What technical analysis cannot do

This shelf has handed you a large toolkit — every chart type, every candle, every pattern, every indicator, leading and lagging. It would be a poor teacher that spent all that time and never told you, plainly, what the tools cannot do. That is this module's whole job: to name the ceiling.

Here is the ceiling in one sentence. Technical analysis reads behaviour and shifts odds; it does not predict, it contains no edge in the tool itself, and even its famous "self-fulfilling" power runs out fast. Everything the shelf taught remains useful inside those limits. Sold beyond them — as prophecy, as a machine that prints money — it becomes exactly the trap the industry around it is built on.

The reason to end the shelf here, before you build your own checklist, is simple. A tool you understand the limits of is safe. A tool you believe can do more than it can is dangerous in direct proportion to how skilled you are with it — because skill makes the false belief more convincing, not less.

Start with the biggest claim, the one that sells the most courses: that a chart can predict the next move. It cannot, and the reason is not a matter of skill or better indicators. It is structural.

A chart is made of one substance: prices that have already happened. Every line, every pattern, every oscillator is a rearrangement of past data. But the next move is not decided by past data. It is decided by things that are not in the chart at all — an earnings surprise, a policy change, a large fund's decision to buy or sell, a piece of news that lands tomorrow morning. The chart cannot contain information that does not exist yet. So the most it can honestly do is describe what has happened and nudge the odds of what comes next. It cannot forecast, because a pattern shifts probabilities, it never issues a promise.

Now the second claim, subtler and half-true: that patterns "work" because so many people watch them — the self-fulfilling prophecy. There is something real here. When thousands of traders watch the same round number or the same moving average, their coordinated buying and selling really can make a level hold, for a while. A self-fulfilling pattern is exactly this: a level that partly works because the crowd watching it acts on it together.

But — and this is the part the sellers skip — that crowd is small. It is dwarfed by a single determined institution, or by a wave of news-driven supply, or by an index fund forced to sell. The self-fulfilling effect holds a level only until it meets something bigger than the crowd, and then it breaks as if it were never there. The prophecy is real and strictly limited, because real economic pressure eventually overwhelms whatever the chart-watchers were doing.

The third and deepest limit is that there is no edge inside the tool itself. Every indicator on this shelf is public. RSI, MACD, every candlestick, every pattern — anyone can see them, for free, at the same instant. A signal visible to everyone cannot be a private advantage. This is why the shelf keeps returning to the same wall: there is no single tool, however clever, that removes the need for judgement and risk control. The edge, if you have one at all, was never in the chart.

The three limits — no forecast, self-fulfilling only to a point, no edge in the tool — are easiest to see together on a single behaviour every reader has watched: a support level that holds, and holds, and then does not.

Read the three limits directly off it. No edge in the tool: the level at 150 was visible to everyone; the fact that it held three times gave no reader a private advantage, because every reader could see the same line. Self-fulfilling only to a point: the holds were partly real — watchers bought there and their buying defended it — but the effect was finite, and a single wave of supply outweighed the whole crowd. No forecast: nothing on the chart before the fourth test could tell you that this touch would break rather than hold. The break was decided by supply that arrived from outside the chart, information the chart could not contain until it was already price.

Notice what the table is not saying. The left column is genuinely valuable — reading behaviour, shifting odds, timing, risk. The right column is not a list of the tool's failures; it is a list of jobs the tool was never built for. The danger is only ever in asking the tool to do a right-column job.

Watch the ceiling meet a confident reader. illustrative

A trader has watched a support level near ₹150 hold three separate times. Each bounce deepened his conviction. He has a story for it — "big buyers defend this line" — and the story is even partly true, because plenty of traders watch that round level and buy there together. By the third hold, the level feels less like a probability and more like a law. He decides it cannot break, and sizes his next buy accordingly: large, with a stop placed carelessly far away, because why protect against something that cannot happen?

Then the fourth test is different. A block of stock hits the market — a fund reducing a position, indifferent to the chart and to the crowd watching it. The buying that defended 150 three times is simply outweighed. The level breaks, and because so many traders had the same "it can't break" story, their stops trigger together and feed the fall. Price slices through 150 and keeps going. The very crowdedness that made the level hold made its failure faster.

The honest reading was available the whole time. The three holds were real evidence that the level had attracted buyers — useful, worth noting. They were never evidence that it would hold again, because the self-fulfilling effect is always finite and the next seller's size is never on the chart. A reader who understood the ceiling would have used the level exactly as far as it goes: a place where the odds of a bounce were a little better, protected by a real stop, sized so that the inevitable eventual break was survivable. He treated a probability as a certainty, and the ceiling collected the difference.

It is worth being precise about the specific things the toolkit cannot reach, because each one is a place people quietly expect it to help and are quietly failed.

It cannot see the news that has not happened. The single largest driver of tomorrow's price — a result, a policy, a shock — is by definition absent from a chart of yesterday's prices. No amount of indicator-stacking conjures it into view.

It cannot survive the rare, violent day on its own. Markets deliver far more extreme moves than any smooth reading suggests, and a single gap can leap straight over the neat level where your stop was resting. The chart drew the level; the gap ignored it.

It cannot rescue a strategy that only ever worked on paper. A tool that looked perfect in history can be pure fitting to a past that will not repeat, and no live chart will warn you — the warning had to come from honest out-of-sample testing, before you ever traded it.

And it cannot supply the one thing every reader wants it to: certainty. The tool's honest output is better odds, never a sure thing. Everything that turns better odds into a durable result — the stop, the size, the discipline, the humility to be wrong often — lives outside the chart, in you.

Every myth about charts is really a refusal to accept one of the three limits. Named plainly, they lose their grip.

1. Believing the chart predicts. It describes and shifts odds; it does not forecast. Anyone selling prediction is selling certainty the tool cannot hold.

2. Treating a self-fulfilling level as unbreakable. The crowd's coordinated buying is real but small. Every watched level breaks eventually, and the more crowded the belief, the faster the break when supply arrives.

3. Hunting for the one perfect indicator. A public tool cannot be a private edge. The search for the holy-grail indicator is a search for something that, by its nature, cannot exist.

4. Mistaking a run of luck for skill in the tool. A signal that "worked" five times running may be doing so by chance; the tool is the same whether the last five were luck or not. Count the sample before you trust the streak.

5. Blaming the drawing when a level fails. "I drew the trendline wrong" preserves the fantasy that a correctly drawn line would never fail. Levels fail because the crowd is finite, not because your pencil slipped.