample is smaller than the observation count. A percentile from overlapping windows is directional evidence, not a precise probability.

Placing implied volatility on the same cone is where it becomes a desk tool: it shows what the market is charging against what the market has historically delivered at that horizon, which is the variance risk premium made visible rather than asserted.

CONVol ConeLIVE FROM THE DESKOpen on the desk
Realised volatility against its own distribution per horizon, with implied shown alongside.