The Best Technical Indicators: Tested Over 100 Years of Data

Is RSI better than MACD? Are Bollinger Bands more reliable than Donchian Channels? And when it comes to moving averages, does the exponential version really outperform the simple one?
We had the same questions. So, we put the most popular technical indicators to the test, analyzing nearly a century of market data.
The results? We ranked them into three categories: the most reliable technical indicators, known for their high win rates and accuracy; the most consistent indicators, which delivered steady gains over time; and the highest-performing indicators, maximizing overall returns.
Curious to see where your favorite trading indicators stand? Let’s dive in.
Key Takeaways
The results here are based on 100 years of Dow Jones data, using simplified assumptions, offering a strong starting point for exploration. Backtesting your own strategies with specific parameters and context is always a good idea.
RSI and Bollinger Bands proved to be the most reliable indicators, consistently delivering high win rates across both testing periods.
Donchian Channels and Williams %R (Williams Percent Range) stood out for their consistency, maintaining steady performance over time in real-world and controlled conditions.
Ichimoku and EMA (Exponential Moving Average) emerged as the highest-performing indicators, delivering the best overall returns by balancing accuracy and profitability.
No single indicator works perfectly in every situation—choose the one that aligns with your trading style, whether you value reliability, consistency, or maximizing returns.
Our Methodology
Finding the best technical indicators isn’t easy—but that’s exactly why we took on the challenge. Using ProRealTime’s backtesting tool, we aimed to discover which indicators truly stand out.
To ensure we worked with data that’s rich and meaningful, we focused on the Dow Jones Industrial Average (DJIA)—a stock market index with over a century of history.
To make the test as reliable as possible, we analyzed nearly 100 years of Dow Jones Industrial Average data. We split the timeline into two phases:
- In-Sample Period (October 1, 1928 – December 31, 1995):
We used this period to develop and refine the strategies. - Out-of-Sample Period (January 2, 1996 – December 31, 2024):
This was the real-world test. We applied the strategies to this new data to see if they held up.
Using the same data to build and test strategies can lead to overfitting—where a strategy is too tailored to the past and fails in the future.
By splitting the timeline, we tackled this issue head-on. The In-Sample Period allowed us to test ideas and optimize performance using historical data, while the Out-of-Sample Period served as the real-world test, showing how the strategies performed on unseen data.
For simplicity, every strategy we tested was long-only, meaning we only took positions when the signals pointed to buying opportunities. And, to keep things straightforward, each trade represented a single unit of the index.
The amount of fake money used was $10,000, and the trades were done without any leverage.
The goal? To find the technical indicator that performs best over nearly a century of market data. It’s been an exciting journey digging into the data and uncovering actionable insights from the results.
How Trading Signals Were Assigned to Indicators
We kept the trading signals straightforward, reflecting how these indicators are commonly used in real-world scenarios:
Technical indicator | Buy signal | Exit Buy signal |
---|---|---|
Simple Moving Average (50) | Bullish MA | Bearish MA |
Exponential Moving Average (50) | Bullish EMA | Bullish EMA |
RSI (14) | RSI less than or equal to 30 | RSI greater than or equal to 70 |
MACD (12,26,9) | MACD bullish crossover and its signal | MACD bearish crossover and its signal |
Stochastic (14,3,5) | Bullish stochastic crossover and its signal | Bearish stochastic crossover and its signal |
Bollinger Bands (20,2) | Price lower than the lower limit | Price higher than the upper limit |
Donchian Channels (20) | Price higher than the upper limit | Price lower than the lower limit |
Parabolic SAR (0.02, 0.02, 0.2) | SAR green line | SAR red line |
Ichimoku (9,26,52) | Price higher than Senkou Span B | Price lower than Senkou Span A |
ADX (14) | ADX ADXR crossover indicating an upside trend | ADX ADXR crossover indicating a downside trend |
Williams Percent Range (%R) (14) | Buy at an oversold level | Sell at an overbought level |
Commodity Channel Index (CCI) (20) | Buy at an oversold level | Sell at an overbought level |
Momentum (12) | Moves into positive territory | Moves to negative territory |
TRIX (15,9) | Moves into positive territory | Moves to negative territory |
Limitations to the Methodology
Every methodology has its quirks, and ours is no exception. While our approach to analyzing the Dow over nearly a century is robust, there are a few limitations worth keeping in mind to fully appreciate the results.
By focusing solely on the Dow, we inevitably faced survivorship bias since the index evolves over time with only successful companies sticking around.
Additionally, limiting ourselves to long-only strategies means we missed out on opportunities to explore the bearish side of the market, while fixed position sizing left no room to adapt for volatility or price swings.
Transaction costs, slippage and bid-ask spreads were left out of the equation, and overfitting is always a concern. Plus, evaluating just one index doesn’t capture the full breadth of market behavior or how strategies and indicators might perform across different sectors or asset classes.
The Results

Past performance doesn’t guarantee future success—markets evolve over time. We tested one type of buy-and-exit system, but many variations could yield different insights. While the methodology is solid, it’s not flawless. These aren’t dealbreakers, but they’re important to consider as you interpret the findings.
In-Sample Results: October 1, 1928, to December 31, 1995
Strategy | Win Rate | Average Gain | Average Loss | Gain/Loss Ratio | Return Rate |
---|---|---|---|---|---|
SMA (50) | 30.85% | 4.03 | 0.96 | 4.2 | 1.6 |
EMA (50) | 31.27% | 3.76 | 0.74 | 5.08 | 1.9 |
RSI (14) | 71.64% | 5.12 | 5.89 | 0.87 | 1.34 |
MACD | 37.56% | 7.46 | 2.6 | 2.87 | 1.45 |
Stochastics | 46.28% | 1.62 | 1 | 1.62 | 1.21 |
Bollinger Bands | 74.32% | 3.83 | 4.95 | 0.77 | 1.32 |
Donchian Channels | 70.00% | 3.41 | 4.58 | 0.74 | 1.22 |
Ichimoku | 46.00% | 6.87 | 1.68 | 4.09 | 2.34 |
Parabolic SAR | 45.10% | 2.39 | 1.69 | 1.41 | 1.09 |
ADX | 53.32% | 2.82 | 1.95 | 1.45 | 1.3 |
WPR | 69.66% | 2.34 | 2.55 | 0.92 | 1.34 |
CCI (20) | 34.60% | 4.66 | 1.2 | 3.88 | 1.69 |
Momentum | 42.55% | 2.57 | 1.18 | 2.18 | 1.35 |
Trix (15, 9) | 42.93% | 8.11 | 3 | 2.7 | 1.59 |
Out-of-Sample Results: January 2 1996 to December 31 2024
Strategy | Win Rate | Average Gain | Average Loss | Gain/Loss Ratio | Return Rate |
---|---|---|---|---|---|
SMA (50) | 26.25% | 71.25 | 17.22 | 4.14 | 1.35 |
EMA (50) | 30.07% | 61.75 | 18.48 | 3.34 | 1.31 |
RSI (14) | 87.11% | 116.5 | 150.74 | 0.77 | 1.54 |
MACD | 42.70% | 103.71 | 51.69 | 2.01 | 1.28 |
Stochastics | 43.42% | 31.64 | 19.47 | 1.63 | 1.14 |
Bollinger Bands (BB) | 81.25% | 68.48 | 95.21 | 0.72 | 1.4 |
Donchian Channels | 78.26% | 53.73 | 75.58 | 0.71 | 1.34 |
Ichimoku | 38.68% | 81.92 | 38.74 | 2.11 | 1.2 |
Parabolic SAR | 44.25% | 50.6 | 31.55 | 1.6 | 1.15 |
ADX | 53.81% | 46.68 | 38.58 | 1.21 | 1.19 |
WPR | 73.66% | 38.87 | 55.68 | 0.7 | 1.25 |
CCI (20) | 35.69% | 72.13 | 29.12 | 2.48 | 1.24 |
Momentum | 40.65% | 39.49 | 18.72 | 2.11 | 1.26 |
Trix (15, 9) | 44.32% | 104.85 | 55.62 | 1.89 | 1.28 |
Which Technical Indicators are the Most Reliable?
As it relates to the most reliable technical indicators, it’s all about their ability to maintain a high win rate both in the in-sample and out-of-sample periods.
These are the indicators that you can rely on to call the right shots, proving their worth not just in controlled testing but also in the unpredictable twists and turns of real market conditions. High reliability like this is what separates the truly dependable tools from the rest.
Technical Indicator | Win Rate |
---|---|
RSI(14) | 79.4% |
BB | 77.8% |
Donchian Channels | 74.1% |
WPR | 71.7% |
ADX | 53.6% |
Stochastics | 44.9% |
Parabolic SAR | 44.7% |
Trix(15,9) | 43.6% |
Ichimoku | 42.3% |
Momentum | 41.6% |
MACD | 40.1% |
CCI(20) | 35.1% |
EMA(50) | 30.7% |
SMA(50) | 28.6% |
Which Technical Indicators Are the Most Consistent?
When we talk about “most consistent” technical indicators, we’re looking for the ones that deliver a solid average gain across the board – both in the in-sample and out of sample periods. These are the indicators that prove themselves over time, not just in ideal conditions but also when tested against the unpredictability of real-world markets. High average gains in both periods are the true hallmark of a consistent performer.
Technical Indicator | Win Rate |
---|---|
RSI(14) | 79.4% |
BB | 77.8% |
Donchian Channels | 74.1% |
WPR | 71.7% |
ADX | 53.6% |
Stochastics | 44.9% |
Parabolic SAR | 44.7% |
Trix(15,9) | 43.6% |
Ichimoku | 42.3% |
Momentum | 41.6% |
MACD | 40.1% |
CCI(20) | 35.1% |
EMA(50) | 30.7% |
SMA(50) | 28.6% |
Which are the Best Technical Indicators?
For us, the best technical indicators are those that deliver the highest return rate, which we calculate using a straightforward formula:
Return Rate = Win Rate x(Average Gain/Average Loss + 1)
This approach balances how often the indicator wins trades with the size of the average profits versus losses, giving us a clear picture of overall effectiveness.
It’s not just about winning often-it’s about making sure the wins outweigh the losses in a meaningful way.
Technical Indicator | Return Rate |
---|---|
Ichimoku | 1.77 |
EMA(50) | 1.60 |
SMA(50) | 1.48 |
CCI(20) | 1.47 |
RSI(14) | 1.44 |
Trix(15,9) | 1.43 |
MACD | 1.37 |
BB | 1.36 |
Momentum | 1.31 |
WPR | 1.29 |
Donchian Channels | 1.28 |
ADX | 1.25 |
Stochastics | 1.18 |
Parabolic SAR | 1.12 |
Final thoughts
Technical indicators are powerful tools, but their effectiveness depends on how, when, and where you use them. The rankings in this article provide a strong starting point, but they’re not a one-size-fits-all solution.
Markets evolve, and so should your approach. Backtesting your strategies on different assets, time frames, and conditions is crucial to finding what works best for you. Whether you prioritize reliability, consistency, or high returns, success lies in understanding the strengths and limitations of each tool—and continuously refining your approach.
Want to see how these indicators perform with your trading style? Platforms like ProRealTime let you test and fine-tune strategies with just a few clicks. Take the insights from this study and make them work for you.
Stay curious, stay adaptable, and remember: the best traders aren’t those who find perfect tools but those who use them wisely.

Maxime holds two master’s degrees from the SKEMA Business School and FFBC: a Master of Management and a Master of International Financial Analysis. As founder and editor-in-chief of NewTrading.fr, he writes daily about financial trading.

