PRODUCT NOTE / NOVEMBER 15, 2026

Volatility Intelligence: What the Vol Surface Is Actually Saying

Volatility Intelligence: What the Vol Surface Is Actually Saying

Volatility Intelligence: What the Vol Surface Is Actually Saying

Inside the new Volatility Intelligence page: implied versus realized volatility, trustworthy IV ranks, and a read you can defend.

Inside the new Volatility Intelligence page: implied versus realized volatility, trustworthy IV ranks, and a read you can defend.

Inside the new Volatility Intelligence page: implied versus realized volatility, trustworthy IV ranks, and a read you can defend.

BY NICHOLAS DUFF

01 / WHY THE NUMBER IS NOT THE READ

Ask ten futures traders what implied volatility is doing and you will get ten numbers. Ask what tenor that number came from, over what lookback it was ranked, and how many like-for-like observations sit behind the rank, and most will go quiet. That gap is where bad structure decisions get made. The new Volatility Intelligence page in Pulse exists to close it: one instrument at a time, built entirely from Pulse’s own stored option prints, and deliberately honest about what it does and does not know.

Ask ten futures traders what implied volatility is doing and you will get ten numbers. Ask what tenor that number came from, over what lookback it was ranked, and how many like-for-like observations sit behind the rank, and most will go quiet. That gap is where bad structure decisions get made. The new Volatility Intelligence page in Pulse exists to close it: one instrument at a time, built entirely from Pulse’s own stored option prints, and deliberately honest about what it does and does not know.

An implied volatility figure on its own tells you almost nothing. Thirty-two percent IV on crude can be cheap, rich, or meaningless depending on three things: tenor, history, and sample size. A 12-day option and a 40-day option are not the same market; elevated only means something against the instrument’s own comparable distribution; and a percentile calculated from six observations is a story, not a statistic. Most screens skip those distinctions. Volatility Intelligence refuses to.

An implied volatility figure on its own tells you almost nothing. Thirty-two percent IV on crude can be cheap, rich, or meaningless depending on three things: tenor, history, and sample size. A 12-day option and a 40-day option are not the same market; elevated only means something against the instrument’s own comparable distribution; and a percentile calculated from six observations is a story, not a statistic. Most screens skip those distinctions. Volatility Intelligence refuses to.

02 / HOW IT WORKS

Implied volatility is ranked like-for-like. Every stored print carries a days-to-expiration value, and today’s ATM implied read is compared only against sessions within seven days of that same tenor across rolling 60, 120, and 252-session windows. Rank and percentile are published only when at least twenty tenor-matched sessions support them. Below that, the page says Limited Sample and withholds classification; pooled all-tenor figures stay visible only as exploratory context.

Implied volatility is ranked like-for-like. Every stored print carries a days-to-expiration value, and today’s ATM implied read is compared only against sessions within seven days of that same tenor across rolling 60, 120, and 252-session windows. Rank and percentile are published only when at least twenty tenor-matched sessions support them. Below that, the page says Limited Sample and withholds classification; pooled all-tenor figures stay visible only as exploratory context.

Realized volatility is built from the underlying’s own session closes, never from implied values. It uses 10, 20, and 30-session lookbacks annualized over 252 sessions. Returns that cross a contract rollover or a data gap longer than four days are excluded, so a roll does not masquerade as a six-percent move. A lookback is shown only when at least sixty percent of its window is usable.

Realized volatility is built from the underlying’s own session closes, never from implied values. It uses 10, 20, and 30-session lookbacks annualized over 252 sessions. Returns that cross a contract rollover or a data gap longer than four days are excluded, so a roll does not masquerade as a six-percent move. A lookback is shown only when at least sixty percent of its window is usable.

03 / THE DECISION STAYS YOURS

The spread is named precisely: IV at tenor X versus RV over Y sessions, never a bare number. Descriptive bands then place rank, spread, term structure, and trend in context. Ranks under 25 read as historically low implied vol; above 70, historically elevated. Term structure is flagged flat when the front and back sit within half a volatility point.

The spread is named precisely: IV at tenor X versus RV over Y sessions, never a bare number. Descriptive bands then place rank, spread, term structure, and trend in context. Ranks under 25 read as historically low implied vol; above 70, historically elevated. Term structure is flagged flat when the front and back sit within half a volatility point.

Each tracked market receives a validation status: Validated, Limited Sample, Mixed Tenor, Stale, or Unavailable. Only validated, tenor-matched measurements feed the screening logic. If a symbol’s history cannot support a credible rank, the page says so plainly instead of filling the space. A data-quality panel also shows what entered the calculation, what was excluded, and how old the read is.

Each tracked market receives a validation status: Validated, Limited Sample, Mixed Tenor, Stale, or Unavailable. Only validated, tenor-matched measurements feed the screening logic. If a symbol’s history cannot support a credible rank, the page says so plainly instead of filling the space. A data-quality panel also shows what entered the calculation, what was excluded, and how old the read is.

If the underlying print is more than 45 minutes old, the measurement is marked stale rather than silently treated as current. The page refreshes every ten minutes while open. Visible limits are part of the read: you know when you are working from a strong sample and when you are looking at a thin one.

The Premium Strategy Research panel screens structure types against the current volatility state and then stops. It classifies; it does not recommend. Choosing between a defined-risk debit structure and a premium-selling structure requires a forward volatility forecast, and Pulse does not hold one. Publishing that as a trade call would dress up an assumption as an edge.

The benefit is faster structure triage without false precision. Tenor-matched ranking and a twenty-observation minimum remove phantom IV-percentile readings. Roll-adjusted, gap-filtered realized vol gives the implied-versus-realized spread its actual meaning. Every figure retains its tenor, window, and sample count, creating an audit trail rather than a guess.

A practical routine is simple: check validation status first, then implied rank, then whether realized volatility confirms or contradicts it, then term structure. If implied is historically elevated and realized is not keeping pace, premium-selling structures move up the list. If implied is low and realized is running hotter, long-vol structures get the attention. The page gives you the state of the surface; the decision stays yours.

Volatility is not a dial you read off a screen. It is a distribution, a tenor, and a sample. Volatility Intelligence treats it that way. Nothing on this page recommends a trade, and nothing on it is backtested performance.

Volatility is not a dial you read off a screen. It is a distribution, a tenor, and a sample. Volatility Intelligence treats it that way. Nothing on this page recommends a trade, and nothing on it is backtested performance.

Pulse — Market Clarity is brought to you by BearsTerritory.io