Indicators

Add an indicator to a trading app and something seductive happens. A new line appears — smooth, deliberate, faintly scientific — and it seems to know things the raw price did not tell you. It bends and crosses and colours itself red or green, and it is easy to feel that you have switched on an instrument that reads the market the way a thermometer reads a fever.

So the question for this module is blunt: when an indicator flashes a signal, where did that signal come from? Is it new information arriving from outside — or is it the same price you were already looking at, run through a formula and drawn again in a different shape?

An indicator is a formula applied to past price, past volume, or both. That is the whole of it. You feed in numbers the market has already produced — yesterday's closes, this week's turnover — and the formula gives you back a new number, which the app plots as a line. Nothing enters the formula that the price and volume did not already contain. So nothing can come out of it that they did not already contain either.

This is the single idea the rest of the module defends: an indicator is a re-presentation of data you already had, not a source of new data. It re-arranges the past into a more readable shape. It cannot reach beyond the prices it was built from, because those prices are its only ingredients. A momentum line, an average, a band around the price — each is the same chart you were already reading, folded and smoothed and re-plotted.

That is why this module sits where it does — after you have learned to read a chart, a candlestick, and volume, but before RSI, MACD, and Bollinger Bands get a module of their own. Learn the honest nature of the whole family first, and the specific tools later become easy to keep in their place. The most reliable way to use an indicator well is to first ask how it would fool you — and the answer always starts with: by feeling like more than a re-drawing of the price.

Every indicator runs the same little pipeline, and seeing it drawn out drains most of the mystique. illustrative

First an input: some slice of price or volume history — say the closing prices of the last 20 days. Then a formula with a lookback period: how many past values to include, and how to combine them — average them, compare highs to lows, measure the gap between two averages. The formula almost always involves smoothing: blending many past values so the output glides instead of jumping. Out comes a line, plotted for you to read. And then the only part that matters — your read: what, if anything, the line is worth in this particular market.

Indicators come in two families, and the difference is just where the line is drawn.

An overlay sits on top of the price, in the same scale as the rupees on the chart. A moving average is the classic overlay: draw the average of the last N closes right over the candles, and it traces a calmer version of the same path. Bollinger Bands are overlays too — a moving average with a band above and below that widens when price gets jumpy. Because they share the price's scale, overlays answer questions like "is price above or below its own recent average?"

An oscillator is drawn on a separate scale below the price, usually bouncing inside a fixed range like 0 to 100. RSI and MACD and the stochastics live here. Because they are re-scaled, oscillators answer questions like "how stretched is price compared to its own recent behaviour?" — high in the range, low in the range, and so on.

But notice what unites the two families, because it matters more than what divides them: overlay or oscillator, both are computed from the same price and volume. Drawing the line in a different place does not give it a different source. It is the same data, wearing a different costume.

Take the simplest indicator of all — a moving average — and follow the arithmetic, because it reveals a property the whole family shares.

A 5-day moving average is just this: add the last five closing prices, divide by five. Tomorrow, drop the oldest, add the newest, divide again. That is the entire calculation — no more than you would do to split a restaurant bill.

Now watch what that averaging does in time. Suppose a stock climbs steadily to ₹160, peaks, and starts falling. On the day price peaks at ₹160, the five-day average is still dragging in four earlier, lower prices — so the average has not peaked yet. It keeps rising for a few more days, because the high price is still fresh in its window, until enough falling days pull it down. The average turns after the price turns. Always. This delay is the lag, and it is not a flaw to be tuned away — it is the direct consequence of averaging the past.

Every indicator inherits this in some form. RSI averages recent gains and losses — so it lags. MACD is the gap between two averages — so it lags twice over. The line always tells you where the price has been, dressed up to look like where it is going. Keep that straight and you will never again mistake a moving average that "turned up" for a moving average that predicted anything.

Here is the lag, made touchable. Below is a jagged composite price that climbs to a peak and falls, with a moving average drawn on top of it. illustrative Drag the lookback and watch two things move together.

First, the line never turns before the price — it cannot, because it is built only from prices that have already happened. Its high can only arrive after the price's high. Second, as you widen the window, the average gets smoother and more confident-looking, and it turns later. That extra smoothness is not extra knowledge. It is simply more of the past folded into the line.

The takeaway the widget makes on its own: you are not choosing between a good line and a bad one. You are choosing where to sit on a trade-off between reacting early and reacting cleanly — and no setting escapes the lag, because no average of the past can lead the present.

An indicator's usefulness is not a fixed property of the indicator. The very same line can be genuinely helpful in one market and actively misleading in another, and the thing that flips it is the market state.

In a trending market — price grinding persistently in one direction — a lagging line can earn a modest keep. A moving average that turns late still spends most of a long trend on the right side of it, so "late but usually right" is a tolerable summary. The lag costs you the turns, but the trend is long enough to forgive it.

In a range-bound, choppy market — price sloshing sideways with no direction — the same tools fall apart. An oscillator built to say "overbought near the top of the range, oversold near the bottom" will flip from buy to sell to buy again and again, firing signal after signal, none of which leads anywhere, because there is no trend for them to catch. The indicator is not broken. It is doing precisely what its formula says — and its formula is mismatched to a market that is going nowhere.

This is why "what did the indicator say?" is never a complete question. The complete question is: what data made this line, what does it lag, and what state is the market in right now? The same reading — an oscillator at 70 — is a shrug in a trend and a siren in a range. The number did not change. The context did.

Take one concrete case, because it exposes how easily lag gets misread as either magic or failure. illustrative

A composite stock has sold off hard, bottomed, and started to bounce. Three days into the recovery, a 14-period momentum oscillator finally turns up and crosses out of "oversold." A beginner watching it sees the cross and thinks: there — the indicator called the bottom. A second beginner, who was watching the price directly and already saw the bounce, thinks: useless — it turned up three days after the low.

Both have misread the same event. The oscillator did not call the bottom; it averaged three days of rising prices and, once that was enough to lift the formula, it turned. It could not have turned earlier, because on the day of the low there were no rising days yet to feed it. It is not magic (it followed the price) and it is not failure (following the price a few days late is exactly its job). It is a lagging summary behaving normally.

Now add a slower, 28-period version of the same oscillator. It turns up a few days later still — but it also did not fire during two false rallies the month before, when the faster line jumped and then flopped. Which is better? Neither, plainly. The fast line is earlier and noisier; the slow line is later and cleaner. Early and smooth are trade-offs, not moral qualities — a longer lookback buys you fewer false alarms at the price of a later turn, every single time.

Being clear about what indicators are fences off the errors that cost beginners the most.

An indicator cannot create information that is not in its inputs. Feed it only price and volume, and price and volume are all it can ever reflect — never the earnings quality, the debt, the promoter's plans, or the next big buyer's decision. It re-arranges the visible past; it cannot import the invisible future.

It cannot lead the price it is built from. An overlay or oscillator computed from past prices is arithmetically downstream of those prices. It can confirm a move after the fact and summarise a trend in progress. It cannot get ahead of the data that defines it.

It cannot issue a command. A cross, a threshold, a colour change — these are re-descriptions of what price already did, not instructions about what you should do. The decision to act, and the reason for it, has to come from you and from evidence the indicator never saw.

And it cannot rescue a bad thesis with a good signal. If the reason to own something is weak, no arrangement of moving averages makes it strong. At its honest best, an indicator helps you time an action you already had good grounds to take — the subject of a later module. It is a stopwatch, not a compass.

The same handful of illusions catch indicator users again and again. Name them and they lose their grip.

1. Treating the line as an oracle. The signal feels like it arrived from outside. It did not. It is the price you were already watching, run through a formula and re-drawn. No indicator knows more than the data it was fed.

2. Stacking duplicates and calling it confirmation. Three indicators built from the same prices are not three witnesses — they are one dataset in three costumes. Their "agreement" is guaranteed by construction, not by independent evidence.

3. Fitting indicators until one agrees with you. There are hundreds of indicators and settings. Keep adding them and one will eventually flash the signal you were hoping for. That is not the market talking; it is you shopping for a line that confirms a decision you already made.

4. Forgetting the lag. A line that "turned up" reports the past, a step behind. Reading a delayed echo as a forecast is the most common and most expensive indicator error.

5. Ignoring the market state. An oscillator flipping in a range is not broken — it is mismatched. The same reading means opposite things in a trend and in chop, and the indicator never tells you which you are in. You have to.

6. Treating thresholds as laws. '70 is overbought, 30 is oversold' are conventions, not rules the market obeys. Price can stay 'overbought' for weeks in a strong trend while the number screams sell.