Updated 13 October 2026
The best technical indicators for algorithmic trading are the ones that can be written as exact rules and tested on data. Moving averages and MACD describe trend, RSI measures momentum, and Bollinger Bands and ATR measure volatility. No indicator works on its own. What decides the result is how it is combined with entry, exit and position size.
Key facts
- Bollinger Bands use a 20-period moving average and bands two standard deviations away as standard settings (John Bollinger, October 2026).
- Under ESMA's rules for retail clients, CFDs on major indices and gold have a maximum leverage of 20:1 and major currency pairs 30:1 (ESMA, 2018).
- Our own ProRealTime account went from 35,000 kr to 218,747 kr after fees between December 2023 and 30 September 2026, with a largest decline of −23.5 %.
- Our two momentum breakout algos had opposite results in 2026 to the end of September, Nasdaq M1 +39,050 kr and DJ30 M1 −13,541 kr (live data, 1 October 2026).
What is a technical indicator in algorithmic trading?
A technical indicator is a formula that turns price, and sometimes volume, into a number a computer can compare against a rule. A moving average, for example, is the average closing price of the last 50 bars. The algorithm can then check on every new bar whether price is above or below it.
In algorithmic trading an indicator is a condition in code, and the code needs exact numbers for the period, the threshold and what happens when the condition is met. So the useful question is which indicator can be defined precisely enough to test. If you want to see how a condition becomes an order, read how a trading bot works.
Which technical indicators are most used in algorithmic trading?
Five indicator families cover most automated strategies. They are moving averages, RSI, MACD, Bollinger Bands and ATR, plus simple price channels for breakouts.
| Indicator | Measures | Standard setting | Typical rule in code |
|---|---|---|---|
| Moving average (SMA, EMA) | Trend | For example 50 and 200 bars | Go long when the fast average crosses above the slow one |
| RSI | Momentum | 14 bars, levels 70 and 30 | Filter out longs when RSI is above 70 |
| MACD | Trend and momentum | 12, 26 and 9 bars | Enter when the MACD line crosses the signal line |
| Bollinger Bands | Volatility | 20 bars, 2 standard deviations | Act on a close outside a band or on narrowing bands |
| ATR | Volatility | 14 bars | Set stop distance as a multiple of ATR |
| Price channel | Breakouts | Highest high and lowest low of N bars | Enter when price breaks the channel |
Moving averages smooth price so the direction becomes clear. A simple moving average weights every bar equally, while an exponential one gives more weight to recent bars and reacts faster. A faster average gives earlier signals and more false ones.
RSI, the Relative Strength Index, compares recent gains with recent losses on a scale from 0 to 100. Readings above 70 are often called overbought and below 30 oversold. In a strong trend RSI can stay above 70 for a long time, so a rule that sells every time it gets there will trade against the trend.
MACD is the difference between a 12-bar and a 26-bar exponential average, with a 9-bar average of that difference as the signal line. It is built from moving averages, so it lags price in the same way.
John Bollinger describes his bands as a 20-period average with bands two standard deviations above and below. ATR, the Average True Range, measures how far price typically moves per bar, which makes it useful for stop distance and position size rather than for direction.
Which indicators fit trend following and which fit breakouts?
Trend following strategies mostly use moving averages and MACD, while breakout strategies mostly use price channels, Bollinger Band width and ATR. RSI and band touches are more common in mean reversion, which bets that price returns towards an average.
A trend follower accepts many small losses in sideways markets in exchange for a few large gains when a trend runs. A momentum breakout strategy waits for price to leave a range with force and enters in that direction.
Most of our own algos are trend followers, and two of them, Nasdaq M1 and DJ30 M1, are momentum breakouts. We do not publish the exact rules or indicators in the code, so this article describes how the indicator families work in general, not what is inside a specific algo.
How many indicators should one algorithm use?
Usually few. Every indicator adds parameters, and every parameter is another way to fit the rules to past data instead of to how markets behave.
Two indicators that measure the same thing, such as MACD and a moving average crossover, add little information. A rule set that only works with one exact period, say a 17-bar average but not 15 or 20, is a warning sign of overfitting, meaning the rules describe the history rather than the market.
How do you test an indicator before trading live?
You run it in a backtest, check it on data it was not built on, and then run it live with small size. A backtest is a simulation on historical data. It shows how the rules would have behaved, not how they will behave.
Include spread, overnight financing and slippage, since a rule that trades often can look strong before costs and weak after. Then decide position size and stop loss before you start, which our article on risk management in algo trading covers in more detail.
Leverage matters as much as the indicator. Under ESMA's product intervention, retail CFDs on major indices and gold are limited to 20:1 leverage, and national regulators found that 74 % to 89 % of retail CFD accounts typically lose money.
What does our live account show about indicator-based rules?
It shows that rules of the same type can give very different results on different markets in the same period. The figures below come from our live data per 30 September 2026.
| Our ProRealTime account | Value |
|---|---|
| Start, December 2023 | 35,000 kr |
| End of September 2026, after fees | 218,747 kr (+525 %) |
| Largest decline from peak | −23.5 % (February to April 2024) |
| Positive months | 20 of 34 |
| 2026 to end of September | −0.2 % (−433 kr), 4 of 9 months positive |
| Nasdaq M1, momentum breakout, 2026 | +39,050 kr |
| DJ30 M1, momentum breakout, 2026 | −13,541 kr |
| Average orders per day | 7.7 |
The two breakout algos trade a similar idea on two US indices, and in 2026 one gained 39,050 kr while the other lost 13,541 kr. That spread is a reminder that an indicator setup is never a finished edge. It depends on the market, the period and the risk per trade, which is under 2 % in our account according to the live results page.
Past results do not predict future results, and losing periods like September 2026, −11.8 %, are a normal part of trading. Every algo and every year is listed on A+ Algos in numbers, and the systems are in our ProRealTime algos collection. You can try one of our algos free for 14 days before you buy a licence.
FAQ
What is the single best indicator for algo trading?
There is no single best indicator. Moving averages are the most common because they are simple to code and test, but every indicator only describes past price. What matters more is whether the full rule set, with entry, exit, stop and position size, holds up on data it was not built on and then in live trading after costs.
Is RSI good for automated trading?
RSI is easy to code and widely used, often as a filter rather than as the main signal. Its weakness is that in a strong trend it can stay above 70 or below 30 for a long time. A rule that always sells overbought readings will then trade against the trend. Test RSI rules across both trending and sideways periods before trusting them.
Can technical indicators predict the market?
No. Indicators are calculated from past prices, so they describe what has happened, not what will happen. An algorithm uses them to apply the same rule every time, which removes ad hoc decisions but not risk. A backtest of an indicator strategy is a simulation, and live results with real costs can be considerably weaker.
Which indicators do A+ Algos use?
We do not publish the exact rules or indicators inside our algos. What we can say is that most of them are trend followers and two, Nasdaq M1 and DJ30 M1, are momentum breakouts. They trade CFDs on indices, gold and EUR/USD, long and short, and we publish the live result of every algo every month after fees.
Summary
Moving averages, RSI, MACD, Bollinger Bands and ATR are the most used technical indicators in algorithmic trading because they can be written as exact rules and tested. None of them works alone, and fewer, well-tested rules usually beat many combined indicators. Our own live account since December 2023 shows that results vary a lot between markets and periods.
See the results live. Every month since December 2023 we publish the result of our own algos after fees on the live results page. Our most traded algo is Nasdaq M3 for ProRealTime.
Trading CFDs involves high risk and you can lose money. Past results, live or backtested, do not predict future returns.