Gooaye EP705 Notes: The Numbers Are Beautiful and the Stock Won't Move, Because Everyone Onboard Already Knows

Contents

1. First, where the saying stops
2. How he reasons from “it won’t move” back to “everyone onboard already knows”
3. Why the second name runs harder than the leader
4. What makes a dead stock move
5. Where the reasoning breaks
6. How do you measure “crowded”
7. A few other things worth keeping
8. Sources worth checking
9. The One Thing to Take Away

Great wisdom does not look like wisdom; great strategy does not look like strategy; great courage does not look like courage; great profit does not look like profit.

—— Six Secret Teachings, “Civil Strategy: Opening Instruction” (attributed to Jiang Ziya; compiled in the Warring States period, translation mine)

In the 2026-10-14 episode of Gooaye (EP705), host Hsieh Meng-kung said the tape since early September has been kind to anyone willing to fade momentum: pick up what people throw away, trim a little when everyone charges in, and the index keeps handing you another chance to buy while still making new highs over time. What he spent the most time on was something else — several small-cap equipment names he holds have posted violent monthly revenue for months, with forward EPS easy to pencil in, and the stocks simply don’t move, because the seats are filled with the trading community and institutional money, everyone holding the same information, and nobody willing to put up more cash. He offered two ways out: print a number above what the market expects, or let one hard selloff shake out the impatient. The call comes with a precondition — the index is still grinding higher. When the trend turns down, clean positioning won’t save a stock either.

First, where the saying stops

He knows “don’t go where the crowd is” starts an argument the moment he says it, because “crowded” has no definition. The crowd in 0050, VOO or VT is beyond counting, and you can still buy them. Everyone now knows to buy AI, which doesn’t mean you should deliberately catch beaten-down consumer or auto names instead. He is talking about positioning at the single-stock level: inside the small equipment names he trades, which specific tickers have the whole trading and institutional crowd already onboard.

How he reasons from “it won’t move” back to “everyone onboard already knows”

The chain is short enough to be worth keeping verbatim. For a stock to go up, new money has to come in. Those equipment names print violent revenue, month after month, and their forward EPS looks excellent — and everyone onboard can do that arithmetic too.

His test is to use himself as the sample: if this month’s revenue came in another 50% or 100% higher, would you want to add? Yes. But the number it actually printed matches what you expected, you already know it’s good, and you have no plan to sell something else to buy more. Everyone onboard is parked in that same state, so the bid is simply absent.

Hence his reading when good news lands and the stock sits still: that number has already been bought. The uglier version is good news followed by a drop, because the people stuck higher up use the news to trim. The stock looks heavy; it can’t lift, and it won’t commit either.

Why the second name runs harder than the leader

This is the part that landed hardest for me. Anyone who has traded a theme has seen it: you know the industry, you know which company genuinely benefits and takes the most orders, and the leader doesn’t run. The other one, with fewer orders and a spot further down your own quality ranking, runs the most. Stretch the time frame and the leader eventually grinds up, slowly.

The difference is who is sitting there. The leader’s seats hold a crowd with synchronized information; new news reaches them without creating disagreement, and therefore without creating trades. The second name has fewer holders and scattered information, so the same headline still leaves someone feeling they got a bargain. Picking the biggest beneficiary and picking the stock that moves are two different jobs — I had been treating them as one.

What makes a dead stock move

He gave two routes, and both require changing who is onboard or what they know.

The first is a beat. Plenty of companies habitually guide low and deliver high: capable of 90, they tell you 75 or 80, because a high bar is hard to clear later. The odds of such a company beating are decent. In his words, the market already knew you were good-looking; it didn’t know you were going to win the pageant, and that’s the day the car drives off.

The second is a washout. The one-shot drops like this July, or last April, are more than the fearful, the impatient and the in-and-out crowd can hold, and they get asked to leave. Those names that chop sideways for months and then rip straight up after a crash came from exactly this.

Where the reasoning breaks

My first thought on finishing was that the positioning explanation and the fundamental explanation look identical in the moment. A stock that ignores good earnings might be full of people who already know, or might be about to print numbers that get worse.

He left himself a hook in this very episode. Memory is the tightest, strongest thing in the market, and the stocks are relatively weak. He didn’t assign that to positioning. His line was to watch whether the coming earnings carry it. If the tightest, strongest name can’t deliver, be careful with every price-hike theme, because the market will rotate from trading supply-chain bottlenecks toward hunting for things that actually produce value.

So the method carries an honest limit: in the moment you can’t tell which cause you’re looking at, and you wait for the next print to check your answer. If it carries, the earlier stillness was positioning, and you traded patience for room. If it doesn’t, the stillness meant the market smelled it before you did.

That also answers the question I keep wanting to ask — when you’re underwater, do you cut or hold? Rephrase it: what is your reason for holding? If the reason is “the numbers will get better,” the next earnings print is your answer key. If the reason is “it’s already fallen enough,” nothing will ever verify it for you. A listener question this episode had exactly that shape: he saw theme A moving, bought B instead, B went nowhere and then lower, and he asked whether to cut, add, or trust his own eye. The answer: your thesis was that B goes up when A goes up; A went up and B didn’t, so the thesis was wrong, and the thesis is what needs fixing.

How do you measure “crowded”

This is the least transferable piece of the whole method, and the place where he hands you no tool. He judges crowding by feel from inside the circle — which tickers everyone in the trading chats is already long. Most people can’t measure that.

What you can do a step back is watch proxies: months of good numbers with no price response, good news closing red on the day, headlines every day while volume shrinks. None of these are precise, and they’re good enough as a reminder not to size up here. His own practice is to keep buying such names, just never heavy — heavy means falling asleep waiting up there.

A few other things worth keeping

Broadcom’s CEO, his casino metaphor, and his own contradiction. At the DBS forum he said many investors don’t know what they’re invested in, set extreme expectations anyway, and then discover they’re in a casino; he also said most companies find no return on AI, like a kid in an arcade burning ten dollars of tokens in ten minutes and learning nothing. Hsieh put a question mark on the second half: the man sells shovels, and the harder enterprises burn tokens, the more money flows to his customers, who come back for more XPUs. The episode also relayed his September arithmetic — a hundred billion spent on frontier models earning a hundred and twenty billion, a hundred billion on open-weight models earning thirty — which means the more enterprises waste, the better his customers’ revenue looks.

AI collapsed the cost of entering an unfamiliar field. Three years ago, understanding why HBM stacks, what process the stacking needs, and where its weaknesses are took stacks of papers and sell-side reports. Now one round of questions gets him to roughly a 60-out-of-100 industry baseline, with contacts filling in the front-edge details. He flags two thresholds: a free model, or one without heavy reasoning, hands back different quality; and what you habitually feed it is a second threshold.

Whatever you reward is what grows. Companies that made “who burned the most tokens” a performance metric got people burning tokens with nothing to show. He scales it up to taxes: tax short-term trading and short-term liquidity thins; stack levies on dividends and more people sell before the ex-date and buy back after, with brokers happily building the cross-trade service for you.

He won’t put stops where everyone else puts them. No fixed 10% or 20%, no single moving average, because those are the levels most people watch and the easiest place to get shaken out. His line: he doesn’t mind eating a pullback, he minds missing the fish. And the same 15% drop means one thing when you’re falling alone and another when the whole index is falling with you.

Someone running 4x leverage lives in a different world than you. A 1% index drop is 5% on his book, so he tells you it’s a crash. Borrow his emotions and you’ll size your book by someone else’s judgment.

Signing up for a race is the adult version of a test. His tennis serve is now landing above 80% of the time, so he wants to enter the local club ladder and a neighborhood cup, cannon fodder odds and all. As a kid he hated exams; as an adult he wants something that checks whether he actually got better. Listening to that, I thought: in investing, the thing that forces you to check your answer is called the next earnings report.

Sources worth checking

• Gooaye EP705 (2026-10-14)
• Broadcom CEO Hock Tan’s remarks at the DBS Asia Future Forward forum (relayed in this episode; covered by several financial outlets)
• The September Broadcom arithmetic relayed in this episode, comparing returns on frontier versus open-weight models

The One Thing to Take Away

Good news only has a price while somebody still doesn’t know it. Once everyone can compute the same number, that number is already in the price, and however good it is, it buys you no new bid.

Here’s something I’ve tried that works outside investing too. Next time you go into a conversation that matters — a raise, talking a family member into a different decision, pitching a plan to a friend — write two columns before you open your mouth: on the left, three things the other person already knows; on the right, the one thing only you know. Then give your opening to the thing on the right. However hard you push the three on the left, their reaction to those has already happened.

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.