Bitcoin RSI and MACD explained: how to read RSI, MACD and EMA

Bitcoin RSI (Relative Strength Index) is a 0⁠–⁠100 momentum gauge that compares the size of BTC's recent gains with its recent losses, usually over 14 candles. Readings above 70 are traditionally called overbought, below 30 oversold, and around 50 neutral. MACD and EMA fill in the rest of the picture: the EMA (exponential moving average) shows the direction of the trend, while MACD shows whether that trend is speeding up or fading. As of October 20, 2026, our Binance-based data put bitcoin's 14-day RSI at 46.6, in neutral territory, with MACD below its signal line. Below is how each indicator is calculated, how traders read it, and what an eight-year backtest says about whether the classic signals actually work on bitcoin.

What RSI is and how it is calculated

J. Welles Wilder introduced RSI in his 1978 book New Concepts in Technical Trading Systems (StockCharts ChartSchool). The math fits in five steps:

1. For each candle, take the change in closing price. Rises count as gains, falls (as positive numbers) count as losses.
2. The first average gain and average loss are simple 14-period averages.
3. After that, apply Wilder smoothing: (previous average × 13 + current value) / 14.
4. RS = average gain / average loss.
5. RSI = 100 − 100 / (1 + RS).

Fourteen periods is the default both in Wilder's original and on TradingView. On a daily chart that is two weeks; on a 4-hour chart, a little over two days. Shorter settings push RSI into the extreme zones more often and produce more false signals.

How to read RSI: 70/30, divergences and trend ranges

There are four classic ways to read the indicator:

The trend ranges come from analyst Constance Brown, and her mentor Andrew Cardwell observed that bearish divergences are more typical of bull markets and usually lead to a brief correction rather than a true reversal (ChartSchool). That matters for bitcoin, where RSI can stay above 70 for weeks during a strong rally.

MACD: two averages and a signal line

Gerald Appel developed MACD (Moving Average Convergence Divergence) in the late 1970s (ChartSchool). The standard settings are 12, 26 and 9:

• MACD line = 12-period EMA − 26-period EMA;
• signal line = 9-period EMA of the MACD line;
• histogram = MACD line − signal line.

A MACD line crossing above the signal line is read as bullish momentum, a cross below as bearish. A move through zero means the 12-period EMA has moved above or below the 26-period EMA. MACD has no upper or lower bound, so it is a poor tool for spotting overbought conditions, and because it is measured in dollars, values cannot be compared across assets: bitcoin's MACD runs in the hundreds of dollars while ether's runs in single digits.

EMA: trend, levels and the golden cross

An exponential moving average gives each new close a weight of 2 / (n + 1), which is about 9.5% for a 20-period EMA, so it reacts faster than a simple moving average of the same length (ChartSchool). Common periods:

When the 50-day average crosses above the 200-day, traders call it a golden cross; a cross below is a death cross. These are slow signals: on Binance daily candles from March 2018 to October 2026, the 50- and 200-day EMAs crossed only 14 times (our calculation from the Binance API). For longer-term cycle gauges, see our guide to MVRV and the Mayer Multiple.

What bitcoin's indicators look like on October 20, 2026

The technical analysis section of our daily brief calculates these indicators from Binance BTC/USDT daily candles. As of October 20, 2026, 4:00 AM ET:

The picture is mixed. Our dashboard recorded the latest 50/200 EMA golden cross on September 14, 2026, and price remains above both long averages. But by our calculation MACD crossed below its signal line on September 29, and the MACD line itself has drifted to roughly zero, meaning the 12- and 26-period EMAs have nearly converged. Over the past 120 days RSI ranged from 30.3 (June 30) to 85.9 (August 21): heavily overbought in August, momentum spent now. In short, the indicators describe sideways action inside a broader uptrend, not a signal either way. Live values are on our Bitcoin page and in the market pulse.

Do RSI and MACD work on bitcoin? An eight-year backtest

We calculated 14-day RSI and MACD(12, 26, 9) on every closed Binance BTC/USDT daily candle from August 17, 2017 to October 19, 2026, then measured the price change over the 30 days after each signal (signals counted from March 5, 2018, once the averages had enough history).

The results run against the textbook. "Oversold" readings below 30 did not, on average, lead to a bounce: a month later price was slightly lower, worse than a random day. "Overbought" readings above 70 more often meant the rally kept going, because bitcoin's strong moves tend to extend. Bullish MACD crosses did a bit better than the baseline and bearish ones a bit worse, but a couple of percentage points is not much given the spread of outcomes. There were 229 MACD crosses in 8.6 years, about 26 a year, and most of them were noise.

Academic work points the same way. Wei, Kyriakou, Sermpinis and Stasinakis, writing in the International Journal of Finance and Economics (2024), tested 7,846 technical rules on 12 cryptocurrencies; only a handful showed genuine out-of-sample profitability, and the authors conclude that their findings question the value of technical analysis on crypto (University of Glasgow eprints). Our look at bitcoin fractals reached a similarly cautious verdict.

Common mistakes

• Buying just because RSI is under 30. On bitcoin this has not worked on average. In 2026, RSI fell below 30 three times: January 31, February 12 and June 2. Thirty days later price was 12.6% and 7.8% lower after the first and third signals, and 7.5% higher only after the second.
• Selling just because RSI is over 70. In a strong trend RSI can stay overbought for weeks; it was above 70 on 10.4% of all days since 2018.
• Comparing numbers across websites. EMA and RSI depend on how much history the calculation starts from and on the exchange feed. Our 200-day EMA seeded with 400 candles is $76,236; seeded with the full history since 2017 it comes out near $75,970.
• Mixing timeframes. RSI 70 on a 15-minute chart and on a daily chart are different events.
• Curve-fitting the settings. Try hundreds of period combinations and one will always "work" on past data, then fail going forward.
• Ignoring derivatives. Price-based indicators can't see open interest, funding or liquidations, which often drive the move.

FAQ

What is a normal RSI for bitcoin? Between 30 and 70 is neutral, and around 50 means no clear momentum. In bull markets, bitcoin's RSI often holds in a 40⁠–⁠90 range.

What are the best MACD settings for crypto? Most platforms use the standard 12, 26, 9, so those are the signals everyone sees. Shorter periods give more signals and more noise.

EMA or SMA? An EMA reacts faster to recent prices; an SMA is smoother. For long levels like the 200-day, the difference is small.

Can you trade on these indicators alone? Our backtest and academic research both suggest that, on their own, they offer no reliable edge. They are best used to describe the current state of the market, not as ready-made trade signals.

This is not investment advice.