"It's down 60%" is not a reason to buy anything. It is not even a fact about the company — it is a fact about the last few years of its share price, and it is compatible with a business that is thriving, one that is dying, and one being taken over.
We know because we took all three apart this year, and reached three different answers. Same symptom, three diagnoses. Here is how they separate.
The market changed its mind, the accounts didn't
Adobe — down about 64% from its 2021 peak, as at our 31 July 2026 report
When we priced Adobe at $250.41 on 31 July 2026, the shares were roughly 64% below their 2021 peak. Over the same period revenue had risen about 60% and earnings per share about 74%. Gross margin was at an all-time high of 89%, on a 36% return on invested capital.
Nothing in the accounts had broken. What collapsed was the multiple — the market decided generative AI would make the product unnecessary, and stopped paying for profits it expected to disappear. By our report the shares changed hands at about ten times free cash flow.
This is the only one of the three shapes that can be a bargain, and it is the rarest. It requires the market to be wrong about the future while being right about the present, which happens — but far less often than the number of people claiming to have found it would suggest.
Seven weeks later we revisited Adobe after its third quarter, and found the first real evidence in the gross margin: 88.1%, down 1.2 points, its first break in eight quarters, because Adobe now pays to run other companies' AI models inside its own product. We cut our score from 7.4 to 6.8.
That is what this shape looks like when it is working properly. The thesis did not die — it got a first piece of evidence against it, and the evidence was in a line nobody was arguing about.
The business actually deteriorated
Nike — $44.57 in July, $36.36 in September. Our own call, tested.
In July we wrote about Nike at $44.57, near a decade low, and said watch the turn — with a zone around $40 where we would be interested. By 17 September 2026 the price was $36.36, down another 18.4%, and the zone had arrived.
We did not buy the idea. Here is why, in the language of the accounts: clean earnings per share of $1.58 against a dividend of $1.64 — a payout above 100% of clean earnings. The buyback switched off. Gross margin, stripped of a one-off refund, at 40.2% and still falling. A chief financial officer three weeks into the job.
The price hit our number and the evidence did not. That is the whole distinction. A price falling because a business is getting worse is not a discount — it is the market being approximately right, in advance, and the "cheapness" is just the accounts catching up to the chart.
A price target reached by deterioration is not a price target reached.
The disruptor got disrupted — and someone bid for it
PayPal — 82% below its 2021 peak, as at our 2 July 2026 report
PayPal at $56.15 on 2 July 2026 was the most extreme fall of the three: 82% below its 2021 peak. On the ratios it looked irresistible — about ten times earnings, an 11% free cash flow yield, net cash on the balance sheet.
But the fall was not an opinion, and it was not a collapse either. The moat had genuinely eroded: the company that once disrupted payments had been disrupted in turn. And by the time we wrote, the situation had a fourth element that neither of the other two cases had — a live $53 billion takeover bid, which the board had called too low, and which was holding the price up.
Our verdict was neither "bargain" nor "trap": not a trap, just no longer a bargain at this price. The cheapness was real and it had already been noticed by someone with more money than us. A stock supported by a bid is not priced on its business any more — it is priced on the negotiation, and that is a different game with different rules.
Two questions that separate all three
| Did profits fall? | Did the multiple fall? | What it is | |
|---|---|---|---|
| Adobe | No — rose ~74% since 2021 | Yes, heavily | A change of opinion. Can be mispricing. |
| Nike | Yes — dividend above clean EPS | Yes | A change of business. Cheapness is the accounts catching up. |
| PayPal | Moat eroded, cash intact | Yes, to ~10× | A change of position — real value, already found by a bidder. |
The mechanical version takes ten minutes. Pull up earnings per share for the last five years and the price-to-earnings multiple for the same period. The price fall is those two numbers multiplied together, and separating them tells you almost everything:
- Multiple down, earnings up → the market is pricing a future. Your job is to decide whether that future is real. This is where bargains live, and where the best-argued mistakes live too.
- Multiple down, earnings down → the market is pricing a present it can already see. Usually you are early, not clever.
- Multiple flat, earnings down → nothing is on sale. The price fell exactly as much as the business did.
- Multiple down, earnings up, and a bidder in the room → you are no longer valuing a business, you are guessing at a negotiation.
Being right about the business is not enough
We got Adobe's business right in July and the shares went nowhere for seven weeks. We identified Nike's zone in July and it arrived in September carrying evidence that made it uninvestable. We called PayPal cheap and someone richer had already called it cheaper.
None of those are failures of analysis. They are what analysis actually looks like: three careful answers, of which one was a hold, one was a pass, and one was a shrug. The reason to do the work is not that it converts falls into opportunities. It is that it stops you buying the two out of three that were never opportunities at all.
Each stamped with the date and the price it was written at, so the calls can be checked.
What to do with this on Monday
Take the worst position in your portfolio and do the ten-minute split: earnings then versus earnings now, multiple then versus multiple now. Most people have never done it for something they own, and the answer is frequently not the one the story in their head has been telling them.
If the earnings are up and only the multiple fell, you have a thesis to defend and a reason to hold. If the earnings are down too, you are not holding a bargain — you are holding a company, at whatever it is now worth, and the fall you already suffered is not an argument for anything.
Frequently asked
How do I tell a bargain from a falling knife?
Separate the price fall into its two causes: the multiple the market pays, and the profit it is paying for. If profits kept rising while the multiple collapsed, you are looking at a change of opinion. If profits fell with the price, you are looking at a change of business. The first is sometimes an opportunity; the second is usually just early.
Is a stock down 60% automatically cheap?
No. Down 60% tells you what happened to the price, and nothing about what the business is worth. A company whose earnings have halved and whose multiple is unchanged has fallen 50% and is exactly as expensive as it was. The starting point matters too: a great deal of what falls 60% had risen 200% first.
What is a value trap?
A company that looks statistically cheap on figures that are still falling. The ratios improve on paper every quarter while the business deteriorates underneath them, so the stock is permanently 'cheap' and permanently going down. The tell is that the cheapness comes from the denominator shrinking, not from the price being wrong.
Should I wait for a stock to stop falling before buying?
Waiting for confirmation costs you part of the discount, and it removes the worst outcome — buying a business that is still deteriorating. If the thesis depends on evidence that has not appeared yet, waiting for the evidence is not timidity, it is the thesis. If the evidence is already in the accounts, the price chart is not information.