When indicators disagree

Sooner or later it happens: one indicator shouts "sell", another "buy" and a third shrugs. Professor Wick stands between them like the referee at a tug of war. Good news: you do not have to decide in that moment. You should have decided earlier.

A real conflict: February 2024#

On 23 February 2024 the MACD line crossed below its signal line (2,176 vs 2,206) – a classic "sell" signal. Three days later, on 26 February, RSI 14 stood at 76.6, i.e. "overbought". Two oscillators, two reasons to sell.

But the price (50,744 on 23 February) was far above the 50-day moving average (45,242), and the average was pointing up. The trend said clearly: it is rising.

Whoever listened to the oscillators sold around 51,000. On 27 February MACD crossed back above its signal, RSI jumped to 81.4 – and on 14 March the price made a high of 73,777. In a strong trend, oscillators stay "overbought" for a long time.

It does not always work out like this. Another time the same conflict would end in a fall. That is exactly why the question is not which indicator is "right", but which rule you set in advance.

Who has the final word#

1. Trend as a filter. First ask the trend line which way you are trading at all. Above a rising SMA 50 you look only for buys, below a falling one only for sells – or nothing.
2. Oscillator for timing. In the direction of the trend, an oscillator can help you pick the moment – for example buying when RSI dips to 40 in a rise and turns up.
3. Volume as a check. Take a move without volume with a pinch of salt.
4. When they disagree, do nothing. No trade is a decision too – often the cheapest one.

The market regime matters#

Each indicator suits a different market regime. In a trend, averages and trend following work better; oscillators there report "overbought" for weeks. In a sideways market it is the other way round: averages keep crossing, and mean reversion tends to pay – buying near the lower edge, selling near the upper one. The same signal helps in one regime and hurts in the other.

Test your rules#

The rule "the trend filters, the oscillator times" sounds sensible – but whether it really works on your market and timeframe only shows in a test over many trades. How to turn an idea into exact rules is shown in the lesson From idea to rules, and what to watch out for when testing in Backtest pitfalls.

What you take away from the course#

• Every indicator is a formula from past prices – which is why it lags.
• Indicators belong to three families: trend, oscillators, volume. Two from the same family are not two confirmations.
• The period decides between speed and calm; a fast indicator is wrong more often.
• You know Binance's default settings and how to change them – but you change them by your plan, not by the chart.
• Divergences and conflicts are warnings, not orders. The rule you set in advance decides.

Common myths#

• "One of them is always right, I just do not know which." None is always right. The rule decides, not the indicator.
• "I will add a third indicator to break the tie." That only adds another voice to the argument.
• "An overbought market has to fall." In a strong trend it can stay overbought for weeks.