Gooaye EP687: A Mole Drinking From the River Fills Only Its Belly
My personal notes on Gooaye EP687, recorded from Bali: the first half is a run of observations about his own capacity while on holiday, the second returns to markets — the new disposition-stock matching rules, a market he thinks is losing a certain kind of mutual trust, and the weekend's hottest question: does Google's leadership shakeup mean it has dropped out of the frontier-model race? His reading runs opposite to the market's, in four steps. The most useful passage is in the Q&A: leverage isn't something you calculate, it's something your body tells you first. Educational notes, not investment advice.
The tailor-bird builds its nest in the deep forest, but only uses a single branch;
the mole drinks at the river, but only takes what fills its belly.
— Zhuangzi, “Free and Easy Wandering” (English rendering after James Legge, 1891; public domain)
These are my personal notes on Gooaye EP687 (released 2026-08-12). This is not a transcript and not official content. If you want the full thing, please support the original show. What follows is what the episode prompted in me, plus my own framing.
What this episode is about
He recorded it from Bali.
The first half is a run of observations about himself while on holiday: how far apart his life actually is from a rich person’s, why buying out a venue isn’t necessarily more fun, and a failure he finds funny in hindsight — he had said on air, repeatedly, that this trip was going to be a proper holiday, and then cleared his index position on the plane to go buy single names instead.
The second half returns to markets. Taiwan’s new matching rules for disposition stocks went live, and he spent a session watching what the game looks like under them; then his sense that this market is losing a certain kind of mutual trust; and finally the weekend’s hottest topic — whether Google’s leadership shakeup means it has dropped out of the frontier-model race. In between sits a long run of listener questions, from leverage and research method all the way to unemployment and family.
Listening through, I think both halves are saying the same thing: know where your own capacity ends. The holiday half is about the capacity of consumption, the market half about the capacity of position size and of understanding, and the Google section about a company deciding which kind of business it wants to hold.
The main points
1. The holiday experiment produced two conclusions, both about limits. The outward one: strip out real estate (Taiwanese property spans an enormous range) and past a certain level of assets, life stops differing all that much. He uses business class as the example — the single most enjoyable moment is the first upgrade out of economy; if you have only ever flown business, you don’t actually know what’s good about it, because measured against your own sofa it isn’t much. The inward one is more interesting: he had planned to buy out a private area at the resort, since neither he nor his wife likes crowds — and his mother-in-law wanted the exact opposite. She came precisely to see people. What you experience as an upgrade may be a downgrade to someone else. And what he can’t quit isn’t stocks, it’s research: while everyone else went to the kids’ club, he lay on a deck chair talking industry with friends and checking the tape. One detail is easy to skate past — he says he can be “a bit more graceful” about it now because there’s no leverage in the book, so being slightly offside doesn’t matter. Do the same thing fully levered and it stops being a holiday.
2. A rule change doesn’t turn someone who couldn’t into someone who can. After disposition stocks moved to one match every two minutes, the most visible feature he saw was a flood of fake orders — everyone probing the edges of the new rules. But that isn’t the point for most people. The point is this: every rule change creates a set of beneficiaries and a set of victims, and retail is usually on the victim side, because you don’t suddenly become profitable just because the rules moved. The people who extract money under a new regime are the ones who actually studied the new regime. As for all the intraday deception, he offers a good image: it’s a crowd of giants brawling, and you happened to be standing beside them and caught the wind off a punch — the punch was never aimed at you, you were just there.
3. The collapse of the Zeus rule: when good news becomes a distribution tool. He borrows a concept from the Odyssey to describe the current market — the Zeus rule being, roughly, that everyone holds to a shared understanding and doesn’t turn on each other: good revenue should mean the stock goes up. But recently a name printed revenue well above expectations and didn’t limit up, and everyone began to wonder whether a lot of people had known all along. Once good news becomes an opportunity to distribute, then next time — you know revenue will be good, and you know others will dump into it, so you’d be an idiot not to dump first. Once trust breaks, the game recurses forward, and eventually nobody is willing to step up first. The example from his own circle is sharper: a friend walked him through the fundamentals yesterday and said this one couldn’t lose; asked today, he’d already sold, because “I keep getting played, I can’t take it.” He calls it fundamental day-trading. His judgment: this state repairs itself only once the index makes new highs; until then the tape stays rangebound and awkward, though groups are in fact forming — and looking at it through an ultra-short-term lens is what will trap you.
4. Google didn’t leave the game; it changed where it stands. This is the most fully reasoned stretch of the episode. The market read the DeepMind leadership change plus a group departing to found a startup as a concession of defeat. His reading is the opposite, in steps. Step one, look at the shape of the appointment: the successor carries an SVP title, not CEO, meaning he reports directly into the parent — what used to be a two-track structure between the US headquarters and London has been folded into one. That’s tightening control, not letting go. Step two, put Berkshire’s stake next to it. Buffett said out loud that the position was his own decision, and after the new CEO took over they added substantially and subscribed to a large additional issue. Which raises a question: since when does Berkshire buy call options on frontier models? It doesn’t. What it likes is railroads — heavy assets, oligopoly, regulated returns. Step three, so run it backwards: if Google is pushing resources toward cloud and compute, then what Berkshire bought isn’t the model at all, it’s the railroad and the power grid of the digital era. Step four, look at the structure of the new venture the departing team founded: Google is a founding investor and also its cloud partner. So this isn’t people walking out — it looks more like moving the most cash-hungry, most uncertain slice of research outside the corporate body while keeping the equity stake and the tenant relationship. The risk goes out; the rent-collecting seat stays; and reported cash flow may in fact turn positive sooner. He notes Meta is doing something structurally different but directionally similar, and offers two signals to track from here: upward revisions in TPU procurement, and new procurement indicators on Meta’s side.
5. But the pivot has a price, and he raised it himself. If a company recasts itself as an infrastructure provider, the “it might win” slice of the valuation has no reason to remain, and the multiple shouldn’t be as high — so a de-rating is genuinely possible. Then he immediately supplies a counterexample: the market spent weeks arguing that Apple was the most conservative on AI investment and therefore the loser, and when earnings landed, the money ran toward Apple, while the heavy spenders underperformed. So he leaves the conclusion in an honest place: your argument can win and you can still not make money. He adds a longer-dated variable — the competition from China won’t necessarily respect sanctions, and with copying plus subsidy it can drive prices very low, so last cycle’s script of “infrastructure, then platform, then software, and software is where the money is” may not run again this time.
6. Leverage announces itself through your body. A listener asked about suddenly not wanting to run leverage anymore, and the answer is the most practical stretch of the episode. His own cash equity book carries no margin at all, and although his pledged credit line keeps growing he has never drawn it down fully — not because of a calculation, but because the body tells you first: you stop sleeping well, you start getting dragged around by intraday swings and international headlines. Then the line worth keeping: when you notice that you have become an expert on everything under the sun, fluent in the state of some distant strait, you have probably done something wrong somewhere. His explanation is that noise comes at you because you are anxious — the tape drops and you rush to find a reason, when sometimes the tape simply goes up and down. Meanwhile plenty of large operators understand exactly one industry, put the position on and go drink wine and ski, and end up making money. The difference isn’t knowing more; it’s being able to filter noise out.
7. “Didn’t we already argue about this?” — being locked out of the trade by your own knowledge. Someone asked how anyone could possibly know all thousand-plus listed Taiwanese names. His answer: there is nothing new in markets, themes just cycle — optical, thermal, equipment, substrates, passives, and around again to optical and thermal. But the cycling has a side effect: when a theme you thoroughly researched two years ago gets dragged out again, you think “we already did this” and don’t get on — your knowledge has become your blind spot. His rebuttal is simple: new people keep entering the market, and the financials can keep printing higher. He also mentions researching the bicycle supply chain once — derailleurs, carbon fibre — and never making money on it, but that understanding wasn’t wasted; it stacks onto the next time.
Extending the thought
1. The slice of valuation that gets removed is, structurally, an option
For a company doing both infrastructure and frontier research, the multiple the market assigns is really two things added together: cash flow it is collecting today, and an option on “it might win.” Those two are different in kind — the cash-flow piece can be checked against revenue, margin, and the payback period on capex; the option piece can’t be measured with the same ruler, because its value comes precisely from uncertainty.
So when a company moves high-risk research off its own books, what happens isn’t that the valuation gets lower — the composition of the valuation gets swapped: the option shrinks and cash flow’s weight rises. That matters because it changes the question you should be asking. Not “how much did it fall, is it cheap yet,” but “which piece fell.” If what fell was the option, and the option was your reason for owning it, then cheapness is beside the point: your thesis is gone. Conversely, if what you wanted all along was rent-like cash flow, this decline is the narrative moving toward you.
Why not the other explanation — that this is just short-term sentiment? You distinguish them by checking whether capital allocation actually changed: procurement, capex, investment structure. Those are real money, not talking. That is exactly what makes the two tracking signals from the episode valuable — they can be falsified.
And the kill condition, written down in advance: if a frontier model produces a generational lead, pricing power over compute migrates toward the model layer and the landlord logic breaks — the rent collector becomes the side without negotiating power.
2. How much news you read is a function of your position size
“When you become an international affairs expert, you’ve probably done something wrong” deserves to be unpacked into something operable.
Most people treat information consumption as exogenous: markets are turbulent, so I read more. But the real causality often runs the other way — your position size sets your anxiety level, and your anxiety level sets how much news you consume. The same geopolitical headline gets scrolled past when you’re small and gets three analyses and two videos when you’re fully levered. The volume of news didn’t change. You did.
How do you turn that into a test? “Am I scared?” doesn’t work, because fear is subjective and people tough it out. The more mechanical question is: of everything I read this week, how many items actually changed a decision? If you’ve spent several weeks consuming macro analysis and not once added, trimmed, or moved a stop because of it, you weren’t researching — you were taking a placebo. Its real function is anxiety relief, not judgment formation.
Why not the other explanation? Some people genuinely do read heavily because the research demands it. The same question separates them: real research outputs a decision, or an explicit “no action, waiting for condition X.” Anxious reading outputs only more reading.
One further inference follows: if noise volume is a function of position size, then the most effective way to reduce noise isn’t “building discipline” — it’s cutting the position down to the size your discipline can actually cover. That’s the real content of the contrast in the episode. The large operator isn’t psychologically tougher; what he holds sits inside his radius of understanding, which is why he can put it on and go skiing.
3. In a market that has lost trust, is a longer time frame a solution or an evasion?
The Zeus-rule passage is worth pushing one step further. Once good news becomes an occasion to distribute, the game recurses: I know revenue will be good; I know you know; I know you’ll sell it to me. So nobody steps up first, and good news becomes a selling point. Over short horizons that recursion has no endpoint, because every additional layer holds.
But the recursion has a natural boundary: the earnings actually have to be reported. Revenue can be front-run by three days; it cannot be front-run by three quarters. If a company’s profits are genuinely rising, price eventually has to meet that number. So “it repairs itself once the index makes new highs” reads to me as more than a mood call — there’s a mechanism in it: in a rising-index environment the opportunity cost of distributing goes up, because after you sell you may not get back in, and the incentive to defect falls.
Is this just “hold long enough and you’re fine” in disguise? No, and there are two explicit failure conditions. First, if the index keeps grinding sideways in a range, this repair never arrives, and stretching your horizon only spares you a few round-trip slaps; it doesn’t make you money. Second and more important: stretching your horizon presupposes that the earnings will actually materialise. If what you’re holding is a narrative rather than a number, time is working against you, not for you.
Which is why the operational content of this section is really another detail from the episode: he says the last drawdown left a mark, so he now deliberately picks names that can put up results in 2026 and 2027; the ones that only deliver in 2028 or 2029 will rise too, but he doesn’t want them. Time frame isn’t a question of attitude. It’s a question of whether you can survive until the answers are checked.
Worth reading
- Gooaye 股癌 — the show itself, on Apple Podcasts, Spotify and elsewhere. For the full context, please support the original
- Zhuangzi, “Free and Easy Wandering” — source of the opening quote. Full text and commentary at the Chinese Text Project (ctext.org); public domain
- Taiwan Stock Exchange, disposition securities section — the rules, the current list and the matching mechanics, first-hand and free
- SEC EDGAR and company investor relations pages — quarterlies, annual reports and earnings call transcripts for Alphabet and Meta. To verify whether capex and cloud revenue actually changed, primary filings beat news coverage
- Taiwan’s Market Observation Post System — monthly revenue and financials for listed companies, for checking whether the numbers behind a rotating theme are keeping up
- Homer, the Odyssey — origin of the “Zeus rule” reference in the episode; public domain translations are freely available
- Aswath Damodaran’s data page (NYU Stern) — free public datasets on industry multiples and cost of capital, useful for calibrating what “expensive” means
Disclaimer: These are personal listening notes and study notes, offered as educational content. They do not constitute investment advice, an offer, or a solicitation. Companies and industries mentioned appear only to convey the context of the episode’s discussion; no specific security is recommended and no price target is given. Investing carries risk, past performance does not indicate future results. Judge independently according to your own financial situation and risk tolerance, and consult a qualified professional where appropriate. The author may hold positions in the types of assets discussed.
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.