CM Williams VIX Fix Across Stocks, Crypto and Forex
CM Williams VIX Fix can be calculated from closing prices and lows without an options feed. That makes it possible to study the same price-excursion formula across stocks, crypto and forex. It does not make the resulting signals equally useful in all three markets.
The first comparison should be the data and rule definition. A daily stock bar, a daily crypto bar and a broker's daily forex bar can represent different sessions and cutoffs.
What is being compared?
Chris Moody's original TradingView publication credits Larry Williams for the underlying idea. "CM" identifies the script's author; it does not identify a separate asset class or a guaranteed trading method.
For lookback N, the core formula is:
WVF = 100 × (highest close over N bars − current low) / highest close over N bars
With a highest close of 200 and current low of 190, WVF is 5. Multiplying every price by the same positive constant leaves that percentage unchanged. Changing the bar aggregation, session or adjustment method can change the inputs and the result.
WVF is a price-based distance, not options-implied volatility and not the conventional standard deviation of realized returns. Cboe's VIX methodology instead derives 30-day expected volatility from SPX options. A shared nickname such as "fear gauge" does not make the two calculations equivalent.
Keep the market data comparable
| Market | Record before comparing WVF signals |
|---|---|
| Stocks and ETFs | Exchange/feed, regular or extended session, adjustment settings and corporate-action treatment |
| Crypto | Exchange, pair, quote currency, spot versus derivative contract and daily cutoff |
| Forex | Broker/feed, quote basis, session cutoff and weekend handling |
| Futures | Specific contract versus continuous series, roll method and adjustment policy |
A split or a contract adjustment can alter a rolling price reference. A thin or erroneous low can also create a large WVF value. Inspect unusual input bars before interpreting them as market-wide panic.
A lookback of 22 means 22 bars, not always one calendar month. Changing from daily to four-hour bars changes the observation horizon. Applying the same number on two timeframes is not a controlled test of timeframe alone.
Confirm which script you installed
On TradingView, search for the original indicator and verify ChrisMoody as its author. Save the publication link and settings with your test record. Other scripts may add filters, alerts or inverted calculations.
For example, Williams Vix Fix + Inverse by Alorse describes inverse functionality as an addition to the CM version. Do not assume a "Highs Not Lows" switch exists in every WVF script. A small value of the original downside formula is not automatically a market-top or short-selling signal.
The current Pineify Williams VIX Fix component provides the raw value, upper band and scaled high/low range outputs. It has no inversion switch. It can help construct a custom script, but any modified implementation must be compared with the selected reference before its results are pooled.
For the detailed Pineify controls and formulas, see the WVF calculation and setup guide. The indicator's catalog access is plan dependent; this comparison does not claim a free entitlement or a verified customer count.
A green bar is a condition, not an order
In Pineify's current drawing, lime means WVF is at least its upper Bollinger-style band or at least its scaled rolling maximum threshold. The latter is often labeled "Percentile," but multiplying the maximum by a factor is not percentile rank.
Three candidate events should be evaluated separately:
- WVF remains above a threshold.
- WVF crosses above that threshold for the first time.
- WVF returns below it after a period of high readings.
They can occur on different bars at different prices. A sequence of high readings during one decline is also not the same as several independent predictions.
Before enabling an alert, inspect the events that the chosen version actually exposes. TradingView's alert documentation explains that alerts use a stored script-and-settings snapshot. Recreate an alert after relevant changes. An indicator condition does not guarantee notification delivery or a broker fill.
Evaluate a cross-market hypothesis
Suppose the question is whether first upward threshold crossings are followed by positive five-bar returns. This is a hypothetical study specification, not a reported result.
- Select instruments and dates before checking the results; include unsuccessful or discontinued instruments where relevant to the universe.
- Define whether the event is evaluated at the bar close and whether the return begins at that close or a later executable price.
- Specify how overlapping events are handled and whether each instrument receives equal weight.
- Compare with unconditional five-bar returns over the same periods, rather than assuming every positive outcome is caused by the signal.
- Report the event count, return distribution and adverse excursions by market; do not hide a weak market inside a pooled average.
- Use a separate interval to evaluate any parameters selected from the initial sample.
Adding a Stochastic cross, ADX filter or support condition creates another hypothesis. Retain the original rule as the comparison and disclose how many combinations were tried. No evidence here establishes the earlier stock-specific 18-bar or crypto-specific 16-bar presets as optimal.
PercentRank requires a separate definition
An empirical rank counts observations relative to a value. It is not "percent of the range" and is not the same as the script's scaled maximum line. A rule such as rank above 98 over a ten-observation window must specify whether the current observation is included, how ties are handled and which denominator is used. With such a short window, attainable ranks are coarse.
Do not describe a rank-based strategy as quantified evidence unless its data, exact ranking convention, orders and results are available. It can be a research idea, but no measured performance is supplied here.
If turning the event study into a trading strategy, add position sizing, costs, stop and exit rules, financing where relevant, and fill assumptions. TradingView's strategy documentation describes the emulator; a historical signal plot alone is not an execution test.
Porting to MetaTrader
Pine Script does not run directly as an MT4 or MT5 indicator. A MetaTrader implementation requires separately written or converted code. No specific verified MT4 download is supplied here, so the feature set of an unnamed port cannot be assumed.
Compare the port's highest-close window, current-low input, standard-deviation convention, history indexing and bar timestamps with the reference. Matching colors on a screenshot is insufficient. Retain a small table of raw values on aligned closed bars, then test live-bar behavior and alerts separately.
Frequently asked questions
Can the calculation run on crypto or forex data?
It can be evaluated where suitable close and low data are available and the denominator is meaningful. Calculation availability does not establish predictive accuracy.
Does WVF calculate implied volatility like the VIX?
No. WVF uses historical price levels; the VIX uses SPX options to estimate expected volatility.
Does the original downside formula find market tops?
A low value alone does not establish a top. Inverse variants use additional logic and should be identified and evaluated separately.
Are daily and four-hour signals independent confirmation?
They often use overlapping price information. Agreement does not by itself establish independent evidence or a higher probability of profit.