Averaging Down Stocks: When It Fails, When It Works

The first time it works, it feels brilliant. A good stock dips 15% on a noisy quarter, you buy more, it recovers, and your average cost looks clever. You learn a lesson — "buy the dip" — and you file it away as a rule. The trouble is that the lesson was the wrong one. What worked was not the act of buying lower. It was that the business was fine and the price was wrong. Those are completely different things, and the next time you reach for the same move, the business may not be fine at all.

Averaging down — adding to a position as the price falls — is one of the most respectable-sounding ways to lose money in the Indian market. It borrows the language of value investing ("be greedy when others are fearful") to dress up a behaviour that is often pure loss aversion: lowering your average cost so the paper loss looks smaller, on a screen, today. And the deterioration you're averaging into is frequently visible in plain sight. Across more than 1,500 listed Indian companies tracked on Inve, thousands of management commitments are logged, and on a transcript record only about two years deep, barely half of them are delivered as stated while hundreds were quietly dropped and never mentioned again (Inve data, as of 2026-06-12). That delivery rate is a snapshot of an evolving record, not a settled completeness figure — but the direction is clear enough: averaging down into one of those dropped commitments is not conviction. It is throwing good money after a thesis the company has already stopped honouring.

Take a real one. Vedant Fashions — the company behind Manyavar — guided its investors in October 2024 to keep expanding retail floor space at "plus/minus 1%, 2%" around its long-run 14–15% rate (Vedant Fashions Q2 FY25 concall, 30 Oct 2024). By the January-quarter call, that target had been walked down to "decent growth," then to 8–10% gross additions, then quietly recast as "improving the quality of our retail footprint" — and by the Q3 FY26 call it had vanished from the script entirely (Inve Promise Tracker, status walk on the retail-area target; illustration, not a view on the stock). An investor anchored to the ₹848 the stock once touched, watching it fall toward ₹411, could have averaged down the whole way (market data, as of 24 Jun 2026). The chart said "cheaper." The transcripts said "the thing your thesis rested on is being abandoned in slow motion." This piece is about telling those two apart — before you click buy.

Why does averaging down feel like discipline when it isn't?

Because it wears the costume of the thing it most resembles in name only. Real value investing buys more of a sound business when the price falls relative to unchanged or improving value. Averaging down, as most people practise it, buys more of a position when the price falls — full stop — without re-checking whether the value fell faster.

The emotional engine underneath is loss aversion married to anchoring. You paid ₹800. It's ₹411. The loss stings — losses are felt about twice as keenly as equal gains — and the ₹800 anchor makes the red number feel like a wound that must be healed. Buying more at ₹411 brings your average down toward ₹600. The wound looks smaller. Nothing about the business has changed, but the number that hurts has moved closer, and that relief is what you're actually buying. It is medication for a feeling, billed to your portfolio as a strategy.

Here is the homely version, and the one to commit to. Averaging down on a deteriorating business is like a batsman who keeps playing the same loose drive that has already got him out twice — except now he's told himself he'll hit the next one harder to make up for the runs he lost. The shot didn't fail because he played it gently. It failed because it was the wrong shot for the ball. Doubling the force doubles the risk of the same dismissal. More conviction applied to a broken thesis is not bravery. It's the same mistake, with more money on it.

When does averaging down actually compound — and when does it dig the hole deeper?

The whole question reduces to one distinction: is the price falling because the market is wrong, or because the business is? Get that right and averaging down is one of the most powerful things you can do. Get it wrong and it is the fastest way to turn a small mistake into a portfolio-sized one. Two real falls from the last year make the difference concrete.

Averaging down compounds when the thesis is intact and the price isn't. Through FY26, Infosys traded down from a 52-week high of ₹1,728 to a low of ₹1,026 (market data, as of 24 Jun 2026) — a roughly two-fifths haircut. But look at what management was actually doing while the screen bled. Its FY26 constant-currency revenue-growth guidance was set at 0–3% in the Q3 FY25 call, then revised up quarter after quarter: 1–3% at Q4 FY25, 2–3% at Q1 FY26, 3–3.5% by Q2 FY26 (Inve data, FY26 guidance walk, as of 2026-06-12). Operating margin held flat at 24% across FY24, FY25 and FY26 on a standard definition — comfortably inside, and a couple of points above, management's own volunteered FY26 operating-margin band of 20–22%, which Inve's Promise Tracker records as achieved — and sales grew 9.6% in the most recent year (company filings, FY24–FY26). The guidance was being reaffirmed and then revised upward; the metric they volunteer most — margin — never wobbled and stayed at or above its guided band. A falling price laid over a rising guide is the textbook dip worth adding into. But "the guide went up" is a necessary, not a sufficient, condition: guidance can be reaffirmed and still turn out wrong, and one intact metric does not validate a whole thesis — so a rising guide is a reason to keep looking, not a green light to add on leverage or to concentrate. (Illustration, not a view on the stock — copy the pattern of testing the record, not the position itself.)

Averaging down digs when the price is falling because the fundamentals are. Vedant Fashions is the mirror image. Operating margin compressed from 48.2% in FY24 to 46.4% in FY25 to 43.8% in FY26 — and because margin fell faster than sales rose, absolute operating profit shrank from ₹658 crore to ₹629 crore even as sales crept up 5% (Inve data, FY24–FY26). Same-store sales growth, the number a retailer lives or dies on, came in at just 1.8% for the first nine months of FY26 (Vedant Fashions Q3 FY26 concall, 13 Feb 2026). And the floor-space expansion target that justified the premium multiple drifted from 14–15% to silence over four calls. Here, every rupee you add buys more of a business whose value is falling beneath the price — the classic value trap, and averaging down is what converts a loss you could have cut into a loss that defines your year. (Illustration, not a view on the stock.)

The tragedy is that the two falls look identical on a price chart. A 40% fall is a 40% fall whether the cause is a passing fear or a permanent impairment. The chart cannot tell you which. Only the record can — the guidance, the delivery, the dodges, read across quarters. Which is exactly the work most people skip in the moment they most need it, because the falling price creates urgency and urgency kills analysis.

How do you tell a temporary dip from a deteriorating business?

You invert the question. Don't ask "is this cheap now?" — a falling price will always make something look cheaper. Ask instead: "what would it look like if this business were quietly breaking — and does the record rule that out?" Then go check the record against that test, not against your hope.

Three things to read before adding a single share, all of which sit in the primary source rather than the price:

The guidance walk. Pull every forward commitment management has made over the last 4–6 quarters and line them up. Is the same target being reaffirmed (cyclical dip, thesis intact) or has it drifted down and then vanished (structural problem, thesis breaking)? Vedant Fashions is the walk in miniature. In October 2024 the answer to "does 15% retail-area growth continue?" was: "from a midterm perspective… plus/minus 1%, 2%, we will be in that sort of a retail area expansion mode only" (Q2 FY25 concall). Sixteen months later, the same management's framing was: "this year, strategically, we were focusing on improving the quality of our retail footprint… store expansion should start normalizing from next 2-3 quarters" (Q3 FY26 concall, 13 Feb 2026). Read those two sentences back to back and the target didn't get missed — it got redefined into prose. The net floor space they actually added that quarter was about 5,500 square feet, against the 170,000–180,000 they once guided per year (Vedant Fashions Q3 FY26 concall). A target that only ever moves one way and then goes silent — whether it's floor space, a margin band, or a debt-reduction commitment — is the single clearest sign you are averaging into deterioration. roughly a third to nearly half of the companies Inve tracks have at least one commitment that was simply ghosted (Inve data, as of 2026-06-12); the names that tempt you to average down are often the ones in that group.

The Q&A on the sore spot. Find the question an analyst keeps asking — receivables, margin walk, client concentration, the segment underperforming — and read the answers across quarters. Direct answers that get more specific suggest a business confronting its problem. Answers that get vaguer, or a topic deflected for three straight quarters, suggest a problem getting worse out of sight. Inve has logged roughly 1,300 such dodges — evasive, deflected or downplayed Q&A responses — across its tracked calls (Inve data, as of 2026-06-12), each tagged to a topic and a quarter — a persistent evasion streak is a flag no price chart carries, and the graded Q&A in each concall summary surfaces it quarter by quarter.

The metric that went quiet. Identify a number management used to volunteer proudly and check whether it's still being mentioned. Vedant Fashions had guided a 4–5% store-closure rate for FY25 as part of a "cleanup"; the final figure was simply never reported back to shareholders (Inve Promise Tracker, store-closure-rate target). Things that stop being said when they would be embarrassing are louder than anything that gets said.

The convergence is the verdict. A receding-then-ghosted floor-space target, plus margins compressing three years running, plus same-store growth collapsing to 1.8% — that is not noise to buy into. That is a thesis dissolving in real time, and the cheaper price is the market knowing it before you do.

What's the rule that keeps averaging down honest?

Decide the basis for adding before you look at the price. Write it into your process: "I will add to this position only if, re-reading the latest results and guidance from cash, I would open the position at today's price — and only if the specific things my thesis depends on are still being delivered." Notice that the buy price and the size of the paper loss appear nowhere in that rule. The "would I open it fresh at today's price?" test exists to defeat a specific bias: the endowment effect — placing a higher value on a position simply because you already own it, even though you would never buy it fresh at this price. If the only reason to add is that the average cost will look better, you are not averaging down. You are anchoring, and paying for the privilege.

Inve's Promise Tracker exists for exactly the check this rule demands — it holds each commitment's original wording against its current verdict (Achieved, On Track, Missed, or Ghosted), so the difference between a dip worth buying and a thesis worth abandoning is visible without re-reading a year of transcripts under the pressure of a red screen. A portfolio of 15–25 stocks is 15–25 guidance walks to keep straight; nobody does that reliably from memory at the exact moment urgency is highest.

Where we could be wrong: cutting every falling position is its own expensive error, and our two-year transcript window can flatter the value-trap case. The Infosys fall above is the steelman for averaging down — a ~40% drawdown sitting on top of guidance that rose four calls running and a margin that never moved; an investor who sold that dip on the chart alone would have sold a business doing exactly what it said it would. Markets do over-react; good businesses do get thrown out with the sector. And a guidance "walk" of four or six quarters is a read on how management is communicating now, not a lifetime verdict — a target can go quiet for one bad year and come back. The discipline is not "never average down." It is "average down only when the value is intact and the price is wrong — never when the price is right and the value is the thing falling." The same chart sits behind both. The record tells them apart.

What should you believe before you add?

Not "it's cheaper than I paid." Cheaper-than-you-paid is information about your past, not the company's future. Ask the owner's question instead: what must I believe, five years out, for adding here to work — and does the record support it or contradict it? For Infosys the believable story was on the page — margin steady, guidance rising. For Vedant Fashions the story you'd have to believe (floor space resuming its old growth, margins re-expanding, same-store sales recovering off 1.8%) is the precise thing management stopped committing to on the record. If you can defend the five-year story from the guidance walk and the Q&A, averaging down is conviction. If your honest answer is "I just don't want to book the loss," then the cheaper price is a trap, and the only thing you're lowering is the number that hurts — right up until it hurts a great deal more.

Frequently asked questions