Anchoring Bias in Investing: Can't Sell a Loser

A friend bought a mid-cap auto-ancillary stock at ₹400 in late 2023. It is ₹262 today. He has read the deteriorating results, he can see the margin slipping, and yet every time we talk he says the same thing: "Let it come back to ₹400, then I'll exit." Notice what is doing the work in that sentence. Not the business. Not the guidance. A price he himself paid two years ago, which the market has no memory of and no obligation to honour.

That is anchoring — the mind's habit of fixing on the first number it sees and judging everything afterward in relation to it. In investing it is not a personality flaw; it is a near-universal default, and it costs real money. There is a second anchor too, one almost no behavioural-finance article warns you about: the number management hands you on a concall and then quietly walks away from. This piece is about both — where the anchors come from, why they are so sticky, and how to replace them with a question the market actually answers.

What exactly is an anchor, and why does the buy price become one?

An anchor is a reference point your brain grabs early and then refuses to let go of. The classic experiment: spin a wheel that lands on a random number, then ask people to estimate something unrelated — the share of African nations in the UN, say. People who saw a high number guess higher. The wheel told them nothing. They anchored anyway.

Your purchase price is the most powerful anchor in your portfolio, and it is the most useless. The market sets price on the present value of future cash flows. It does not know — or care — that you paid ₹400. The ₹400 lives only in your demat statement and your head. Yet it becomes the line that separates "winner" from "loser," "vindicated" from "wrong," in a way that has nothing to do with whether the business is worth owning at ₹262 today.

Here is the homely version. Think of the buy price like the sticker price on a flat you bought at the top of the cycle. The locality has since lost its planned metro line, two builders have gone bankrupt, and identical flats now transact 30% lower. You can refuse every offer below your purchase price for as long as you like — but you are not "protecting" anything. You are simply choosing not to act on information the market already has. The flat is worth what someone will pay, today, knowing what they now know. So is the stock.

How does anchoring turn into the inability to sell?

Anchoring rarely travels alone. It teams up with loss aversion — the well-documented finding that the pain of a loss is felt roughly twice as keenly as the pleasure of an equivalent gain. Put the two together and you get the single most expensive behaviour in retail investing: holding losers far too long and selling winners far too early.

The chain runs like this. You anchor on the buy price. Falling below it registers as a "loss" — but only a paper loss, you tell yourself, not a real one, because you haven't sold. Selling would convert the anchor's verdict into a fact, and loss aversion makes that conversion feel unbearable. So you wait. "It'll come back." The waiting is not analysis. It is the mind protecting itself from an admission, using a number that means nothing to anyone but you.

The cruel part is the arithmetic. To "get back to even" from a 35% drawdown, the stock has to rise about 54%. You are demanding a harder recovery from a business whose prospects have, by your own reading, got worse — purely to make a two-year-old anchor whole. Meanwhile that capital sits dead. The cost is not just the ₹138 of paper loss per share; it is every quarter of compounding that the money could have earned elsewhere while you waited for a number to come back.

Invert the question the way Munger would. Don't ask "will it recover to ₹400?" Ask: "if I held cash today and someone offered me this exact business — these results, this order book, this guidance record — at ₹262, would I buy it?" If the honest answer is no, you are not holding an investment. You are holding an anchor.

The second anchor nobody warns you about: management's target

The purchase price is the anchor every behavioural-finance article mentions. There is a second one, specific to how Indian companies communicate, that almost none do: the guidance number management gave you, which quietly gets revised — or dropped — while you keep believing the original.

A management team says on a concall, "we are targeting margins close to 18%." You anchor on it. You build your thesis around it. Then, call by call, the language softens, the number drifts, and eventually it is simply not mentioned. No headline announces the change. There is no press release that says "we are abandoning our earlier target." The anchor stays lodged in your head long after the company has walked away from it.

This is anchoring you cannot fix by being disciplined about your own buy price, because the anchor was planted by someone else and the revision was designed not to be noticed. And it is far more common than most investors assume. Of 13,280 management commitments tracked on Inve across 1,519 listed Indian companies, 934 were simply ghosted — stated once on a call and then never referenced again — and roughly one company in three (35%) carries at least one such quietly-dropped commitment (Inve data, 2026-06-24). Each of those is an anchor sitting in some investor's thesis, doing damage, while the company has moved on.

A real said-versus-did: how one auto-ancillary's margin anchor drifted

Look at CIE Automotive India, a listed auto-components maker, because its record is public and instructive. On the July 2024 concall, management framed a fresh high-water mark with evident pride: India EBITDA margin "without any onetime impact has crossed 18% for the first time and it was 18.1% in Q2 CY24" (CIE Automotive India Q1 FY25 concall, 19 July 2024). A few months later, on the 25 October 2024 (Q3 & 9M CY24) concall, management was explicit about where that settles: "We were able to keep this close to 18%, a little bit less than 18%. So that's the target that we have" (Inve data, Q3/9M CY24 concall, 25 October 2024). An investor following these calls had a clean, confident number to anchor to. This is an illustration of how a guidance anchor behaves over time, not a view on CIE Automotive India shares.

Then watch what the number actually did, quarter after quarter:

Every figure is management's own reported number, call by call (Inve data, Q1 FY25 through Q4 FY26). The target was never formally withdrawn; it was simply never met, and the framing migrated from "that's the target" to softer language about being "around that 17.5% to 18% kind of margin" (Inve data, Q4 FY25). An investor who anchored to "close to 18%" in mid-2024 and never tracked the walk would, eighteen months later, still be carrying a number the business had quietly drifted below for six straight quarters. Six quarters of a target missed is not a scandal — margins move with the cycle — but it is precisely the kind of slow drift that an anchored mind never notices, because nothing ever announced the miss.

The same company shows the purer form of the trap too. On that 25 October 2024 call, management told an analyst that lifting India's export mix to 15% "should be quite easy in the next years" (Inve data, Q3/9M CY24 concall, 25 October 2024). On every call since, the specific 15% target has gone unaddressed — discussed around the edges, never reaffirmed against the number (Inve data, through Q4 FY26). That is a guidance anchor that didn't drift; it simply went silent. If it lives anywhere now, it lives in the heads of investors who heard "quite easy," wrote it into their thesis, and were never told it had lapsed.

How do you replace an anchor with something useful?

You cannot un-see your buy price. The brain does not have a delete key. What you can do is build a process that forces a fresh reference point on a schedule, so the anchor stops being the only number in the room.

Re-underwrite, don't re-anchor. Once a quarter, for each holding, write one paragraph answering: "Given everything I know today, would I open this position at today's price?" Note that the buy price appears nowhere in that question. This is the single most useful anti-anchoring habit, and it takes ten minutes per stock if you have the inputs in front of you — the latest concall summary gives you the quarter's guidance and graded Q&A without the full transcript.

Anchor to the thesis, not the price. When you buy, write down the two or three things that must stay true for the investment to work — the margin trajectory management guided, the order-book growth, the debt-reduction commitment. Those are your real reference points. If they break, the buy price is irrelevant; if they hold, a lower price is an opportunity, not a wound. The discipline is to track the thesis, not the ticker.

Track the guidance anchor explicitly. Because the management-target anchor is planted by someone else and revised silently, you need a memory layer that holds the original wording against what is being said now. Inve's Promise Tracker logs each commitment with its birth quarter, original quote, and current verdict — Achieved, On Track, Missed, or Ghosted — so a "close to 18%" target that has quietly drifted, or a "quite easy" export goal that simply went quiet, surfaces as a flag rather than living on, undisturbed, in your head. A diversified portfolio of 15–25 stocks means roughly that many concalls a quarter, hundreds of pages of forward commitments; nobody re-checks all of them against last year's wording by hand. (For the longer-run discipline of judging managements by what they deliver, see management guidance versus return on capital.)

Where this argument could be wrong

Anchoring is not the only reason to hold through a drawdown, and "the margin missed for six quarters" is not, on its own, a reason to sell. Sometimes the thesis genuinely is intact, the market is wrong about a temporary problem, and the right move is to do nothing — or, on a re-underwritten basis, add to the position. A target slipping from 18.1% to 17.6% in a cyclical auto-supplier, with margins still comfortably in the high teens, may be exactly the kind of noise a patient owner should ignore; the honest reading of CIE's own commentary is that demand cycles, not mismanagement, drove most of that drift. The data point is not "they failed." It is "the number you anchored to is no longer the number they stand behind — and you should know which is which before you decide."

So the test is never "am I down?" It is "would I buy this at today's price, knowing what I now know?" A "yes" backed by a re-read of the actual record is conviction. A "no" that you override because the screen is red is anchoring wearing the costume of conviction — often propped up by confirmation bias, where you go looking only for the reasons the stock will come back. We are confident about the method here; we make no claim about where any particular share price goes next.

What's the one habit that breaks the trap?

Separate the decision from the entry price entirely. Cover the buy-price column on your portfolio screen — literally, with a sticky note if you must — and look only at the business: the latest results, the guidance trajectory, the questions management dodged. Decide what you would do if you were starting from cash. Then uncover the column. Most of the time the anchor will have been quietly steering you, and you will not have noticed until you took it away.

The losers you can't sell are rarely held because of fresh analysis. They are held because a number you paid, and a target someone gave you, refuse to leave. The fix is not willpower. It is forcing a current reference point onto the table every quarter — your own re-underwriting, and management's own words checked against what they actually delivered. So here is the owner's question to end on: of every stock you hold, which targets are you still anchored to that management itself stopped repeating quarters ago?

Frequently asked questions