Skip to main content

True Strength Index: The Indicator That Actually Gets It Right (Mostly)

· 7 min read
Pineify Team
Pineify Team

You know what? I've got a love-hate thing with momentum indicators. Most of them? Complete garbage in choppy markets. But TSI—this weird little oscillator that most traders scroll right past—it's different. Not perfect, mind you. Just... different in a way that actually matters.

True Strength Index

The Guy Who Accidentally Built Something Useful​

So there's this dude, William Blau. Back in '91, probably sitting in his pajamas (I like to imagine), he drops this thing in Technical Analysis of Stocks & Commodities. Then—get this—four years later he writes a whole book about it. Momentum, Direction, and Divergence. Catchy title, right?

Here's what he did that was actually clever: instead of slapping one moving average on price changes like RSI does (yawn), he went full inception mode. Two exponential moving averages. One inside the other. Sounds simple, but—well, you know how sometimes the simplest things are the hardest to think of first.

Why "True Strength" Is Such a Pretentious Name​

Look, I didn't name it. But here's the deal: while other oscillators are basically shouting "LOOK, PRICE CHANGED!" TSI does this weird normalization thing. It divides smoothed momentum by smoothed absolute momentum. What does that even mean? Think of it like... instead of just measuring how far your car swerved, it measures how far you swerved relative to all the little wiggles you normally make.

End result? This thing bounces between -100 and +100 without having seizures every time Bitcoin sneezes.

The Math Nobody Actually Does By Hand​

Alright, here's where I lose half of you. But stick with me—it's not that scary.

Blau's original settings (25, 13, 7) work like this:

  1. Price Change = Today's close minus yesterday's close (mind-blowing, I know)
  2. First smoothing = Slap a 25-period EMA on that
  3. Second smoothing = Hit it again with a 13-period EMA
  4. Absolute Price Change = Same thing, but make everything positive
  5. Repeat steps 2-3 on the absolute stuff
  6. TSI = 100 × (the double-smoothed thing ÷ the double-smoothed absolute thing)
  7. Signal line = 7-period EMA of the TSI line

Real talk? If you're scalping, try 7/5/5. If you're more of a "check my positions once a week" person, maybe 35/21. But honestly? Test everything. Markets don't read textbooks.

The Best Pine Script Generator

Reading This Thing Without Losing Your Mind​

The Zero Line Dance​

  • Above zero = bulls are feeling frisky
  • Below zero = bears are running the show

That's... literally it. I mean, there's more nuance, but if you're looking for simple, start there.

Signal Line Crossovers (Or: Why Your Stop Loss Gets Hit Anyway)​

When TSI crosses above its signal line—supposedly bullish. Crosses below—bearish. Here's the thing though: this works great in trending markets. In chop? You're gonna get whipsawed so hard your broker will send you flowers.

The ±25 Zones (Where Dreams Go to Die)​

Blau said ±25 is where things get frothy. Reality? TSI can camp above 25 for weeks in a strong trend. Ask anyone who tried shorting TSLA in 2020 because "RSI was overbought." Yeah, how'd that work out?

Divergences: The Art of Being Right Too Early​

Higher lows on TSI + lower lows on price = potential bounce. Lower highs on TSI + higher highs on price = potential drop. Sounds easy. Except divergences can persist longer than your marriage. Always, always wait for confirmation.

TSI vs. The Usual Suspects (Spoiler: Nobody Wins)​

ThingTSIRSIMACD
Noise FilterDouble EMA (fancy)Single SMA (basic)EMA spread (confusing)
Scale-100 to +100 (centered)0-100 (why?)Goes forever
LagMediumLow-ishHigh enough to time travel

Look, they all work until they don't. TSI just tends to produce fewer false signals in sideways markets. But if you think any indicator is gonna make you money by itself, I've got some NFTs to sell you.

Settings That Actually Make Sense (By Timeframe)​

  • 1-5 min charts: 7/5/5 if you enjoy stress
  • 1-4 hour charts: 25/13/7 (Blau's original—still solid)
  • Weekly charts: 35/21/7 or just... go outside sometimes

But seriously, backtest everything. Your market isn't my market. What works on corn futures might be garbage on Dogecoin.

Five Strategies That Won't Immediately Bankrupt You​

The "Just Follow the Trend" Strategy​

Long when both price and TSI are above zero. Exit when TSI drops below zero. Revolutionary, I know. But you'd be surprised how many people try to get fancy and end up broke.

The "Oversold Bounce" Play​

Wait for TSI to drop below -20, then cross back above the signal line. Works great in strong uptrends. Works terribly in bear markets. Context matters, who knew?

Divergence + Candlestick Combo​

Spot a bullish divergence, wait for a bullish engulfing candle, then cross your fingers. Actually works better than it should, probably because most people are too lazy to wait for both.

The "Both Indicators Agree" Filter​

Only take trades when both TSI and RSI cross their centerlines in the same direction. Yeah, you'll miss some moves. You'll also miss a lot of fakeouts.

Multi-Timeframe Sanity Check​

Weekly TSI says uptrend? Only take daily long signals. Weekly says downtrend? Maybe don't fight the tide. Groundbreaking stuff, really.

Building These Things (Without Wanting to Throw Your Laptop)​

Pine Script is... an acquired taste. If you'd rather not spend your weekend debugging missing semicolons, there's this thing called Pineify. Type "give me a TSI with divergences" and boom—working code. Not sponsored, just lazy.

Advanced Stuff for People Who've Read This Far​

  1. Percentage TSI: Rescales to 0-100 because apparently some people can't handle negative numbers
  2. Histogram Mode: Shows the gap between TSI and signal line. Looks fancy, tells you the same thing
  3. Adaptive TSI: Changes periods based on volatility. Cool in theory, curve-fit in practice

How to Mess This Up (A Comprehensive Guide)​

  • Chasing extremes: TSI > 25 doesn't mean "short immediately." It means "maybe pay attention"
  • Ignoring everything else: TSI + support/resistance + volume = useful. TSI alone = gambling
  • Over-optimizing: Your perfect backtest settings will fail spectacularly next week

Questions I Get at 2 AM on Discord​

"Is TSI better than MACD?" Sometimes? TSI turns faster because it's looking at momentum of momentum. MACD is comparing EMAs. Different tools, different jobs.

"Does it work on crypto?" Sure, if your crypto has reliable closes. So... maybe stick to the top 50 coins.

"Can I just set and forget?" Buddy, if you find an indicator you can set and forget, call me. I've got a bridge to sell you too.

Look, TSI isn't gonna make you rich. No indicator will. But it's one of the better momentum tools out there—mostly because it's too boring for most YouTube traders to hype. And honestly? That's probably a good thing.

Use it as part of a system. Test everything. Don't risk money you can't afford to lose. And maybe—just maybe—spend less time looking for the perfect indicator and more time learning how markets actually work.

But what do I know? I'm just some guy on the internet.