True Strength Index (TSI): Double-Smoothed Momentum for Cleaner Signals

William Blau wasn't trying to invent a flashy new indicator. He was trying to fix an old one. Raw momentum — the simple difference between today's close and the close N bars ago — is useful but noisy. One-layer smoothing (a single EMA applied to momentum) helps, but still lets short-term spikes bleed through. Blau's insight was that applying a second EMA to already-smoothed data removes a fundamentally different type of noise without destroying the underlying signal. He published this concept in his 1991 article and later expanded on it in his 1995 book Momentum, Direction and Divergence.

Here's how the calculation works, step by step:

Step 1 — Calculate price change (PC). For each bar, subtract the previous close from the current close. If EUR/USD closes at 1.0920 today and closed at 1.0915 yesterday, the price change is +0.0005. Simple momentum.

Step 2 — First smoothing. Apply a 25-period EMA to the price change series. This produces a smoothed momentum line that removes the shortest-term noise. Blau chose 25 periods because it approximately covers one month of daily data — long enough to establish directional context.

Step 3 — Second smoothing. Apply a 13-period EMA to the result from Step 2. This second pass targets the medium-frequency noise that survived the first smoothing. The 13-period setting covers roughly half a month, creating a nested smoothing structure that progressively filters out noise at different frequencies.

Step 4 — Normalize. Repeat Steps 1-3 using the absolute value of price change instead of the raw price change. This gives you the double-smoothed absolute momentum. Then divide the double-smoothed PC by the double-smoothed absolute PC, and multiply by 100.

The normalization step is what bounds the TSI between -100 and +100 and makes it comparable across instruments. A TSI reading of +35 on EUR/USD and +35 on gold means the same thing: momentum is 35% of its maximum possible strength in the bullish direction over the lookback window. This is a significant advantage over raw momentum, which produces readings in price units that can't be compared across different markets.

Why 25 and 13 specifically? Blau selected these as defaults because they correspond to common trading cycles — roughly one month and half a month on daily charts. But they're not sacred numbers. Here's how adjustments affect the indicator:

The faster settings (13/7/5) work better on intraday charts like H1 where you need quicker responses. The default 25/13/7 is ideal for H4 and D1 swing trading. The slower settings suit weekly charts or position trading where you want to ignore everything except major trend shifts.

One thing worth understanding: the double-smoothing process introduces lag. The TSI will always respond more slowly than a raw momentum indicator or a single-smoothed oscillator like RSI. That's the fundamental tradeoff Blau accepted — you get fewer false signals, but you also get later signals. For swing traders on H4 and D1, this tradeoff is generally favorable because the reduction in noise more than compensates for the slight delay. For scalpers on M5 who need split-second timing, the lag becomes a liability.

The normalization against absolute price change is the mathematical trick that makes the whole thing work. Without it, you'd have a smoothed momentum line that oscillates around zero with no fixed boundaries — useful, but hard to judge whether a reading is extreme. By dividing by the absolute movement, Blau created a ratio that tells you what percentage of total price movement has been directional. A TSI of +60 means 60% of all price movement over the smoothing window has been upward. A TSI of -20 means 20% has been downward on net. That's genuinely informative.

The TSI gives you two separate signal systems on one indicator panel: zero line crossovers for directional bias, and signal line crossovers for timing. Understanding when to use each — and when to combine them — separates mechanical chart-reading from intelligent analysis.

Zero line crossovers work like a directional filter. When the TSI crosses above zero, double-smoothed momentum is positive — meaning the trend of price changes, after all that smoothing, is net upward. When it crosses below zero, the trend is net downward. This is the TSI equivalent of asking "which team is winning?" It doesn't tell you the score or whether the game is about to turn, just who has the current advantage.

On USD/JPY D1, imagine the TSI has been negative for three weeks during a steady downtrend. Price drops from 152.80 to 148.50 and the TSI bottoms at -28. Then the selling slows, buyers start stepping in, and the TSI gradually rises toward zero. The moment it crosses above zero, the double-smoothed momentum has officially flipped positive. This crossover confirms that the downtrend has lost its structural momentum — not just paused, but actually reversed at the smoothed level. Because of the double smoothing, this zero cross happens later than it would on raw momentum, but it's also far less likely to be a fake-out.

The problem with zero line crossovers alone is that they're binary and slow. You know the direction shifted, but the cross often happens after a significant portion of the move has already occurred. Enter the signal line.

The signal line is a 7-period EMA of the TSI itself. It smooths the TSI one more time — yes, that's triple smoothing of the original data — and serves as a faster trigger line. When the TSI crosses above its signal line, short-term momentum is accelerating relative to the recent trend. When it crosses below, momentum is decelerating.

Here's where it gets practical. Signal line crossovers above the zero line are bullish continuation signals. They suggest a pullback within an uptrend is ending and momentum is resuming. These are higher-probability setups because you're trading in the direction of the established trend. On GBP/USD H4, if the TSI is at +18 and crosses above its signal line, that's a buy signal within a bullish context. The trend is up, and the recent pause is over.

Signal line crossovers below the zero line are bearish continuation signals. The TSI at -22 crossing below its signal line means selling pressure is re-accelerating within a downtrend. These are your short entry triggers in a confirmed bearish regime.

The highest-probability signal combines both layers: a zero line cross confirmed by a signal line cross in the same direction. When the TSI crosses above zero AND is above its signal line, you have both the directional shift and the momentum acceleration agreeing. This double confirmation filter eliminates most of the marginal signals that plague single-indicator systems.

A step-by-step example on EUR/USD H4: The TSI has been below zero for two weeks. It begins rising from -32 and crosses its signal line at -15 — an early warning that selling pressure is fading. Seven bars later, the TSI crosses above zero at +3, still above the signal line. Now you have the full confirmation: trend has shifted (zero cross) and momentum is accelerating in the new direction (signal line position). This is your entry zone. Your stop goes below the most recent swing low, and your target is the next resistance level.

Overbought and oversold zones add one more layer. While the TSI technically ranges from -100 to +100, extreme readings are rare. Blau suggested treating +25 and -25 as overbought and oversold thresholds. Most TSI readings on forex pairs fall between +30 and -30. When the TSI reaches +25 or higher, the trend is strong but may be extended. When it drops to -25 or lower, selling has been aggressive and a bounce becomes increasingly probable.

But here's the nuance that catches beginners: an overbought reading doesn't mean "sell." In strong trends, the TSI can stay above +25 for extended periods, generating signal line crossovers above the overbought zone that are actually excellent continuation entries. Only when the TSI starts dropping from overbought territory and crosses below the signal line should you consider the extension as a potential exhaustion signal. The same applies in reverse for oversold readings.

For MetaTrader 5 users: the TSI is not included by default. You'll need to add it from the Code Base or custom indicators library. TradingView has it built in under "True Strength Index." The default settings on both platforms typically use 25/13/7, matching Blau's original parameters.

Divergence is where the TSI earns its keep compared to noisier oscillators. Because the double smoothing filters out minor price fluctuations, the divergences that do appear on the TSI tend to be genuine structural shifts rather than noise-induced mirages. When you spot a TSI divergence on H4 or D1, it carries significantly more weight than the same divergence pattern on raw momentum or even RSI.

Regular bullish divergence forms when price makes a lower low while the TSI makes a higher low. The market printed a new extreme in price, but the momentum behind that push was weaker than the previous one. The bears swung hard, but the bat connected with less force. On AUD/USD D1, picture price dropping from 0.6520 to 0.6310 with the TSI bottoming at -31. Price bounces to 0.6420, then drops again to 0.6280 — a lower low. But the TSI at that second bottom reads only -22 — a higher low. The selling pushed price further, but the double-smoothed momentum says the move was weaker. That's your setup.

Regular bearish divergence is the mirror: price makes a higher high while the TSI makes a lower high. EUR/USD D1 pushes to 1.1050 with the TSI peaking at +28. Three weeks later, price edges to 1.1085 — a higher high. But the TSI only reaches +19. The rally looks healthy on the price chart, but the momentum engine is producing less power on each push. The trend isn't dead yet, but it's running on fumes.

Hidden divergence signals continuation rather than reversal. Hidden bullish divergence occurs when price makes a higher low (respecting the uptrend) but the TSI dips to a lower low. This looks alarming — momentum hit a new low! — but it actually means the pullback reset the indicator deeply while price held its structural support. It's a coiled spring. When the TSI turns back up, the trend often resumes with force. These hidden divergence setups are favorites among swing traders because they offer entries within established trends at favorable prices.

Now, here's the critical rule that will save your account: divergence is a warning, not a trigger. A divergence tells you the trend is weakening, but "weakening" and "reversing" are not the same thing. A trend showing bearish divergence on D1 can grind higher for another two or three weeks before actually turning. If you short the moment you spot the divergence, you'll often get run over by the last leg of the trend.

The solution is to require confirmation before acting on any divergence:

Confirmation Method 1 — Signal line crossover. After spotting a bearish divergence, wait for the TSI to cross below its signal line. This confirms that momentum has actually started to decline, not just flatten. On GBP/JPY H4, a bearish divergence followed by a signal line crossover within 5-8 bars is a high-probability short setup.

Confirmation Method 2 — Zero line break. For stronger conviction, wait for the TSI to cross the zero line in the divergence direction. If you see bullish divergence, wait for the TSI to cross above zero before entering long. This is the most conservative approach — you'll miss the first portion of the move, but you'll avoid the false divergences that never materialize into actual reversals.

Confirmation Method 3 — Price structure break. Pair the TSI divergence with a trendline break or support/resistance breach on the price chart. Bearish divergence combined with a break below a rising trendline is one of the most reliable reversal setups in technical analysis. The divergence tells you the engine is failing; the trendline break tells you the wheels have come off.

A complete divergence trade on USD/CAD H4: Price makes a higher high at 1.3680 while the TSI forms a lower high (bearish divergence). You don't short immediately. Instead, you wait for the TSI to cross below its signal line, which happens three bars later at 1.3655. Your entry is the close of that bar. Stop loss goes above the recent high at 1.3695. Target is the nearest H4 support at 1.3520. Risk-to-reward: approximately 1:3.4.

The timeframe matters enormously for divergence reliability. On M15, you'll find TSI divergences constantly — many of which resolve with a 15-pip blip before the trend continues. On H4 and D1, divergences are rarer but significantly more meaningful because the double smoothing has already filtered out the noise that creates false divergences on lower timeframes. For swing trading, H4 is the minimum timeframe for acting on TSI divergence. D1 is optimal if your holding period allows for multi-day trades.

One more nuance: multi-swing divergence is stronger than two-point divergence. If price makes three successively higher highs while the TSI makes three successively lower highs, the bearish divergence carries more weight than a simple two-point version. The trend has been weakening progressively across multiple impulse waves, not just one. These extended divergences often precede the largest reversals.

Theory is nice, but you're here to trade. Let's build a complete TSI-based swing trading setup that works on H4 and D1 charts with clear rules for entry, stop loss, target, and position management. This isn't the only way to use the TSI, but it's a structured approach that leverages the indicator's strengths — smooth momentum reads and reliable signal line crossovers — while managing its weakness, which is lag.

The Setup: TSI Trend-Following Swing Strategy

Timeframe: H4 or D1 Indicator settings: TSI (25, 13, 7) Trend filter: 50-period EMA on the price chart Entry trigger: TSI signal line crossover in the direction of the trend Confirmation: TSI above zero for longs, below zero for shorts

Entry Rules — Long:

1. Price is above the 50 EMA (uptrend confirmed)
2. TSI is above zero (momentum confirms the trend)
3. TSI has pulled back toward or below the signal line during a price pullback
4. TSI crosses back above the signal line (momentum resumes)
5. Enter on the close of the crossover bar

Entry Rules — Short:

1. Price is below the 50 EMA (downtrend confirmed)
2. TSI is below zero (momentum confirms the trend)
3. TSI has rallied toward or above the signal line during a price bounce
4. TSI crosses back below the signal line (selling resumes)
5. Enter on the close of the crossover bar

Let's walk through a full long trade on EUR/USD H4:

The 50 EMA is rising and price sits above it at 1.0840. The TSI reads +14, above both zero and the signal line — momentum is bullish. Price pulls back from 1.0880 to 1.0830 over three days. During this pullback, the TSI dips from +21 to +8, crossing below its signal line at +11. This is the "reload" phase — the trend is intact but momentum has paused. Two bars later, the TSI crosses back above its signal line at +12. Entry on the close of that bar at approximately 1.0845. Stop loss below the pullback low at 1.0810 (35 pips). First target at the previous swing high of 1.0880 (35 pips, 1:1). Second target at 1.0930 if the trend extends (85 pips, 1:2.4).

Stop Loss Placement:

Place your stop below the most recent swing low for longs (above the recent swing high for shorts). Don't use a fixed pip value — the market's volatility determines the appropriate stop distance. On H4 EUR/USD, a typical pullback swing is 30-50 pips, so your stop will usually fall in that range. On D1, expect 60-120 pip stops. If the stop distance makes the position too large for your risk tolerance (1-2% per trade), reduce position size rather than tightening the stop.

Trade Management:

Once the trade is open, the TSI continues providing useful management information:

• TSI rising and above signal line: The trade is working. Hold the position and let it run.
• TSI flattening near +25 or higher: Momentum may be peaking. Consider taking partial profits (50% at Target 1).
• TSI crosses below signal line while still above zero: Momentum is fading but the trend is intact. Trail your stop to break-even or the most recent minor swing low.
• TSI crosses below zero: The trend structure has broken. Exit remaining position regardless of where price sits relative to your target.

Multi-Timeframe Enhancement:

For higher-probability entries, add a D1 directional filter to your H4 entries:

1. Check the D1 TSI direction: Is it above zero and above its signal line? If yes, the higher timeframe supports bullish trades.
2. Drop to H4 for entry timing: Wait for the H4 TSI signal line crossover as described above.
3. This alignment — D1 momentum bullish + H4 entry trigger — filters out counter-trend trades that work against the larger cycle.

What to Avoid:

This strategy struggles during sideways markets. When the 50 EMA is flat and price is chopping above and below it, the TSI will produce signal line crossovers in both directions with little follow-through. The fix is simple: if the 50 EMA has been essentially flat (slope near zero) for more than 20 bars, step aside and wait for a new trend to establish.

Also avoid entering when the TSI is already at extreme readings (+25 or above for longs, -25 or below for shorts) at the time of the signal line crossover. A signal line crossover at TSI +28 means you're joining the trend after it's already run hard. The best entries happen when the crossover occurs with the TSI between +5 and +20 for longs — enough to confirm the trend direction but not so far into overbought that the easy part of the move is behind you.

Expected Performance Characteristics:

This is a trend-following system, which means it wins less than half its trades but the winners are larger than the losers. Expect a win rate around 40-50% with an average winner roughly 1.5-2x the average loser. The edge comes from the positive risk-reward ratio, not from being right most of the time. Strings of losses during choppy markets are normal — the 50 EMA filter and the zero-line requirement keep these drawdowns manageable.

The TSI and MACD are often mentioned in the same breath because they share a family trait: both apply multiple layers of exponential smoothing to price data. But the way they apply that smoothing, and what they measure underneath it, leads to meaningfully different behavior on your chart. Understanding the distinction helps you pick the right tool — or decide to use both.

What MACD Actually Measures:

MACD takes two EMAs of price itself (typically 12 and 26 periods), subtracts the longer from the shorter, and plots the difference. The MACD line shows whether short-term price is above or below long-term price — essentially a smoothed measure of trend displacement. The signal line (9-period EMA of MACD) provides crossover triggers. The histogram shows the distance between MACD and its signal line.

MACD's "double smoothing" is indirect: each EMA is a single smoothing of price, and the MACD line represents the difference between two single-smoothed values. The signal line adds a third layer of smoothing.

What TSI Actually Measures:

TSI applies two sequential EMAs directly to momentum (price change), not to price itself. It smooths the speed of price movement, then smooths it again. The normalization against absolute price change converts this into a bounded percentage measure of directional strength.

The key difference: MACD measures trend displacement (where price is relative to its own averages). TSI measures trend velocity (how fast price is changing, double-smoothed). Displacement and velocity are related but not identical. A car can be far from its starting point (high displacement, high MACD) but slowing down (low velocity, declining TSI). Conversely, a car can be near its starting point but accelerating rapidly (low displacement, rising TSI).

Practical Behavior Differences:

In a trending market, MACD tends to produce signals earlier because it responds to price position relative to moving averages, which shifts as soon as price starts moving. TSI responds to the change in price change — a second derivative of sorts — which takes longer to register but confirms the move more reliably.

Consider a scenario on GBP/USD H4: price breaks above a consolidation range and starts trending up. MACD will cross its signal line early in the move as the 12 EMA separates from the 26 EMA. TSI will confirm a few bars later, once the double-smoothed momentum calculation registers the sustained directional change. If the breakout is genuine, both signals lead to profitable trades, but MACD gets you in earlier. If the breakout is false and price reverses back into the range, MACD triggers a losing trade while TSI's slower response may have kept you on the sidelines.

In a choppy, range-bound market, this dynamic reverses. MACD's faster response produces more whipsaw crossovers — the 12 and 26 EMAs weave back and forth as price oscillates, and the MACD line dances above and below zero with each swing. TSI's heavier smoothing absorbs much of this noise, producing fewer crossovers during the chop. If you've ever complained that MACD generates too many false signals during ranges, TSI was literally designed to solve that problem.

Divergence Detection:

Both indicators can identify divergence, but TSI divergences tend to be cleaner. Because the TSI line is smoother, its peaks and troughs are more distinct and easier to connect with trendlines. MACD divergences can be muddied by the histogram and by minor oscillations in the MACD line that create ambiguous peaks. When scanning for divergence on D1 charts, TSI typically produces fewer but more reliable divergence signals.

When to Use Each:

• Use TSI when you're swing trading on H4/D1 and want fewer, higher-conviction signals. Use TSI when your primary concern is avoiding false signals in volatile or choppy conditions. Use TSI when you need to compare momentum strength across different instruments.

• Use MACD when you need faster signal generation and are willing to manage more false positives. Use MACD when you're day trading on H1 and the additional lag of TSI costs you too much of each move. Use MACD when you prefer visual simplicity — the histogram makes momentum shifts instantly visible.

• Use both when you want a confirmation system. MACD provides the early signal; TSI confirms it. If MACD crosses its signal line bullish and the TSI is above zero and also crosses its signal line bullish, you have strong momentum agreement from two different calculation methodologies. Signals where both indicators agree are meaningfully more reliable than either indicator alone — you're basically filtering displacement momentum through velocity momentum.

Here's a practical combo framework for H4 swing trading:

1. MACD crosses signal line bullish — place the pair on your watchlist
2. Wait for TSI to cross above its signal line with TSI above zero — entry trigger
3. If TSI confirms within 5 bars of the MACD signal — strong setup, full position
4. If TSI has not confirmed after 8 bars — the move may lack momentum, skip it

This approach uses MACD as a screening tool and TSI as the execution trigger. It leverages MACD's speed for early detection and TSI's smoothness for filtering false signals before you commit capital. It won't catch every move — nothing does — but the moves it catches tend to follow through with conviction.