Built on William Blau's True Strength Index — a double-smoothed momentum oscillator that delivers cleaner divergence reads and earlier signal-line crossovers than MACD, with a color-coded bar plot that reveals momentum phases at a glance.
Section 01 — Overview
BWT Precision Oscillator is built on the True Strength Index — William Blau's 1991 momentum oscillator introduced in Stocks & Commodities magazine and detailed in his book Momentum, Direction, and Divergence. TSI takes raw price momentum (1-bar net price change) and applies two consecutive EMAs, dividing the double-smoothed momentum by the double-smoothed absolute momentum. The result is a bounded oscillator that captures genuine directional pressure while filtering out the high-frequency noise that obscures shorter momentum reads.
The double-smoothing design gives TSI a unique combination of properties. It responds faster than MACD on a comparable lookback because the second smoothing pass acts on the first smoothed value rather than on raw price — the lag compounds less aggressively. It also oscillates more cleanly than RSI or single-smoothed momentum because the high-frequency wiggles that obscure peak/trough structure are filtered before the oscillator value is computed. The result is one of the most readable momentum charts in technical analysis: divergences are obvious, reversals print clearly, and trend-phase shifts are unambiguous.
A signal line — an EMA of the TSI — provides discrete crossover triggers (the same role MACD's signal line plays for that oscillator). Color-coded bars distinguish bullish vs bearish momentum phases, and the centerline (zero) divides the chart into a bullish bias zone above and a bearish bias zone below. The combined read — TSI position relative to zero, TSI vs signal line, color-coded bar phase — encodes more decision-relevant momentum information than any single oscillator can deliver alone.
Screenshot
Image provided by you
Section 02 — Concept Reference
A working glossary of every component, signal, and pattern this oscillator surfaces. The TSI's double-smoothed design produces some of the cleanest reads in the entire momentum family.
William Blau's 1991 oscillator: TSI = 100 × (EMA(EMA(momentum, slow), fast) / EMA(EMA(|momentum|, slow), fast)). The double-smoothing isolates directional momentum from absolute momentum, producing a bounded reading that responds faster than MACD on equivalent lookbacks while filtering the noise that obscures single-smoothed oscillators. Blau introduced it as a "true" strength index in contrast to RSI's purely magnitude-based read.
The defining design choice. Single smoothing (a single EMA on momentum) preserves too much noise to reliably surface divergences. Double smoothing applies a second EMA pass to the already-smoothed momentum, removing high-frequency wiggle while preserving the directional shape. The lag cost is materially lower than two equivalent simple averages chained together because the second EMA acts on already-trended data.
An EMA of the TSI itself, plotted as a separate line. Crossings of TSI through its signal line are discrete trigger events: TSI crossing above signal = bullish trigger; below = bearish trigger. Used the same way MACD's 9-EMA signal is used — but the underlying oscillator is double-smoothed and bounded, producing fewer false crosses.
The most important level on the entire oscillator. TSI above zero = net positive momentum across the smoothing window = bullish bias. Below zero = net negative momentum = bearish bias. A clean cross of zero in the direction of HTF trend is a continuation trigger; a cross against trend is the first warning of a regime change. Always check zero-line position before reading anything else on the oscillator.
Histogram-style bars colored by momentum phase: bullish color when TSI is above signal line (or above zero, depending on configuration), bearish color otherwise. The visual phase encoding lets you read momentum direction at a single glance without parsing two oscillator lines. Color shifts mark the same events as line crossovers but are far easier to scan across a multi-chart layout.
TSI's double-smoothing makes divergences cleaner and more readable than on faster oscillators. Standard divergence (price HH + TSI LH for bearish; price LL + TSI HL for bullish) signals reversal. Hidden divergence (price HL + TSI LL bullish; price LH + TSI HH bearish) signals continuation. Because TSI filters intra-bar noise, the divergent peaks and troughs are cleaner than on RSI or stochastics.
Both are momentum oscillators with signal lines, but they compute differently. MACD = EMA(price, fast) − EMA(price, slow); it measures separation between two price averages, is unbounded, and responds slowly. TSI = double-smoothed momentum normalized by absolute momentum; it is bounded (typically −100 to +100), responds faster on equivalent lookbacks, and produces cleaner divergences. For divergence trading, TSI generally outperforms MACD.
Momentum oscillators detect a change in the rate of price change — which by definition occurs before price reverses direction. TSI's double-smoothing makes that lead time more reliable because the noise filter prevents false-positive lead reads from single-bar wicks. In practice, a clean TSI signal-line cross typically precedes price-action confirmation by 1–3 bars on the chart's native timeframe.
Section 03 — Workflow
Every clean Precision Oscillator setup follows this sequence. The zero-line bias check before signal-line cross is what separates productive entries from random crossover noise.
Section 04 — Parameters
Blau's original TSI specification used (25, 13) for the slow/fast smoothing pair. The signal line is typically a 7- or 9-period EMA of the TSI. These defaults are robust on most timeframes; tune only after confirming standard settings produce good results on the instrument you trade.
| Parameter | Default | Description |
|---|---|---|
| Fast | — | Period for the second (faster) EMA smoothing pass — Blau's original is 13 |
| Slow | — | Period for the first (slower) EMA smoothing pass — Blau's original is 25 |
| Signal Length | — | EMA period for the signal line that produces crossover triggers — typical values 7 or 9 |
Section 05 — Trade Setups
These are the named, repeatable Precision Oscillator entry models. Each one specifies the context, trigger, entry, stop, and target — removing the moment-to-moment discretion that destroys most momentum-based trading.
The cleanest TSI entry available. In a confirmed HTF uptrend, TSI pulls back below zero on a price retracement, then crosses back above zero as price resumes higher. The zero-line cross in the direction of HTF trend is a high-conviction continuation signal because both price and momentum have completed a reset and aligned with the dominant move. Mirror image for downtrends.
The discrete entry trigger. TSI crosses above its signal line while both are positioned to support the trade direction (above zero for longs, below for shorts). The cross is the binary trigger; bar-color phase shift confirms the cross is real, not a brief overlap. Always require zero-line agreement — a cross above signal while both are deep below zero is a counter-trend bounce, not a clean long.
TSI's signature setup. Price prints a higher high (lower low) but TSI prints a clearly lower high (higher low) at structure. Because the oscillator is double-smoothed, the divergent peaks are much cleaner than on RSI or stochastics — single-bar wicks don't create false-positive divergences. The cleaner the visual divergence, the more reliable the signal.
An aggressive precursor to the zero-line cross. In a confirmed uptrend, TSI dips toward zero on a pullback but hooks back up before crossing — the indicator visually rejects the zero line. The hook signals that the pullback was insufficient to break momentum and the trend is about to resume. Provides an earlier entry than waiting for the full cross.
The highest-quality TSI configuration. HTF TSI position defines bias (above zero = bull bias, below = bear). LTF TSI signal cross in agreement with HTF bias provides the entry trigger. The multi-timeframe alignment requirement filters out the bulk of false LTF signals — a 1m signal cross while the 15m TSI is rolling against the trade is exactly the LTF setup that fails most often.
A trend re-entry pattern. TSI in an established uptrend regularly pulls back to test the zero line and resumes higher; the test-and-reject is itself the trade. The setup is similar to the zero-line cross but does not require the full cross — the trade is taken on the rejection of zero from above (longs). Requires confirmed trend; in a range or reversal phase, the pullback to zero often becomes a full cross.
Section 06 — Best Practices
These practices distill the most reliable rules from William Blau's original TSI work and from the practical experience of traders who use it as a primary momentum read. Each one is a filter that improves trade quality.
TSI above zero is bullish bias; below is bearish bias. This single read defines whether subsequent crossovers are continuation triggers or counter-trend bounces. A signal cross above signal line while both are deep below zero is not a long — it is a relief rally inside a downtrend. Read zero-line first, every time.
A TSI/signal crossover with no nearby price level is a wiggle. Pair every cross with a structural reason — prior swing, key MA, BWT Core Level, ICT Order Block, opening range boundary. The oscillator gets you a directional read; the structural level gets you a tradable setup.
The double-smoothing design produces noticeably cleaner divergences than RSI or stochastics. Most TSI literature emphasizes signal-line crossovers, but in practice divergence-anchored trades produce the strongest statistical results. Train your eye to spot price/TSI divergence first; treat the cross as the trigger that resolves the divergence.
A common mistake is treating TSI like MACD with the same defaults. TSI's second smoothing pass acts on already-trended data, so the effective lag is materially lower than two-stage MACD smoothing. Use comparable or slightly faster periods than MACD defaults to take advantage of TSI's lag-efficiency, not the same periods.
TSI excels when it confirms a signal generated by another tool — a price-level setup, an ICT structural break, an opening range breakout. Standalone TSI trades from random crossovers underperform substantially. Build setups where price action defines the trade and TSI gates whether it's high-conviction or low.
In a multi-chart layout, bar color phase shifts are visible instantly while line crossovers require careful inspection. Use the color phase as your alert; use the line cross as the precise trigger. Color flip without subsequent cross is a fakeout; cross with color confirming is real.
Blau's (25, 13) defaults work well on swing-trading timeframes (4H, daily) but may be too slow for active intraday trading on volatile instruments. For 5m futures consider faster slow/fast (often 13/7 or similar). For very slow instruments (low-volatility currencies) lengthen toward 30/15. Backtest period changes deliberately.
When TSI is well above zero with a clear upward slope, mechanical short signals from secondary indicators are very low-probability. Use TSI as a veto filter: if TSI directly contradicts the proposed trade direction with conviction, skip the setup. The veto rule alone eliminates most low-quality entries.
A LTF TSI divergence at a HTF structural level (prior daily swing, 4H supply zone, key weekly level) is a substantially higher-probability setup than the same divergence in mid-range. The combination of momentum exhaustion (divergence) plus price-level reaction (HTF structure) is exactly what makes a high-conviction reversal entry.
A cross of TSI back through its signal line in the opposite direction of an open trade is a meaningful momentum exhaustion signal — useful as an early exit trigger before price-action stops are hit. Particularly valuable on quick winners where banking partial profits on momentum reversal preserves R when followed by a price retrace.
Section 07 — Common Mistakes
These are the recurring failure modes documented across decades of TSI and momentum-oscillator trading. Avoiding them is, on its own, a substantial edge.
Mid-range crosses with TSI hovering around zero produce dozens of crosses per session and almost all are noise. Crosses are meaningful when paired with zero-line bias agreement, structural levels, or divergence — not in isolation.
A bullish cross while TSI is deep below zero is a counter-trend bounce, not a clean long. Always check zero-line position first — it tells you whether the cross is continuation or counter-trend and that distinction governs the trade's expected payoff.
TSI does not have hard OB/OS levels like RSI — but traders sometimes mentally impose +50/−50 as fade levels. Don't. A strongly trending market produces a strongly trending TSI; fade signals based on the oscillator level alone fail repeatedly.
TSI alone tells you direction, not where to enter. Without a price-level reason — prior swing, key MA, OB — the trade has no defined risk anchor and the setup degrades to a pure momentum gamble. Always require structure.
A bar-color flip without a corresponding line cross is a brief overlap, not a real momentum phase change. Use color as the visual alert; require the line cross to confirm. Color-only trades produce many small chops.
Running TSI at very short fast/slow values (e.g., 5/3) on a 1-minute chart obliterates the noise-filtering benefit of double-smoothing — you may as well use a single oscillator. Stay close to Blau's defaults unless you have a tested reason to deviate.
MACD and TSI compute differently. Using MACD defaults (12, 26, 9) on TSI typically produces too-slow signals because TSI's smoothing structure is more lag-efficient. Use TSI-native defaults (Blau's 25/13/9 or your tuned variant).
A TSI signal cross against a strong HTF trend, with no price level to anchor the entry, is one of the lowest-probability setups available. Counter-trend trades require multiple confirmations — divergence, structure, HTF exhaustion — never just the cross.
BWT Precision Indicators require a valid BWT license for NinjaTrader 8. The True Strength Index was developed by William Blau and introduced in Stocks & Commodities magazine in 1991, with full treatment in his book Momentum, Direction, and Divergence. The signal-line and divergence framework described on this page is derived from publicly available work on the TSI. This page is provided for informational and educational purposes only and is not trading advice. Trading futures and other leveraged products involves substantial risk of loss and is not appropriate for all investors. Past performance is not indicative of future results.