Anchoring

Here is a sentence almost every investor has said, out loud or under their breath: "I'll sell when it's back to ₹600." It sounds like a plan. It is really a confession — that a single old number has taken charge of a decision it has no right to make.

Picture the stock behind it. It once traded at ₹600. It sits today at ₹240. The moment you learned the ₹600, something quiet happened in your head: ₹240 stopped being a plain fact about a business and became a discount — a 60% mark-down from where the stock "should" be. The word cheap arrived before a single line of the accounts was read. You did not decide it was cheap; the ₹600 decided for you, and handed you the verdict dressed as your own judgement.

This is the whole subject of this module. A stock does not know what you paid for it. It does not remember its old high, its issue price, or the target some analyst typed into a spreadsheet. Those numbers live in your mind, not in the business — and yet they reach into your judgement and tug it around. Name that pull, and you can start to feel it happening. That is the first defence, and it is a real one.

The unsettling part is how little the anchoring number has to mean. In 1974, Amos Tversky and Daniel Kahneman ran an experiment that has been repeated countless times since. They spun a wheel of fortune marked from 0 to 100 — rigged, though the subjects did not know it, to stop on either 10 or 65 — and then asked people an unrelated question: what percentage of African countries are in the United Nations? People who watched the wheel land on 10 guessed, on average, around 25%. People who watched it land on 65 guessed around 45%.

Sit with that. The wheel was obviously random. Everyone could see it had nothing to do with the UN. And it changed their answers anyway, by a wide margin. This is anchoring-and-adjustment, and it is the mechanism beneath the whole module: a first number lands in the mind, becomes the starting point, and the person adjusts away from it toward the real answer — but stops too soon. The final estimate stays pulled toward wherever they began, even when they began at a spun wheel. Anchoring is that tug; insufficient adjustment is why it survives even when you try to correct for it.

Now change the wheel for the numbers a market throws at you all day — a buy price, an old high, an IPO price, a round figure, an analyst target — and notice that these are worse than the wheel, because they feel meaningful. The wheel at least announced itself as random. The ₹600 looks like information. So you don't even try to shake it off; you adjust down a little from it and call the result analysis. This module comes early on the shelf, right after the wiring itself, because the anchor sets the frame before any of your careful tools switch on. A valuation you build while a big first number is glowing in your mind is a valuation quietly bent toward that number, however honestly you did the arithmetic.

The market is unusually generous with anchors. Each one feels like a landmark — a fixed point you can measure from — and each one is really just a number that arrived first, carrying assumptions that may already be dead. Meet the common five by name, because a named anchor is one you can catch reaching for the wheel.

Your buy price. The most personal and the most stubborn. "I'll sell when it's back to ₹600" treats the price you paid as a level the stock owes you. It owes you nothing; it has never heard of you. Your buy price is a fact about your past, not about the company's future, and building a sell decision on it means letting a number that matters to no one but you overrule the business.

A past 52-week high. The 52-week high is printed on every quote screen, which is precisely what makes it dangerous — it is handed to you as if it were a reference price. But a high is only the most optimistic price a crowd reached in a year, under conditions that may be gone. Measuring today's stock as "40% below its high" tells you about last year's mood, not this year's worth.

A round number. ₹100, ₹500, ₹1,000 — the mind loves a round figure and treats it as a natural resting place. A stock "holding ₹500" or "breaking ₹1,000" feels significant, but the roundness is an accident of our base-ten counting, not a fact about the enterprise.

The IPO price. Being below the issue price feels like safety, as if the listing price were a floor. It is not. An IPO price is set by the seller and their bankers to place shares at the most they can get; it is closer to an asking price than a valuation, and the business can be worth well below it.

An analyst target. A ₹800 target next to a ₹500 price reads like "60% upside" — a measured gap. But a target is one model's output, stuffed with assumptions about growth, margins and the multiple. The gap only looks like a fact because ₹800 has become the anchor; open the model and the upside is only as good as the assumptions inside it.

The thread running through all five: the anchor smuggles in a distance. Cheap, upside, discount, floor — every one of those words is secretly a measurement from a first number, and the moment you notice you are measuring rather than valuing, you have caught the bias in the act.

Watch the pull happen in an ordinary case. illustrative

A consumer-durables company traded at ₹600 two years ago, on 70 times earnings, in the middle of a demand boom — everyone was buying air-conditioners, margins were fat, factories ran full. Today it trades at ₹240, on 32 times earnings. Along the way, margins have halved, inventory has piled up in warehouses, and demand has cooled back to something ordinary. A reader glances at the chart, sees ₹600 then ₹240, sees 70x then 32x, and feels the word forming: cheap.

Notice what did the deciding. Not the accounts — the reader hasn't opened them. The ₹600 and the 70x set the frame, and everything after was measured down from them. But the ₹600 belonged to a world of fat margins and empty warehouses that no longer exists. The 70x was the market paying up for a boom it expected to continue. Strip the anchors away and the real question is naked and much harder: what are these earnings — lower-margin, slower-growth, inventory-heavy — actually worth? It is entirely possible that ₹240 on today's business is dearer than ₹600 was on the old one. The fall is not the case. The business is the case.

This is anchoring-and-adjustment in its market clothes: the ₹600 and the 70x are the spun wheel, and the reader is adjusting down from them instead of valuing from scratch.

Now play with the pull directly. The toy below fills two rooms with people valuing the same illustrative business. One room is shown a first number — the anchor — before it guesses; the other is shown nothing and works from the business alone. Move the anchor and watch the anchored crowd drift after it, while the fresh-eyes room stays put near what the business is worth. Nothing about the company changes. Only the first number does.

There is one reliable way to loosen an anchor, and it is not to argue with it. It is to start somewhere else. Before you look at what the stock has done, build a rough value from the business itself — and build it as a range, not a single number, so no one figure can pretend to be the truth.

The range comes from a handful of honest assumptions, each of which you could defend to a sceptic: how fast can revenue plausibly grow, from pessimistic to fair to optimistic? What margin does the business earn through a full cycle, not just at the top? How much cash does it actually keep after the reinvestment it needs? What does the debt and any likely dilution do to the per-share picture? Turn each into a low case and a high case, and you get not a price but a band — "somewhere between roughly ₹X and ₹Y, and here is what would have to be true for the top of that band." That band is computed with no reference to ₹600, no reference to the 70x, no reference to what you paid. Only then do you turn the market price over and compare.

The test to keep in your pocket is a single question: what would I pay for this business with fresh eyes and no history here — if today were the first day I ever heard of it? A reader who paid ₹600 and a reader who just discovered the stock at ₹240 are looking at the same company and the same future. If the ₹600 reader would hold and the new reader would not buy, the only thing separating them is a number that exists in one person's memory. The fresh-eyes question strips that number out and forces both readers back onto the same ground: the business.

None of this asks you to pretend the past didn't happen. A buy price recorded in a journal, sitting beside the thesis that justified it, is a useful record — it tells you what you believed and lets you check whether the facts have moved. The anchor turns harmful only when the bare number, stripped of its assumptions, starts doing the new thinking. Keep the assumptions and the number is a note. Lose them and the number is a leash.

Freeing yourself from a first number is a real protection. It is also not everything, and pretending otherwise is its own quiet trap.

It does not tell you the business is good. Valuing with fresh eyes protects you from measuring down from ₹600; it does not stop you from building a cheerful valuation on weak assumptions. You can drop every anchor and still be wrong about the company — the fresh-eyes range is only as honest as the numbers you feed it.

It does not mean every reference point is poison. Some anchors are useful when their assumptions travel with them. Last cycle's peak margin is a fair benchmark if you remember what drove it. A past multiple is informative if you carry the conditions that earned it. The skill is not to burn every old number; it is to refuse to let a bare number — one whose assumptions have been quietly stripped off — set your estimate.

And "that's just an anchor" can itself become an anchor. Waved at a genuinely relevant reference — a replacement cost, a long-run margin, a sensible peer multiple — the bias label stops being a defence and becomes a way to dismiss real evidence. The move is not to sneer at every number that has a history. It is to ask, of each one, what assumption lived inside this, and is that assumption still alive?

The same handful of moves catch beginner after beginner. Named once, they are far easier to catch in yourself.

1. Calling a stock cheap because it is below an old high. "Down 60% from ₹600" is a distance, not a value. The old high was a mood; the current price has to be judged against the business, not against the peak.

2. Waiting for your buy price to "come back." Breakeven is a fact about you, not about the company. A stock climbing back to what you paid does not make the thesis right, and a stock that never does is not disobeying — it never knew the number.

3. Treating an IPO price as a floor. The issue price was set to sell shares, not to mark fair value. Being below it is not a margin of safety; it is just being below one particular first number.

4. Reading an analyst target as a measurement. "₹800 target, ₹500 now, so 60% upside" turns one model's assumptions into a fact. The upside is only as real as the assumptions inside the target, which are usually invisible in the headline.

5. Updating the price without updating the assumptions. A stock falls, the reader keeps the old story and simply calls it "cheaper now." The price moved; the thesis didn't. Re-price the business, not just the ticker.