Production System
Methodology: how Vyreon measures options and market activity
The rules behind every Vyreon measurement: compared with each security's own history, point in time, like-with-like comparisons, lawful gaps instead of filled values, and verified before publication.
Vyreon computes a fixed set of measurements for each of 64 US stocks and ETFs once per trading day, after the close, and compares each one with that security's own history. This page explains the rules every measurement follows. The Learn pages explain the measurements themselves.
What Is Measured
| Family | Examples | Computed from |
|---|---|---|
| Implied volatility | level, skew across strikes, term structure across expiries | end-of-day option chains |
| Options positioning | open interest by expiry and strike, how it shifts over five sessions, put/call ratios | end-of-day option chains |
| Options activity | intraday option trading in a fixed set of contracts | 30-minute option intervals |
| Price behavior | daily returns, realized volatility, relative volume, the path of recent sessions | daily bars, adjusted for dividends and splits |
| Technical indicators | RSI, MACD, moving averages, Bollinger and Keltner bands, pivots, candlestick patterns | daily bars, adjusted for dividends and splits |
| Intraday equity activity | intraday path, session VWAP | 15-minute Cboe One equity bars |
| Company context (17 single stocks) | reported fundamentals and filing events, as published at the time | company filings |
| Macro context | inflation, policy rate and industrial production releases | official economic series |
| Derived measurements | gamma exposure and walls, max pain, IV rank, expected move, technical indicators, beta and relative performance | computed from the measurements above |
All inputs are licensed market data. Option implied volatilities are either the provider's or Vyreon's own fit to quoted prices, and each page says which.
Rules Every Measurement Follows
Compared with its own history. A reading is shown with where it sits in that security's recent sessions, typically the last 126 sessions (about six months), as a percentile. Rankings only use sessions where the measurement itself was supported.
Point in time. Each day uses only data that existed by that day's cutoff. Open interest is the count published after the previous session. Company facts count from the moment they were filed, not the period they describe. Nothing is rewritten with hindsight.
Like with like. Comparisons across days use the same contracts. For example, the five-session change in open interest counts only expiries alive on both days, so contracts simply expiring do not show up as a shift.
Lawful gaps, never fabricated values. Every value carries a state: supported, withheld (with a reason such as "too few eligible contracts"), not applicable (a measurement that does not exist for that security, such as company results for an ETF), or not measured. Gaps stay gaps. Vyreon never fills a missing day or value by interpolating, carrying a value forward or substituting a zero. Two stated conventions are not gaps: model inputs that are unobserved (an interest rate or dividend carry) use a stated zero reference, and an option with no reported volume is counted as not traded. Some measurement definitions interpolate within one day's data, for example between expiries to reach a 30-day at-the-money implied volatility; each explainer states where it does.
Verified before publication. Each day is computed in full for all 64 securities, then stored as verified only if every expected measurement is present, and read back from storage before it is served. A day that fails these checks is not published.
Corrections are visible. If a method changes or an error is fixed, the affected days are republished as new versions; earlier versions stay on record.
Unusualness
A reading is "unusual" when it sits far into the tail of that security's own history. The paid weekly review will rank the most unusual readings across all 64 securities. Because many measurements are checked every week, some will look unusual by chance, so each issue will state how many readings were flagged beside how many a typical week in our own history flags. Unusual does not mean important, and it never implies a direction.
What The Measurements Do Not Tell You
- Where the price will go.
- Whether to buy, sell or hold.
- Whose positions open interest belongs to, or whether they are hedges or bets.
- Whether a high put/call ratio is bearish or a low one bullish.
Known Limits
- Intraday equity activity currently uses one exchange group's bars (Cboe One, typically 5–12% of consolidated volume), not the full consolidated tape. Consolidated data is planned for launch.
- Individual option trades are only available from the provider for a few days, so two execution-based measurements are reported only from the day daily capture begins.
- Thin options markets (some ETFs, far-dated expiries) are withheld more often, by design.
- Data errors happen. Provider data can be late or wrong; a reading built on a bad input is wrong too. Report anything that looks off.
Vyreon publishes general market information for research and education. It is not investment advice. See the disclaimer.