True Strength Index (TSI) Trading Strategy

Key Takeaways

• TSI uses double-smoothed EMA of price change vs absolute change
• Classic settings are 25‑period and 13‑period EMAs
• Gold backtest showed 7.8% CAGR with 1.7 profit factor
• Win ratio 40% but average winner 4.5% vs loser 1.6%
• TSI complements MACD and RSI for momentum confirmation

Pulse Analysis

The True Strength Index (TSI) emerged as a refinement of classic momentum oscillators, blending price‑change velocity with exponential smoothing to reduce market noise. By applying a 25‑period EMA to raw price differences and then a 13‑period EMA to that result, the TSI creates a smoother signal that is divided by a double‑smoothed absolute price change, yielding a percentage‑based oscillator that pivots around zero. Unlike the RSI, which focuses on overbought/oversold extremes, the TSI emphasizes the persistence of momentum, making it a useful complement to tools such as MACD that also rely on moving‑average crossovers.

Practitioners interpret the TSI through several lenses. A crossing above zero signals a shift to bullish momentum, while a dip below indicates bearish pressure. Adding a signal line—typically a short EMA of the TSI—produces crossover alerts akin to MACD, offering entry and exit cues. Divergence between price and TSI adds a layer of confirmation, though traders must codify divergence criteria to avoid subjectivity. The default 25/13 EMA settings work well across many markets, yet seasoned users adjust periods based on asset volatility, timeframe, and whether they pursue trend‑following or mean‑reverting tactics.

A concrete illustration comes from a backtest on the SPDR Gold Trust (GLD), where a straightforward rule—buy when TSI crosses above zero and sell on the opposite crossing—generated a 7.8% compound annual growth rate over the sample period. The strategy posted a profit factor of 1.7 and a risk‑adjusted return of 15% after accounting for the 50% market exposure, despite a modest 40% win ratio. The edge stemmed from a high average winner (4.5%) versus a small average loss (1.6%). These results underscore that TSI can add value to systematic portfolios, but practitioners should still conduct out‑of‑sample testing, factor in transaction costs, and combine TSI with complementary filters to mitigate false signals.