What is technical analysis?

Technical analysis is the practice of using a market's own trading history — price, volume, time — to inform decisions, rather than analysing the underlying asset's value. It is not fortune-telling, and it is not astrology. It is closer to reading a footprint than reading a horoscope.

The three assumptions

Classical technical analysis rests on three claims. It is worth stating them plainly, because two are defensible and one is doing a lot of heavy lifting.

What technical analysis is genuinely good at

• Defining risk. This is its most underrated use. A chart gives you a specific, non-arbitrary place where your idea is proven wrong — which is what a stop-loss needs.
• Timing. You can be right about a company and still lose money buying it at the wrong moment. Structure helps you wait.
• Identifying where others are trading. Round numbers, prior highs and moving averages matter partly because enough people watch them to make them matter.
• Imposing consistency. A rule-based approach removes some of the improvisation that wrecks discretionary traders.

What it cannot do

• Predict news. No chart pattern anticipates a surprise regulatory decision or a profit warning.
• Tell you what something is worth. That is fundamental analysis' job, and it is why the two are complements rather than rivals.
• Work identically everywhere. A method calibrated on liquid index futures can fall apart on a thin small-cap.
• Survive without risk management. A 60%-accurate method still produces losing streaks that will end an oversized account.

The criticism you should take seriously

The strongest objection is not “markets are efficient.” It is overfitting: with enough indicators, parameters and hindsight, you can always construct rules that would have worked beautifully on data you have already seen. Those rules then fail on data they have not seen.

The second-strongest is confirmation bias. Charts are visually rich and ambiguous — precisely the conditions under which humans reliably see the pattern they were hoping to find. Every trader has drawn a trendline to fit a conclusion they had already reached.

If your analysis never tells you to stay out, it is not analysis. It is justification.

Self-fulfilling, and why that is fine

A common jab is that technical analysis “only works because people believe in it.” That is partly true and not the criticism it sounds like.

If a hundred thousand traders watch the 200-day moving average, orders genuinely cluster there. The level becomes real in its effects regardless of whether it has any deeper meaning. Money is a shared fiction too; that does not stop it buying things. What matters is whether the clustering is reliable enough to trade around after costs.

How to practise it without fooling yourself

1. 1
   Fewer indicators, not more

   Three indicators measuring the same thing is one indicator with extra steps. Most profitable approaches are visually sparse.

2. 2
   Write the rule before the chart

   Define what you are looking for in advance. Scanning charts for something interesting is how you find noise.

3. 3
   Mark levels on a higher timeframe

   Then drop down to execute. Levels drawn on the chart you trade on are contaminated by what you already want to happen.

4. 4
   Track everything

   Your memory of your own trading is systematically flattering. A written record is not.

5. 5
   Test on data you have not looked at

   If a method only works on the period you designed it on, you have described the past, not discovered an edge.

The rest of this track covers the tools themselves, starting with the only one you genuinely cannot trade without: levels.