Gooaye EP680: Make Yourself Unbeatable First — Why Good Numbers Could Not Save the Stock

In ancient times, those skilled in war first made themselves unbeatable,
and then waited for the enemy to become beatable.
Being unbeatable rests with yourself; being beatable rests with the enemy.
So the skilled warrior can make himself unbeatable,
but cannot make the enemy certain to be beatable.
—— Sun Tzu, The Art of War, “Dispositions”

These are my personal notes on Gooaye EP680 (released 2026-07-18). This is not a transcript and not official content. Please support the original show for the full episode.

What the episode is about

In one line: this episode is about what you do when every card in your hand is good and the other side still will not call.

The first half has nothing to do with the market. The host talks about visiting the SpaceX factory back in 2024, and a piece of merchandise a friend brought back for him — a pair of chopsticks made from the melted-down first-generation Mechazilla catch arms, sold in limited numbers to employees, apparently one pair per person. It came with a small bowl shaped from a rocket receiver. His own joke: given the tape, having both a bowl and chopsticks means one fewer prop to buy before he goes begging by the roadside.

The market half comes after: TSMC reported a quarter with almost nothing to criticise, and fell 6% the next day. More than half his book was in that one name, so the day tripped the defensive rule he had set for himself.

The heartbreak emoji in the title works on two levels. One is the tape. The other is a line near the end, when he talks about his kids — I will never again see the one-year-old version of my child.

The main points

1. A quarter with no visible weakness, followed by a 6% drop. Earnings per share came in at 27.25 and gross margin at 67.7%, both near consensus. The genuine surprise was capital expenditure, revised up from 52–56 billion dollars to 60–64 billion, with full-year revenue growth guidance lifted from 30% to 40%. The host’s point is that almost nobody had positioned for a capex raise, and it should read as good news — yet the next day the equipment suppliers who would collect that money ran toward limit-down. His read: the market is currently in a mode where it reaches for the bearish interpretation first.

2. The company publicly said it wants a competitor to succeed. Asked about advanced packaging on the call, management described the back end as extremely tight. The inference worth keeping: if the back end is the constraint, the front end cannot run at full volume. So if a rival — Intel’s EMIB-T, for instance — actually makes its packaging work, that is good for TSMC. When a company says out loud that it hopes someone else fixes a particular layer, it has effectively confirmed where the bottleneck in the chain is.

3. Defence is a number written in advance, not a mood. His trigger: a 15% drawdown from his own equity high, combined with the index breaking its quarterly moving average. Both conditions landed on the same Friday, and he switched that day. What matters is the reason he gave — not “I think it will fall further.” The slope of that moving average was still positive, which if anything argues for a bounce. He switched because the condition was met.

4. But hiding in the safest name gets you hit too. Through this move he had been pulling money out of small and mid caps and stacking it into TSMC, which genuinely smoothed the damage for a while — the big name even rose for a few days first. Then it dropped 6% in a session. That is a lesson in itself: when everyone retreats to the same shelter, the shelter stops being one and becomes the next concentration risk. He had warned in the previous episode that the newly strong groups deserved caution, and those groups were duly sold off — the same mechanism wearing a different jersey.

5. Defence does not mean cash. His plan is to stop picking individual names and use index futures or the local large-cap ETF instead — owning the whole market. His reasoning is blunt: he does not believe in cash. He has watched the index go from ten thousand to over forty thousand; the more cash you held, the more you lost. Holding cash looks clever in the short run, but if you fail to switch back at the right moment you end up the biggest loser — you avoided a bit of drawdown and missed the entire advance. So he cut leverage and changed instruments rather than leaving.

6. What changed is the multiple, not the industry. Memory, passive components and optical names sold off hard with no matching bad news from the industry; July even brought a wave of price increase notices across substrates and glass fabric. So why did rising prices fail to lift the stocks? His explanation is that heavily leveraged money in Korea blew up first and punctured the bubble, and global sentiment moved into a phase of re-rating. His rough illustration: if the average multiple at the highs was, say, 25 times, the reasonable expectation now might be 18 to 20, with an overshoot to 12 or 13 along the way — but getting back to 25 requires a long convalescence. That is also why he is rotating from story stocks toward earnings stocks: if the multiple cannot expand, only earnings revisions can carry you.

7. Elimination reasoning on the ghost stories. The latest bearish narrative concerns the Chinese model Kimi K3, and it rhymes with the DeepSeek episode: a cheaper, more efficient solution appears, therefore the spending is unnecessary. He thinks it is another misread, on two grounds. K3 itself runs at 2.8 trillion parameters and still needs high-specification servers. And Jevons Paradox applies — when something gets cheaper and uses less per call, more people adopt it and more money goes in, so total consumption rises. He applies the same method elsewhere: people had blamed the Middle East conflict for the drawdown, but oil never spiked, so that explanation fails.

8. The cognitive dissonance of long-term investing. A listener asked about an industrial PC maker in point-of-sale terminals — record margins, software revenue up several times over, stock going nowhere but down. Should he cut? The answer is the best passage in the episode: “The price not matching your expectation is precisely why you are here doing long-term investing.” If you think something is worth 500 and it falls to 100, you buy it; if it goes to 80, you should be pleased. The pain comes from mixing long-term investing with momentum speculation — you select on fundamentals and then grade yourself on price, so of course it hurts. He immediately adds the guardrail: patience only pays if you were right about the object of that patience, otherwise you have simply committed yourself to the wrong partner. Hence: start with a small position.

Where this leads

Defence should be a trigger, not a feeling

The value of “15% drawdown plus a break of the quarterly moving average” is not in those two numbers. It is that they were written before the pain.

The rule converts a question you cannot answer at the worst possible moment — is this the bottom? — into a fact you can look up at any time: how far am I from my high, and is the index above that line? Nobody can answer the first. The second always has an answer. The moment you most need judgement is the moment your judgement is worst, and a pre-written trigger simply moves the decision back to when you were thinking clearly.

The episode also supplies an uncomfortable footnote: where you move the money after the trigger fires is itself a risk decision. He stacked into the safest name, it worked, and then that name hit him. He was not wrong about the name; the problem is that “everyone is hiding here” is itself an accumulation of risk. Crowding in a shelter is inversely related to how sheltering it is. So the rule needs a second layer — not only when to reduce, but where the proceeds go so that you have not simply swapped one concentration for another.

His answer was to own the whole index, which is a direct response to that second layer: when you cannot tell which name leads the rebound, stop betting on which name leads the rebound. You give up the excess return from selection and get, in exchange, the certainty of being present whoever leads. That is an honest trade, not a clever trick.

Separate “the industry broke” from “the multiple compressed”

The most portable framework in this episode is splitting a falling price into two very different causes.

In the first, demand is genuinely gone: orders vanish, customers move, the product is displaced. In the second, demand is intact — prices are even rising — but the multiple the market will pay has shrunk. These look identical from the outside; both are just a stock going down. The responses are opposite. The first calls for selling, because the premise of your thesis has failed. The second calls for waiting, because the thesis is right and only the price is wrong.

The episode supplies a fairly concrete test. Price increase notices still circulating, capex guidance still being raised, industry visibility still intact — then the demand side has not broken. What broke is sentiment and leverage. In that case the only real move is to accept that the multiple will not snap back and shift your attention to whether earnings can keep being revised up.

When should that judgement be overturned? My own falsification conditions: price increases start being withdrawn, capex plans start being cut, or customers start cancelling orders. Any one of those and this stops being a re-rating and becomes an industry turn, at which point the whole “just wait” logic has to be discarded. Writing the failure conditions down first is what stops waiting from quietly becoming denial — because those two also look identical while you are inside them.

Elimination tells you what it is not, never what it is

I like the logic used to rule out the Middle East explanation: if that were the cause, oil should have reacted; oil did not react, so it is not the cause. That is clean falsification — first ask what you should observe if the explanation holds, then go look.

The same method works on Kimi K3. If more efficient models reduced compute demand, you would expect parameter counts and deployment requirements to shrink alongside; K3 is a 2.8 trillion parameter model that still needs high-end servers. Add the historical pattern behind Jevons Paradox — falling unit cost usually expands rather than contracts total consumption — and the causal chain of the ghost story breaks at the first link.

But the boundary of the method deserves to be stated honestly: elimination removes wrong explanations; it does not produce the right one. After all the eliminating, what remains in the episode is “probably a liquidity-driven massacre” — and “liquidity” is frequently the respectable name we give to not knowing. That is not a flaw, it is candour: he says as much himself, that there may be things going on that we do not know about, and that he still respects what the tape is doing.

That is the part worth taking. Clear out the demonstrably wrong narratives with falsifiable tests, admit you do not know what remains, then manage that ignorance with position size rather than with prediction. Which is exactly what Sun Tzu meant: being unbeatable rests with you, being beatable rests with the enemy. You can keep yourself from being knocked down; when your turn to win arrives is not yours to schedule.

One honest boundary

The closing stretch of the episode is about parenting and has nothing to do with markets, and it is the part that stayed with me. He says we are constantly saying goodbye to the earlier versions of our children — the same child, and yet no longer that child. He nearly gets choked up telling it.

I put it at the end here because it is the same point as the market half: what matters is the experience of the process, not loading directly into an outcome. You cannot skip this drawdown and arrive at the next high, in the same way that you cannot skip the exhausting years and arrive at a grown child.

The above is my synthesis, not advice to anyone. The figures cited — earnings data, drawdown percentages, multiple ranges — are self-reported or offered as rough illustrations in the episode, used here to explain reasoning rather than for anyone to copy. I am in this drawdown myself.

References

• Gooaye EP680 (released 2026-07-18) — the source of inspiration for this piece. Please listen to the original episode and support the creator.
• Sun Tzu, The Art of War, “Dispositions” — origin of “first make yourself unbeatable, then wait for the enemy to become beatable”
• Jevons Paradox — from William Stanley Jevons, The Coal Question (1865): efficiency gains tend to expand, not reduce, total resource consumption

---

Disclaimer: This article contains personal listening notes and study reflections for educational purposes. It does not constitute investment advice, an offer, or a solicitation. No specific securities are recommended and no price targets are given. Investing involves risk; past performance does not indicate future results. Please make independent decisions based on your own financial circumstances and risk tolerance, and consult a qualified professional where appropriate. The author may hold positions in the types of assets discussed.

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.