VolBand Pro Volatility Forecast Bands

How far will the market move in the next hour? VolBand Pro forecasts the range of the next hour with calibrated 80% and 95% bands and a ready-to-use stop distance. Validated on 18 forex pairs, M15/M30/H1.

VolBand Pro does not tell you where price is going. It tells you how far it is likely to move, and how sure that estimate is.

For the next hour (M15, M30) or the next four hours (H1), VolBand Pro shows:

• the expected range (median high-low distance),
• an 80% band and a 95% band around it,
• a suggested stop distance for risk-based position sizing.

The forecast is fixed at bar close and never repainted.

Why not ATR?

ATR looks back at recent bars and projects them forward. That is why it is too small right before London or New York opens, and too large in the quiet hours after them.

VolBand Pro knows the time of day. It models how each currency pair behaves in each intraday slot, adjusts for today's and this week's activity, and then calibrates its bands on its own recent errors (adaptive conformal prediction). The bands widen before the session opens, not after the move has started.

Validation (published protocol, pre-registered hypotheses)

• Test set: 18 currency pairs that were not used to build or tune the model, from January 2024 to October 2026, about 17,500 forecasts per pair.
• Calibration:
  • M15: the 80% band contained the realized range 79.5–80.4% of the time, and the 95% band 94.8–95.3%, on 18 of 18 pairs.
  • M30 and H1: calibration within tolerance on 18 of 18 pairs.
• Accuracy vs. baselines (interval score, lower is better):
  • vs. ATR, with identical calibration: better on 18/18 pairs, 16–36% lower error on M15.
  • vs. a time-of-day seasonal model, with identical calibration: better on 18/18 pairs, median error reduction 8.7% on M15, 9.2% on M30, 4.6% on H1.
• Stop distance: with random entries and equal average stop size, the VolBand stop-out rate stayed closer to its target across hours of the day and quiet/volatile days than ATR-based and seasonal stops, on 18 of 18 pairs (M15, M30, H1).
• MT5 implementation: checked bar-by-bar against the reference research code on more than 570,000 forecasts (EURUSD and USDJPY, M15). Maximum difference: 3e-11.

How traders use it

• Stop placement: a stop sized to the volatility expected right now, not to the volatility of the last 14 bars.
• Position sizing: keep the money risk constant while the stop adapts to the session.
• Target reality check: if your target lies beyond the 95% band, you are asking the market for a move that happens less than 1 time in 20 within the horizon.
• Cost check: when the expected range is only a few pips, spread and commission take a large share of it.
• Context: a realized range breaking above the 95% band signals unusual activity (news, jumps).

Inputs

The model parameters are fixed to their validated values. The horizon is set automatically:

• M15: 4 bars
• M30: 2 bars
• H1: 4 bars

For Expert Advisors

All values are available through iCustom / CopyBuffer , in price units:

The same values are also published as terminal global variables ( VOLB1_<symbol>_<period>_… ). Quantum Risk Manager Pro (from version 1.002) can use the VolBand Pro stop directly for its lot calculation.

Limitations (please read)

• VolBand Pro forecasts range, not direction. It gives no buy or sell signals.
• Validated on forex pairs, M15, M30 and H1 only. On other symbols or timeframes the indicator works, but marks the output as "not validated".
• Warm-up: it needs about 250 trading days of history before the first forecast.
• On days when volatility shifts abruptly, ATR adapts faster than VolBand Pro. The advantage of VolBand Pro is the time-of-day structure.
• Using the VolBand stop does not reduce drawdown compared with an ATR stop in our tests. It makes the stop-out frequency more consistent, which is a different property.
• Results on historical data do not guarantee future results. Trading involves risk.