Every indicator on your platform is arithmetic performed on prices you can already see. That is not a criticism — it is the correct expectation. An indicator summarises; it does not know anything the chart does not. The skill is choosing summaries that answer a specific question, and refusing to let them answer questions they cannot.
Four workhorses
- Moving average — the trend, smoothed. Price above a rising long-period MA is a fact about trend; MA crossovers in a range are noise machines.
- RSI — momentum on a 0–100 scale. Useful for divergence and extremes, but in a strong trend it can sit "overbought" for weeks. Overbought is a description, never a sell signal by itself.
- MACD — the relationship between two moving averages, good at showing momentum shifting before price turns.
- ATR — average true range: how far the market actually moves per period. The least glamorous and the most useful — it converts volatility into a number you can size stops with.
How professionals actually use them
As filters, not triggers. A desk might require the higher-timeframe MA to slope in the trade’s direction, use RSI divergence as a warning to tighten stops, and set stop distance as a multiple of ATR so the position breathes with the market. The decision still comes from structure and levels; indicators veto or confirm, they do not originate.
The common misuses
Stacking five indicators that all derive from the same closing prices adds correlation, not information. Trading every MA crossover in a sideways market is paying the spread to be whipsawed. And optimising indicator settings until the backtest looks perfect mostly discovers noise — a topic we return to in the automation lesson. If you cannot state, in one sentence, what question an indicator answers on your chart, remove it. A clean chart with two tools you understand beats a dashboard you obey.