Level 10 · Lesson 8
Volatility
State Index
The derivative oscillator. ATR tells you how volatile the market is. VSI tells you whether that volatility is expanding, decaying, or stable. Different question, different answer, critical distinction.
First — Why This Matters
The Derivative Oscillator
Every oscillator you've met so far in the ATLAS suite measures something about the market. MAE measures where price is relative to an envelope. MSI classifies the current regime. MAZ shows acceptance zones. MPG measures participation. MPR measures pressure. All of them measure a first-order property of the market.
VSI breaks this pattern. It's an oscillator of an oscillator — it measures the rate of change of ATR, not ATR itself. This makes it epistemically different from everything that came before it: it's a second-order signal, a derivative, a measurement of a measurement.
Why does this matter? Because traders systematically conflate volatility with volatility change. A market with ATR = 200 that's been steady for 50 bars is profoundly different from one where ATR is 200 after climbing from 50 twenty bars ago. The raw value is identical. The regime is opposite. VSI is the instrument that makes this distinction visible — and it's the only ATLAS oscillator that operates on this level of abstraction.
🔎 THE VSI AXIOM
ATR tells you where you are. VSI tells you where you're going. Both readings matter. Neither replaces the other.
01 — The Three States
Expansion, Decay, Transition
VSI has three states, not four. Unlike MPR (which adds an off-axis TRAP state for effort absorption), VSI's measurement space is one-dimensional: volatility can rise, fall, or stay stable. No pathological off-axis condition exists because there's no equivalent of “absorbed volatility.” The three states map cleanly to the three positions along the derivative axis.
💡 Why 3 States and Not 4
MPR has 4 states because pressure has a directional dimension (positive/negative) AND a pathology (absorption can coexist with either). VSI has no such pathology: rising vs falling volatility is the entire axis. Adding more states would be over-engineering. The 3-state model is the minimum sufficient encoding for a one-dimensional signal.
02 — Base Volatility
Raw ATR Is Just the Starting Point
VSI starts with ta.atr(14) — the standard Average True Range. ATR measures volatility magnitude as a rolling average of true range over 14 bars. But raw ATR is too noisy for a momentum calculation: single-bar spikes, news events, and wick patterns create jagged fluctuations that would swamp any rate-of-change derivation. That's why VSI applies smoothing BEFORE computing momentum.
💡 Why ATR and Not Standard Deviation
Standard deviation measures dispersion around a mean; ATR measures true range (accounting for gaps and overnight moves). For futures, FX, and 24/7 markets alike, ATR is the more faithful volatility measurement because it includes open-to-prior-close gaps. Using stdev would understate volatility in gap-prone markets.
03 — Smoothing Layer
EMA(ATR, 10) — Isolating the Regime
The Pine source applies ta.ema(volBase, 10) to denoise the ATR series. This isolates the underlying volatility trend from single-bar noise. The smoothed series is what gets fed into the momentum calculation. Why EMA and not SMA? Because EMA weights recent values more heavily, making the smoothed series more responsive to actual regime changes while still filtering random spikes.
💡 The Two-Parameter Lookback
VSI uses two lookbacks: smoothLen (EMA period, default 10) and momLen (momentum period, default 10). The smoothing handles noise; the momentum period defines the comparison window. Tuning smoothLen shorter makes VSI more reactive. Tuning momLen longer makes each reading a broader comparison (how has vol changed over a longer window). Most traders leave both at 10; they're calibrated for the broad asset universe.
04 — Volatility Momentum
The Core Calculation
The defining calculation of VSI: volMomPct = (volSm - volSm[10]) / volSm[10] × 100. This is a standard rate-of-change formula applied to the smoothed ATR series. A reading of +28% means smoothed ATR is 28% higher than it was 10 bars ago. A reading of -15% means smoothed ATR is 15% lower. Zero means no change.
💡 Why Percent and Not Absolute Units
If VSI output absolute ATR change (e.g., “+30 points”), the value would be meaningless across assets — +30 on BTC is different from +30 on SPY. Percent-based output normalizes across any instrument: +28% means the same thing on BTC, EURUSD, gold, or SPY. This is the reason thresholds are also in percent (±5%), making the entire indicator asset-agnostic by construction.
05 — The Second-Derivative Principle ⭐
Measuring a Measurement
This is the conceptual breakthrough that makes VSI unique in the ATLAS suite. Every other indicator you've encountered measures a first-order property of the market directly. Price location, regime classification, acceptance zones, participation, pressure — all are direct measurements of something the market is doing. VSI is different. It measures the rate of change of another measurement. ATR is the underlying signal. VSI is the derivative of that signal.
🏘 The Second-Derivative Doctrine
Traders systematically confuse volatility with volatility change. “High ATR = volatile market” is true but incomplete. The trader who only checks ATR is reading a snapshot; the trader who checks VSI is reading a trajectory. For stop placement, position sizing, breakout anticipation, and compression detection, the trajectory matters more than the snapshot. Two markets with identical ATR can require opposite operational responses — one with decaying vol calls for tight stops and expanded position size (compression is building); one with expanding vol calls for wide stops and reduced size (volatility is running). The Second-Derivative Principle is the recognition that what ATR does matters more than what ATR is, and VSI is the instrument that makes this actionable.
Three operational implications:
- 1. Stop placement becomes regime-aware. Instead of mechanically using “2x ATR” as your stop distance, scale it by VSI: 1.5x in DECAY (tight, because vol is contracting and noise is limited), 2.0x in TRANSITION (baseline), 3.0x in EXPANSION (because vol is running and normal noise will now be larger). This is a direct application of the second-derivative insight.
- 2. Breakout anticipation gains a leading edge. Markets that are compressing (sustained DECAY) are building energy for the next expansion. Watching VSI cross from DECAY through TRANSITION into EXPANSION is often a leading signal of the directional move that volatility expansion is setting up. The move itself comes from MPR or price; the environment preparing for the move comes from VSI.
- 3. Position sizing respects the trajectory, not just the level. Classical position sizing uses fixed-fractional risk and ATR-based stops. The Second-Derivative refinement: scale position size inversely to VSI. When vol is expanding, reduce position size even if your risk-per-trade percentage is the same — because the distribution of outcomes has fatter tails. This isn't captured by ATR alone; it requires VSI.
06 — Threshold Symmetry
±5% By Design
The EXPANSION and DECAY thresholds are symmetric at +5% and -5% by default. Same philosophy as MPR: symmetry is a constructional guarantee of no directional bias. VSI treats volatility expansion and volatility decay with perfect equivalence. If you observe more EXPANSION than DECAY on your chart, that's a property of the market, not of the indicator.
💡 The Deadband Matters
The ±5% deadband creates a TRANSITION zone at the neutral midpoint. Without it, volMomPct hovering near zero would constantly flip between EXPANSION and DECAY (since momentum always has a sign, it would flicker between +0.1% and -0.1%). The deadband is hysteresis: no commitment to a state change until the reading has meaningfully crossed into the new zone.
07 — Stability Filter
How Often Does Momentum Flip Sign?
Same architecture as MPR. Stability = 1 - flipRate where flipRate is the rolling average of momentum sign changes over 20 bars. When stability drops below 0.5, the classifier forces TRANSITION regardless of the current momentum value. The tool refuses to commit to EXPANSION or DECAY when the underlying signal is too noisy.
💡 No Off-Axis Exemption Needed
Unlike MPR (where trap conditions are exempted from the stability override), VSI has no off-axis state, so the override applies uniformly. Every unstable reading becomes TRANSITION. This is a cleaner design made possible by the simpler state space — one more reason the 3-state model is appropriate for VSI.
08 — The Persistence Contract
Shared With MPR — 3 Bars Before Commitment
VSI inherits the Persistence Contract architecture from MPR. Any proposed state change must appear for persistBars consecutive bars (default 3) before becoming the confirmed state. Flickers less than 3 bars long are filtered out. This gives VSI its committed-state feel: when it reports a state change, the market has actually committed to it.
💡 Why Share the Contract Across Oscillators
Consistency of philosophy. Every committed-state oscillator in the ATLAS suite uses the same 3-bar commitment contract because the reliability tradeoff (2-3 bars of lag for flicker-free states) is universally correct. When you see amber in VSI and amber in MPR, you know they both applied the same anti-flicker logic. This cross-tool consistency is what makes the oscillator stack read coherently.
09 — The Volatility Cycle
Compression → Expansion → Peak → Decay
Over enough time, volatility cycles through recognizable phases: STABLE (compression building), EXPANSION (volatility running), DECAY (volatility contracting from peak), and back to STABLE (new compression). VSI makes this cycle visible as a clean sine-like oscillation between the three states. Markets rarely sit at a single volatility level for long; they cycle.
💡 The Cycle Is Universal
This volatility cycling exists in every market at every timeframe — crypto, equities, FX, commodities, 1-minute to daily. The frequency and amplitude differ, but the pattern is the same. This universality is why VSI works as a single tool across the asset universe with just threshold tuning — the underlying cycle structure is invariant.
10 — Application
Stop Placement Adjustment
The most concrete application of VSI. Instead of a fixed ATR multiplier for stops (e.g., always 2x ATR), scale by state: 1.5x in DECAY (tighter, vol contracting), 2.0x in TRANSITION (baseline), 3.0x in EXPANSION (wider, vol running). This single adjustment turns your stop distance from a static number into a regime-aware envelope — a direct operationalization of the Second-Derivative Principle.
💡 The Symmetric Scaling Rule
Whatever multiplier you use for your baseline stop (TRANSITION state), reduce it by ~25% in DECAY and increase it by ~50% in EXPANSION. These ratios are empirically robust across assets and timeframes. The point isn't the exact numbers — the point is that stop distance should scale with VSI, not stay fixed. A static 2x ATR stop will be systematically too tight in expansion and too loose in decay.
11 — The Oscillator Stack
VSI × MPR × MSI Together
The full oscillator confluence stack. MSI tells you the current market regime. MPR tells you the directional pressure state. VSI tells you the volatility trajectory. These three measure orthogonal properties, so their agreement or disagreement is the read. Full confluence (all three pointing the same way) is rare and powerful. Divergences (one says expansion while another says transition) are where the nuance lives.
💡 Reading the Triad
Strongest setup: MSI = Trend Up + MPR = RELEASE + VSI = EXPANSION. All three pointing same way, breakout environment, committed direction. Weakest setup: MSI = Chop + MPR = TRANSITION + VSI = TRANSITION. All three amber, no commitment anywhere, stand aside. Nuanced setups: MSI = Trend Up + MPR = RELEASE + VSI = DECAY suggests a smooth mature trend (vol compressing into the trend, classic late-stage behavior). Use the triad as a full regime diagnostic, not as a signal generator.
12 — Asset Class Calibration
Default Thresholds Aren't Universal
The ±5% defaults are calibrated for medium-to-high-volatility assets (crypto, equities, commodities). On FX majors like EURUSD, the percent-change cadence of ATR is tighter — 5% moves over 10 bars are rare. Left at defaults, VSI on EURUSD 1H will sit in TRANSITION most of the time, making the tool effectively mute. The fix: tune thresholds to ±3% for FX majors, keeping them symmetric.
💡 The Tuning Rule
The target is to see VSI spend roughly 40-50% of time in TRANSITION and the remainder split between EXPANSION and DECAY. If yours is in TRANSITION 90% of the time, tighten thresholds. If yours is in EXPANSION/DECAY 90% of the time (rare, but happens on very high-vol assets at short timeframes), widen them. ALWAYS keep them symmetric — asymmetric thresholds silently bias your regime classification.
13 — Common Mistakes
Four Ways Traders Misuse VSI
Each traces back to the same root: reading VSI as something it isn't — typically as a direction tool or a momentum proxy — instead of reading it as what it is: a volatility-change oscillator.
Reading VSI as a directional signal
VSI is ENVIRONMENTAL, not directional. EXPANSION doesn’t mean bullish. DECAY doesn’t mean bearish. Volatility expansion happens in both up-trends and down-trends — often MORE in down-trends. Treating VSI as a direction tool will systematically wrong-foot you. It’s a stop-placement/position-sizing input; let MPR and MSI handle direction.
Treating ATR and VSI as the same information
ATR = volatility magnitude. VSI = volatility MOMENTUM. They’re different. An ATR of 200 is meaningless without knowing whether it’s rising, falling, or stable. You can have the same ATR value in a vol-expansion regime (wide stops, breakout prep) or a vol-stable regime (normal stops, baseline trading). The Second-Derivative Principle is exactly the insight that these are different pieces of information.
Using default ±5% thresholds on EURUSD
FX majors have tighter percent-change cadence in ATR than crypto/equities. At ±5% default, VSI on EURUSD 1H will rarely leave the TRANSITION zone — making the tool effectively mute. Tune thresholds to ±3% for FX majors (keep them symmetric) to restore responsiveness. This is a calibration issue, not an architectural issue.
Expecting VSI, MPR, and MSI to always agree
They measure orthogonal properties. VSI = volatility change. MPR = directional pressure. MSI = overall regime. Each is a separate diagnostic axis. Smooth trends often show MPR=RELEASE + VSI=TRANSITION (committed direction, stable vol). News moves often show MPR=RELEASE + VSI=EXPANSION. The PATTERN of agreement/disagreement is the read. Expecting confluence on every bar misses the whole point of stacking independent diagnostics.
14 — Cheat Sheet
VSI In One Page
Architecture
Sub-pane oscillator, derivative measurement (rate of change of smoothed ATR), 3-state classifier, percent-based output.
Three States
EXPANSION (teal, +1) vol rising · DECAY (grey, -1) vol falling · TRANSITION (amber, 0) stable or unstable.
Core Formula
volMomPct = (volSm - volSm[10]) / volSm[10] × 100 where volSm = EMA(ATR, 10).
Thresholds (symmetric, tunable)
±5% default. FX majors: ±3%. Always keep symmetric.
Persistence Contract
persistBars = 3 default. Shared with MPR — same anti-flicker guarantee.
Stability Override
When stability < 0.5, force TRANSITION. No off-axis exemption needed.
Second-Derivative Principle (★)
VSI measures a measurement. ATR tells you where you are, VSI tells you where you’re going. Different information.
Primary Application
Stop scaling: 1.5x ATR in DECAY, 2.0x in TRANSITION, 3.0x in EXPANSION. Position sizing inversely to VSI.
Data Window Exports
ATR raw/smoothed · Momentum % · Stability · State · Per-state booleans.
15 — Scenario Game
Reading VSI as a Volatility Derivative
Five scenarios testing whether you read VSI for what it is — a second-order signal measuring volatility change — or whether you're still pattern-matching it to first-order oscillators.
Round 1 of 5
Score: 0/5
Trader A looks at BTC and sees ATR of 200. Trader B looks at BTC and sees ATR of 200. Both think they're reading the same market. Which trader has the critical missing piece of information?
16 — Knowledge Check
Final Quiz — 8 Questions
Question 1 of 8
Volatility State Index (VSI) measures:
Question 2 of 8
The three VSI states are:
Question 3 of 8
The core VSI calculation is:
Question 4 of 8
Why is ATR smoothed with EMA(10) before the momentum calculation?
Question 5 of 8
Like MPR, VSI uses the Persistence Contract. This means:
Question 6 of 8
When stability < 0.5, VSI forces:
Question 7 of 8
What’s the correct operational response to VSI reading EXPANSION?
Question 8 of 8
For FX majors (EURUSD, GBPUSD), the default ±5% thresholds: