Open Volatility Intelligence

Product note

Volatility Intelligence: What the Vol Surface Is Actually Saying

Inside the new Volatility Intelligence page — how Pulse measures implied versus realized volatility, why most "IV rank" numbers are quietly wrong, and what you can do with a read you can trust.

Pulse · November 2026·8 min read

Ask ten futures traders what implied volatility is doing and you will get ten numbers. Ask them 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. It is a single screen, one instrument at a time, built entirely from Pulse’s own stored option prints — and it is deliberately honest about what it does and does not know.

Why a volatility number is not a volatility read

An implied volatility figure on its own tells you almost nothing. 32% IV on crude is cheap, rich, or meaningless depending on three things:

  • Tenor. A 12-day option and a 40-day option on the same underlying are not the same market. Ranking one against the other produces a number that looks precise and is not.
  • History. “Elevated” only means something against the instrument’s own distribution — and only if that distribution was built from comparable prints.
  • Sample size. A percentile calculated from six observations is a story, not a statistic.

Most screens skip those distinctions. Volatility Intelligence refuses to.

How it works

  1. 1

    Implied volatility, ranked like-for-like

    Every stored print carries a single days-to-expiration value. The page compares today’s ATM implied read only against sessions that sat 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 back them. Below that, the figure is labelled Limited Sample and no classification is published at all. A pooled all-tenor number is still shown, but always tagged as exploratory rather than treated as a rank.

  2. 2

    Realized volatility, built from the underlying

    Realized vol comes from the instrument’s own session closes — never from implied values. It is computed over 10, 20 and 30-session lookbacks, annualized across 252 sessions, and every return that crosses a contract rollover or a data gap longer than four days is excluded so a roll does not masquerade as a 6% move. A lookback is only published when at least 60% of its window is usable.

  3. 3

    The spread, named precisely

    The page shows the gap between implied and realized — but always as “IV at tenor X versus RV over Y sessions,” never a bare number. That naming convention is the whole point: a spread you cannot attribute is not evidence.

  4. 4

    Regime context

    Descriptive bands on rank, spread, term structure and trend. Ranks under 25 read as historically low implied vol, above 70 as historically elevated. Term structure is flagged flat when the front and back sit within half a volatility point.

  5. 5

    A validation report across every instrument

    Each tracked market is scored against each measurement and given a status: Validated, Limited Sample, Mixed Tenor, Stale, or Unavailable. Only Validated, tenor-matched measurements are allowed to feed the screening logic. If a symbol’s history cannot support a credible rank, the page says so plainly instead of filling the space.

  6. 6

    Data quality, in the open

    A final panel shows exactly what went into the numbers, what was thrown out, 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 itself every ten minutes while open.

What it deliberately will not do

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 be dressing up an assumption as an edge.

Nothing on this page recommends a trade, and nothing on it is backtested performance.

The benefits

  • Fewer false signals. Tenor-matched ranking plus a twenty-observation minimum removes the most common source of phantom “IV percentile” readings.
  • Realized vol you can defend. Roll-adjusted, gap-filtered and annualized on a consistent basis — so an implied-versus-realized spread means what it claims to.
  • Faster structure triage. One screen answers: is vol cheap or rich by this instrument’s own standards, is the premium being paid for it being realized, and is the term structure flat, steep or inverted.
  • Visible limits. Stale, limited, mixed-tenor and unavailable states are displayed, not hidden. You know when you are reading a strong sample and when you are reading a thin one.
  • An audit trail by design. Every figure carries its tenor, window and sample count, which makes it reviewable after the trade — the difference between a journal entry and a guess.

How traders use it

A practical routine: check the validation status first, then the implied rank, then whether realized vol is confirming or contradicting 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 — and Volatility Intelligence treats it that way.

Pulse is an educational derivatives intelligence platform. Nothing in this article or in the app constitutes financial, investment, or trading advice.

Explore Volatility Intelligence