A Ridge From Here, a Peak From There: How One Reader Gets a Different Mountain Out of the Same Earnings Call

Sideways, a ridge; from the end, a peak —
near or far, high or low, never the same mountain twice.
I cannot tell the true face of Lushan,
for I am standing inside it.
— Su Shi, “Written on the Wall of West Forest Temple” (1084; translation mine)

What this is about

A week ago I wrote up what I’d learned from following Serenity through the late-July AI deleveraging — the week price and the ledger went separate ways. Since then I’ve kept reading. Over the past two weeks he posted eighty-odd times, and close to ninety percent of it was public (my own count, through August 12).

Somewhere in the middle of that pile, the thing worth writing about changed. It wasn’t the industry news anymore — for that, go follow the account itself; firsthand always beats secondhand. It was the way he reads. The market looked at an earnings call and drew one conclusion. He looked at the same page and drew another.

Like the old poem says: the mountain didn’t change. The place you stand did.

One earnings call, two readings

In early August, a US optical transceiver maker held its earnings call. Management’s message, roughly: capacity is expanding hard, most of the new capacity will be in the US — and for now it still isn’t enough, because customer demand runs twenty to forty percent above what they can currently supply (Serenity pulled that passage from the transcript on August 6; I checked it against the call and I’m paraphrasing here. One fair addition: by management’s own numbers, if the capacity targeted for end-2027 arrives on schedule, that gap closes — so the scarcity is present-tense, not permanent).

Most people read that call as a list of problems: not enough capacity, can’t keep up, missing from the first wave of a new technology. All true.

His reading went the other way. If even the most aggressive expander in the industry can’t catch demand, then having qualified capacity at all is becoming scarce — and that scarcity accrues to every laser maker in the chain with independent production.

What I admired most: he didn’t use the second conclusion to cancel the first. He said plainly that both are true at once. This company has its troubles, and the whole chain is tightening. Holding a bearish fact and a bullish implication on the same table without letting them fight — I don’t see that skill in most research.

His dashboard has no price column

On August 8 he wrote a line I copied down: “the company at $140 and the company at $75 are the same company” (my paraphrase). Through July’s crash, apart from prices moved by forced selling, nothing fundamental changed.

Out of context it sounds like a consolation. Inside his method, it’s literal. What he tracks is capacity ramp schedules, customer qualification timelines, lead times, when expansion capex lands on the balance sheet. There is no price column on that dashboard. So when prices break, his thesis contains nothing that can be frightened.

I wrote this down as a discipline: whatever unit your judgment is expressed in is the unit that can falsify it. A judgment written in price gets stress-tested every trading day. A judgment written in capacity and timelines can only be overturned by capacity and timelines. Looking back at the days in July that scared me, the fundamentals had done nothing. I had written my judgment in the shape of a price, and price is what moved.

The rotation has an order, and he writes it down in advance

In late July he posted a roadmap-style projection: whichever layer the capex flood reaches next gets repriced — GPUs a few years ago, memory this year, and then a named list of what comes after, with years attached (co-packaged optics, glass substrates, high-voltage power architecture, all marked 2027).

I don’t know whether his years will prove right — neither does he, and forecasts like that are supposed to be graded by time. What caught my attention was the form: he doesn’t predict direction, he predicts structure. On price moves he openly claims no edge. On which layer snaps first and roughly how wide the gap gets, he commits — specifically enough to be scored later.

That’s the same path I’ve been walking this half-year: follow the supply chain upstream until you find the link that breaks first when demand doubles. Watching someone use the same method, in another market, arrive at an adjacent conclusion is a grounding feeling, less because he agrees with me than because the road evidently goes somewhere.

He puts the contradictions on the table

One more detail from these two weeks deepened my trust.

On whether upstream laser materials are truly scarce, the market holds two stories: Asian epitaxy suppliers say extremely tight; several US module makers say it’s fine. Faced with that, most writers quote the side that flatters their thesis. He posted both, then wrote: “depends who you ask.”

Same treatment for his own book. When old compliance news resurfaced about a company he openly calls a major holding, his comment was that the market is right to discount it. No defense offered.

Willingness to write down the half that hurts you is the fastest test I know for whether a researcher can be trusted. Conclusions can be wrong — everyone’s are, eventually — but the habit of how you select evidence cannot be hidden.

Checking against our own data

After reading him, I went back to my own data to check one thing: does the tightness he reads off US earnings calls line up with what other, independent paths see?

Path one, the Taiwan industry view. A Taiwanese industry podcast, Anchoring Notes, has an episode titled “Compute Isn’t Short — Interconnect Is: Why Laser Chips Are in Shortage” (EP6, guest: Jing). Looking up the chain from Taiwan’s position in the supply base, it lands on the same conclusion: the tightest layer this cycle isn’t compute itself, it’s the layer that connects compute. Two paths with no shared inputs ended up at the same place.

Path two, our own optical-interconnect industry review from January (data as of January 2026). The judgment then: the commercial acceleration of co-packaged optics would drive demand surges in upstream links — continuous-wave laser sources, fiber array units. Seven months later, what Serenity reads out of the transcripts is the same chain, one notch tighter.

Path three, price. From our own price data (through the August 11 close), that optical transceiver maker still trades more than forty percent below its 52-week high. The supply side is shouting shortage; the price side is shouting fear. One of them has to correct, and I don’t know which.

What I do know is what to watch. Borrowing his dashboard: lead times, qualification schedules, and who is prepaying. The stock price is not on the list. Those three correct far more slowly than price, and far more honestly.

Three disciplines to keep

One: the unit of your judgment decides whether you can be frightened. Write it in price and the market falsifies you daily. Write it in capacity and timelines and you get graded once a quarter. The second is not safer, just more honest: capacity and timelines were the only things you ever had a view on. Price was never something you could compute.

Two: audit the evidence-selection habit before the conclusion. Next time you read any analysis — including mine — look for one thing first: did the author write down anything that hurts their own case? If not, discount the conclusion no matter how well it reads.

Three: trading less has data behind it. Reviewing his own first half on August 10, he cited a statistic: infrequent traders returned 18.5% annualized versus 11.4% for frequent ones. I tracked down the original: Barber and Odean, “Trading Is Hazardous to Your Wealth,” The Journal of Finance (2000), built on 66,465 household accounts at a discount broker from 1991 to 1996. The sample is thirty years old, so don’t transplant the exact numbers to today. The direction, though, matches my own experience: the best decisions I made this half-year were the ones where I did nothing.

Sources and notes

• All industry descriptions come from the public posts of X researcher Serenity (July 29 – August 12, 2026; 78 public posts) and public podcasts; his quotes appear here in my own paraphrase and may contain transfer errors — go to the original account, and follow it directly
• The Taiwan-side discussion of the same theme: Anchoring Notes EP6
• On trading frequency and returns: Barber & Odean (2000), Trading Is Hazardous to Your Wealth, The Journal of Finance (sample period 1991–1996)
• Nothing in this piece comes from paid subscription content
• Previous piece: A Summer Insect Cannot Speak of Ice
• Su Shi’s poem is in the public domain; the English translation is mine

---

Disclaimer: This is a personal reading journal for educational purposes only. It is not investment advice, an offer, or a solicitation. It deliberately names no tickers, recommends no securities, and sets no price targets. Figures come from public posts, public podcasts, and my own database work; they may contain transcription errors — always defer to original sources. Investing carries risk. Make your own decisions based on your financial situation and risk tolerance, and consult a qualified professional where appropriate.

Raw Markdown

This article is an educational discussion of investment method. It is not advice to buy or sell any individual security, offers no target prices, and does not analyze any current holding. Investing carries risk; make your own decisions or consult a qualified professional.