True Strength Index, usually shortened to TSI, is a momentum oscillator designed to smooth price change enough to reduce noise while still reacting when momentum meaningfully shifts. It is built from the idea that momentum is the direction and persistence of price changes, not the price level itself. TSI compresses those changes into a bounded style output that traders can compare across instruments and timeframes.
TSI is most useful when you treat it as a momentum confirmation tool rather than a standalone buy sell machine. It helps you judge whether a push higher is supported by strengthening momentum or whether price is rising while momentum is fading. Used correctly, it can reduce chasing and can improve timing by adding structure to “momentum is improving” versus “momentum is weakening.”
How it’s calculated
TSI starts with momentum, typically the close to close change, then applies smoothing twice to both momentum and absolute momentum. The ratio of the smoothed momentum to the smoothed absolute momentum is then scaled, which makes the output comparable across markets with different volatility. Many charting platforms also plot a signal line, which is a moving average of TSI.
m_t=Close_t-Close_{t-1} TSI_t=100\times\frac{EMA_s!\left(EMA_r(m_t)\right)}{EMA_s!\left(EMA_r(|m_t|)\right)} Signal_t=EMA_k(TSI_t)In these formulas, m_t is the one bar momentum, |m_t| is absolute momentum, and EMA_r means an exponential moving average with period r. The two smoothing periods are commonly called r and s, where r is the faster smoothing and s is the slower smoothing applied to the already smoothed series. The signal line period k is usually shorter and is used for crossovers and slope confirmation.
Most used settings and why traders choose them
The most common default you will see is 25 and 13 for the two smoothing periods, often written as TSI 25 13. That combination is popular because it is slow enough to filter a lot of chop but still responds when momentum shifts over multi week swings on daily charts. On intraday charts it tends to behave more like a trend momentum filter than a fast trigger.
A frequent variation is to adjust the faster period first, because that changes responsiveness without completely changing the character of the indicator. Lowering r makes TSI more reactive and increases the number of crossovers and centerline flips, which can help in strong trends but tends to raise false signals in ranges. Increasing r reduces whipsaws but can make entries later, so it often pairs better with trend following approaches that do not need perfect timing.
Signal line settings vary widely, but a common choice is 7 or 13 depending on how smooth you want confirmation to be. A shorter signal line emphasizes early turns and more crossover signals, which is useful only if you also use a price based filter. A longer signal line reduces noise and makes crossovers more meaningful, but you will give back more of the early move.
If you want a baseline to start from, keep the common TSI 25 13 and add a signal line like 7. If your goal is to use TSI mainly as a trend momentum confirmation, focus on the centerline and the slope rather than frequent signal line crossovers. If you want more timing help, keep the same TSI settings and adjust only the signal line first.
How it behaves on charts
TSI tends to swing above and below a centerline, where values above zero indicate that smoothed momentum is positive and values below zero indicate that smoothed momentum is negative. Unlike some oscillators that are interpreted mainly by overbought and oversold zones, TSI is often more effective when read as momentum direction and momentum persistence. The most practical question is whether TSI is rising or falling while price is attempting to trend.
In uptrends, TSI often stays above zero for extended stretches, with pullbacks that dip toward zero and then turn up again. Those turns can align with continuation points if price structure supports the trend. In downtrends, the mirror image is common, with rallies that lift TSI toward zero and then roll back down.
Signal line crossovers can be useful, but they are context dependent. Crossovers that occur while TSI is already on the correct side of zero are more consistent because they align with the broader momentum direction. Crossovers that happen near zero can be noisy and should usually be filtered by trend context or price structure.
TSI can also show divergence, where price makes a higher high but TSI makes a lower high, or price makes a lower low but TSI makes a higher low. Divergence is not a reversal guarantee and it is frequently early. Treat it as an alert to tighten rules, demand better confirmation, or reduce position size rather than a standalone reversal trigger.
For broader momentum context, it pairs naturally with trend tools like SMA and with other momentum comparisons like RSI. The goal is not to stack indicators, but to make each one answer a different question about trend, momentum, and timing.
When it tends to work and why
TSI tends to work best in trending or gently trending markets where momentum waves are persistent. In those regimes, the smoothing helps separate meaningful pushes from short lived noise, and the centerline becomes a useful “momentum bias” reference. Continuation entries tend to be more stable because pullbacks usually do not fully reset momentum.
It also performs reasonably well in markets with cyclical swings, where price oscillates in broad arcs rather than sharp mean reversion. In those conditions, TSI often turns before price breaks structure, which can help you plan. The indicator is doing what it was designed to do, compressing noisy momentum into smoother swings that are easier to compare.
TSI is also helpful when you want confirmation that a breakout attempt is supported by improving momentum. A breakout that occurs while TSI is rising and above zero has a different quality than a breakout where TSI is flat or falling. This does not predict outcomes, but it improves the consistency of your decision framework.
When it tends to fail and why
TSI tends to fail in tight ranges and choppy mean reverting conditions where price changes flip sign frequently. The double smoothing reduces noise, but it cannot fix a market that has no persistent momentum. In those regimes, TSI can hover around zero and generate repeated small crossovers that look meaningful but are mostly random.
It can also underperform in very sharp V shaped reversals because smoothing delays recognition. By the time TSI confirms the turn, a large part of the rebound may already have happened. That is not wrong, it simply means TSI is better suited to confirmation and continuation than to picking bottoms.
Divergence is another common trap. It is easy to treat divergence as a direct reversal signal, but markets can keep trending while momentum gradually decays. In strong trends, divergence may appear multiple times before a real trend change occurs. If you use divergence, it should change your risk management and confirmation requirements, not force an early countertrend trade.
Another failure mode comes from using the same TSI rules across very different volatility conditions without adjustment. If an instrument transitions from steady trend to high volatility chop, your crossover frequency and false signal rate will increase. A simple regime filter based on trend or price structure usually matters more than tweaking the periods.
Practical rules (entries, exits, stops, filters)
These rules are built to keep TSI in a supporting role, so the chart and trend do the heavy lifting and TSI provides confirmation and timing. They are intentionally conservative and are designed to reduce low quality signals in ranges.
- Trend filter Use an uptrend filter such as price above a rising 200 period SMA for long bias, and price below a falling 200 period SMA for short bias
- Long entry After a pullback, enter when TSI turns up and crosses above its signal line while TSI is above zero, and price reclaims a prior swing level or prints a higher low
- Short entry After a bounce, enter when TSI turns down and crosses below its signal line while TSI is below zero, and price rejects a prior swing level or prints a lower high
- Stop placement Place stops beyond the swing that defines the setup, not based on TSI, because price invalidation is clearer than indicator invalidation
- Exit logic Scale out or exit when TSI rolls over and crosses back through the signal line against your position, especially if it also approaches or crosses zero, and price breaks the most recent swing in the opposite direction
If you want a simpler approach, remove the signal line and use only the centerline plus slope. For longs, require TSI above zero and rising, and avoid new entries when TSI is falling even if still above zero. For exits, consider reducing risk when TSI makes a lower high while price makes a higher high, but only act aggressively if price also breaks structure.
To reduce whipsaws, add a no trade zone around zero where you require TSI to be clearly above or below zero before acting. Many false signals happen when momentum is essentially neutral and price is range bound. If you do nothing in that zone, your signal quality usually improves.
Summary
True Strength Index is a smoothed momentum oscillator that compares double smoothed momentum to double smoothed absolute momentum, then scales the result. It is most useful as a confirmation tool for trend continuation, using the zero line as a momentum bias reference and the slope as a persistence cue. The common 25 13 settings are a solid baseline, while the signal line period mainly controls how frequently you see crossovers.
TSI works best when markets trend or swing with persistence and tends to disappoint in tight ranges and noisy mean reversion. Use trend filters, price structure, and conservative triggers, and treat divergence as a risk management alert rather than a reversal command. If you keep TSI in a supporting role, it can improve consistency without turning your process into indicator chasing.