RSI Buy and Sell Signals: How to Identify Them (With Backtest Proof)
.png)
RSI is the most widely used momentum indicator in trading. It is on every charting platform, taught in every course, and drawn on every crypto trader's screen. And most of them are using it wrong, or at least incompletely.
This article covers RSI buy and sell signals and the classic 30/70 setup, then examines a BTCUSDT backtest from 2019 to 2024.
What Is the RSI?
The Relative Strength Index (RSI) is a momentum oscillator developed by J. Welles Wilder Jr. in 1978. It measures the speed and magnitude of recent price changes on a scale from 0 to 100.
The formula compares average gains to average losses over a lookback period. The default is 14 candles. When price rises sharply, RSI climbs toward 100. When price sells off hard, RSI drops toward 0.
Classic RSI Buy and Sell Signals
Wilder defined two zones that traders use as entry and exit triggers. The logic is straightforward: when price has fallen fast enough to push RSI into oversold territory, a bounce is more likely. When RSI hits overbought, the rally may be running out of steam.
RSI drops below 30 → Oversold zone → Potential buy signal
RSI rises above 70 → Overbought zone → Potential sell signal
Many traders refine this by waiting for RSI to cross back through the threshold, buying when RSI crosses back above 30 (confirming recovery) rather than the moment it dips below. This filters out some false signals.
Do RSI Signals Actually Work on Crypto?
Crypto is not stocks. Higher volatility, 24/7 markets, and thinner liquidity mean RSI behaves differently here. During a bear market, RSI can stay oversold for weeks. During a bull run, it can stay overbought for months.
Before building a strategy around RSI signals, you need data. Here's what a systematic backtest actually shows.
Backtest: RSI 30/70 on BTCUSDT (2019–2024)
Strategy setup:

| PARAMETER | VALUE |
|---|---|
| Asset | BTCUSDT, Binance spot |
| Timeframe | Daily (1d bars), Close inputs |
| RSI Period | 14, Wilder RelativeMovingAverage (RMA) |
| Buy Signal | Previous RSI >= 30 AND current RSI < 30 |
| Sell Signal | Previous RSI <= 70 AND current RSI > 70 |
| Test Period | January 1, 2019 – December 31, 2024 (UTC) |
| Data Source | Binance spot |
| Starting Capital | 10,000 USDT |
| Position Sizing | 100% of current equity; one long position |
| Execution Assumption | Signal-candle Close, market orders |
| Trading Fees | 0.1% on each entry and exit |
| Slippage | 5 bps (0.05%) adverse per fill, configured |
Results:

| METRIC | RSI 30/70 BACKTEST | BUY & HOLD REFERENCE (BTC) |
|---|---|---|
| Total Return | +98.94% | +2,364.38% (curve-derived) |
| Closed Trades | 8 | N/A (reference curve) |
| Win Rate | 75.00% | N/A (reference curve) |
| Max Drawdown (magnitude) | 50.37% | 76.63% (curve-derived) |
| Sharpe Ratio | 0.4871 | Not reported |
| Profit Factor | 2.0369 | N/A |
| Payoff Ratio | 0.6790 | N/A |
| CAGR | 12.14% | Not reported |
| Ending Value | 19,893.87 USDT | 246,438.11 USDT |
| SQS | 46 | N/A |
| Total Fees | 251.79 USDT | Not established |

What the Backtest Tells Us
The RSI 30/70 backtest returned +98.94% from 2019 to 2024, taking 10,000 USDT to 19,893.87 USDT. The Bitcoin Buy & Hold reference ended at 246,438.11 USDT, a curve-derived return of +2,364.38%. The strategy therefore finished far below that reference endpoint.
Buy & Hold is the reference curve: 10,000 USDT on January 1, 2019 and 246,438.11 USDT on December 31, 2024. Its +2,364.38% return and 76.63% maximum drawdown are calculated from that curve, not taken from an opaque outperformance score. The reference curve's fee/slippage treatment is not established, so this is a reference comparison, not a proven like-for-like cost-adjusted benchmark. Benchmark Sharpe and CAGR are not reported here; trade count and win rate are not inferred for a reference curve.
The 75.00% win rate represents six winning closed trades and two losing closed trades. The average win is 3,239.29 USDT and average loss magnitude is 4,770.93 USDT, giving a payoff ratio of 0.6790. Winning trades were more frequent but smaller on average. The run has only eight closed trades across the six-year window; the Time in Market metric is 34.22%. A small trade sample does not establish a durable trading edge.
The profit factor is 2.0369, but that does not by itself prove a repeatable edge. Sharpe is 0.4871 and maximum drawdown is 50.37%; the worst closed trade lost 7,966.17 USDT. SQS is 46. All eight exits are recorded as Signal exits, and the run ends flat.
The backtest uses daily Close inputs, market orders, 10,000 USDT starting capital, one long position, and 100% of current equity. Costs are configured as 0.1% taker fees on each entry and exit and 5 bps (0.05%) adverse slippage per fill. Entries and exits fill at the signal candle's Close, not the next day's Open.
How to Improve RSI Signals
Five variations to investigate in separate tests. None was tested in this backtest, so the following suggestions are not demonstrated improvements to these results:
Trend filter: Only take RSI buy signals when price is above the 200-day MA. Avoids bottom-catching in bear markets.
Wider thresholds: Try 20/80 instead of 30/70 on crypto. Reduces false signals during strong trends.
Multi-timeframe confirmation: Confirm a daily RSI signal with 4H RSI alignment before entering
Volume confirmation: RSI signal + above-average volume = higher conviction entry; RSI signal + low volume = skip
RSI divergence: Price makes a new low but RSI does not. This is a stronger reversal signal than threshold crosses alone.
How to Test RSI Signals on CoinQuant
You don't need to code any of this. CoinQuant's no-code Strategy Builder lets you run custom RSI backtests in under a minute:
Sign in at CoinQuant
Open the Strategy Builder
Select your asset (e.g., BTCUSDT) and timeframe
Add the RSI indicator with your chosen period (default: 14)
Set your entry rule: RSI crosses below 30 (or your custom threshold)
Set your exit rule: RSI crosses above 70 (or your custom threshold)
Click Run Backtest. Get the full performance report instantly.
For the backtest discussed here, set January 1, 2019 through December 31, 2024 and use the Wilder RMA, Close-input, current-equity sizing and cost settings above. The exact crossing inequalities are in the setup table. This is not a promise of an identical future run.
The real value comes from testing variations and adding filters before risking capital. A trend filter, adjusted thresholds, or multi-timeframe confirmation can meaningfully change the outcome. Run the numbers first. That is exactly what CoinQuant is built for.
Conclusion
This RSI backtest produced a positive return, but +98.94% did not match the Buy & Hold reference's +2,364.38%. A 75.00% win rate across eight closed trades, profit factor of 2.0369, Sharpe of 0.4871 and 50.37% drawdown do not establish that the signals are reliable or that the outcome could not occur by chance. The test supports a narrower conclusion: evaluate RSI rules against a relevant reference and inspect losses, costs and sample size before using them. Trend filters, wider thresholds and multi-timeframe confirmation remain untested variations here.
That is the difference between traders who follow indicators and traders who know whether those indicators actually work.
Ready to test your RSI setup? Backtest RSI signals on CoinQuant. Free. Get your backtest results in under 60 seconds.
Disclaimer:
This content is for educational and informational purposes only and does not constitute financial, investment, or trading advice. All strategies and examples are for illustrative purposes and do not guarantee results. Always conduct your own research before making financial decisions.
Key Takeaway