Who's Really Behind the 'Steady Index, Bleeding Stocks' Market? Notes on Gooaye EP680 — A 15% Drawdown Playbook, TSMC Mispriced, and a Misread AI 'Threat'

Listen not to the rain thrashing through the grove —
why not walk on, humming, taking my time?
A bamboo cane and straw sandals beat a horse. Who’s afraid?
One straw cloak, and I’ll live out my life through the misty rain. —— Su Shi, “Calming the Waves” (Northern Song, 1082); translation mine

---

The 30-second version: this bout of “the index is holding up but individual stocks are getting destroyed” has a cause most people are getting wrong. The host of Gooaye, one of Taiwan’s sharpest markets podcasts, argues the culprit isn’t any single sector going bad. It’s leverage unwinding. TSMC’s earnings were clean and the stock got sold anyway. And the “AI threat” the market keeps waving around is an old script we’ve seen before. Here are the full notes, with our own read at the end.

About the show: Gooaye

Gooaye is one of the highest signal-to-noise finance podcasts in Taiwan. Its host — known as “Chairman” — is known for blunt, contrarian, risk-first market commentary, with recurring mantras like “don’t fall in love with a stock” and “figure out how you lose before you figure out how you win.” What follows are our notes and extended read after listening to the latest episode; the link is at the end, and we strongly recommend listening to the original.

The big picture: the host’s three reads

Read 1: Down 15%? Flip to defense, by rule

The host was explicit about his own discipline this episode: once his portfolio draws down more than 15% from its high, a preset rule switches him into “defense mode” — stop chasing high-beta single names, and trade the broad index instead via TAIEX futures and 0050 (the Taiwan 50 ETF).

This isn’t a bearish call. It’s risk reduction: protect capital and dampen volatility first, then go back to hunting single stocks once the market structure steadies. It’s the same thing we keep saying, that the discipline of stop-losses and position sizing decides survival more than your read on the market does.

Read 2: The real culprit behind “steady index, bleeding stocks” — a Korean leveraged-ETF blow-up

This was the most valuable stretch of the episode. The host’s read: this wave of carnage in small caps and high-beta names across Taiwan and US markets traces back to a blow-up in large, highly leveraged Korean ETFs, whose spillover has systematically compressed valuation multiples across global risk assets — rather than any single sector printing bad fundamentals.

The evidence is in the divergence:

• The indices held up: the S&P 500 and Taiwan’s TAIEX stayed relatively elevated, propped up by mega-caps and, in Taiwan’s case, the TSMC weighting.
• Single stocks got crushed: small caps and high-beta themes (optical interconnect, power devices, the ASIC/GPU chain) fell far more than the indices.

If this were a single-sector story, you wouldn’t get “index fine, stocks in flames” across the board. When what gets killed is the valuation multiple rather than the earnings estimate, the driver is usually flows, not fundamentals. If you take one thing away from this episode, make it that.

Read 3: The Kimi K3 “AI threat” is a misread — Jevons Paradox, again

The market has been passing around a few scare stories lately: China’s low-cost AI model Kimi K3 (Moonshot AI), mature-node order grabbing, and so on — all recruited as reasons that “AI compute demand is about to shrink.”

The host’s read is blunt: this is the same misjudgment the market made with DeepSeek. At bottom it’s Jevons Paradox —

When a resource is used more efficiently, total consumption often rises rather than falls. In the 19th century, burning coal more efficiently didn’t cut coal use — it exploded, because cheapness opened up whole new applications.

Applied to AI: cheaper, more efficient models don’t shrink compute demand; they broaden AI deployment and enlarge total demand. Low-cost models still carry enormous parameter counts and don’t dent demand for AI servers. Reading “more efficient” straight off as “bearish” is a first-order-thinking trap.

The through-line case: TSMC, mispriced

The host used TSMC as the episode’s main case for the “flows vs. fundamentals” distinction:

• The call was broadly bullish: Q2 gross margin of 67.7% landed in line; CAPEX was raised from the prior 52–56B to 60–64B — a genuine upside surprise, signaling demand strong enough to warrant heavier investment.
• Advanced packaging is capacity-constrained: the call even hinted at welcoming competitors like Intel’s EMIB-T to absorb some back-end packaging pressure, so TSMC’s front-end capacity can ramp. That’s a supply-can’t-keep-up signal, not bad news.
• Yet the stock got hammered: a report that clean, a stock down anyway. The host reads it as a sentiment-driven reaction, not deteriorating fundamentals.

TSMC, in other words, is Read 2 in miniature. The earnings are fine; the multiple is what got killed.

Sector pulse: price-hike signals everywhere, stocks ignoring them

The episode’s sector notes echo the same theme — fundamentals up, prices down:

• Memory: Micron and the Korean giants are running “low guide, high beat”; the group’s shares got slammed, but the industry outlook is still strong.
• Passives, substrates, glass-fiber cloth: a steady stream of price-hike signals — demand is visibly tight.
• ADI (Analog Devices): also issued a price-hike signal in July, showing analog/industrial demand remains firm.

These price-hike signals are hard evidence that demand is still there, yet the stocks broadly don’t reflect it — for the reason above: valuation multiples got systematically compressed worldwide. So the tape reads as “steady index, bleeding stocks,” and whoever mistakes a flows-driven overshoot for deteriorating fundamentals is the one most likely to get shaken out.

Stock-by-stock recap

The following is a recap of the host’s personal views as expressed on the show; the notes in brackets are our one-line additions. None of it is a buy/sell recommendation or price target from us. Do your own work.

Taiwan

• TSMC — bullish. Q2 results and the CAPEX raise both beat, the call had no weak spots, and the sell-off reads as a non-rational reaction rather than weakening fundamentals. [The core demonstration of “flows vs. fundamentals” for the whole episode.]
• FEC (Flytech) — bullish (a new name this episode). A POS/industrial-PC maker with record margins and software revenue up several-fold, whose stock hasn’t reflected the fundamentals; the host suggests long-term investors could scale in gently on weakness and be patient. [A textbook “earnings stock” — it speaks with actual profit, not narrative.]
• MediaTek (MTK) — bullish. Citing a friend’s view, the host thinks the stock is undervalued and deserves a higher multiple.
• The “pun-stock all-in” cautionary tale — a listener went all-in on a stock chosen for a pun/homophone theme; the host used it as a warning against over-betting on a gimmick. [The point isn’t to mock anyone — it’s that staking your net worth on a meme is gambling, not investing.]

US

• Micron — bullish. Same pattern as TSMC (low guide, high beat); the memory group’s shares got hit hard but the industry outlook is still strong — a sentiment-driven overshoot.
• ADI (Analog Devices) — bullish. A July price-hike signal, demand still firm, though the stock hasn’t reflected it amid a weaker mood.
• Intel — neutral. If EMIB-T advanced packaging succeeds, it could absorb some of TSMC’s back-end packaging load — industry-complementary rather than purely competitive.
• Tesla — neutral-to-bullish. A long-term holding the host likes; he touched on Musk’s AI5/AI6 chip cost considerations and their impact on the supply chain.
• SpaceX — bullish. The host keeps adding, and rates its culture and execution speed highly — a long-term conviction holding.
• Apple — neutral (as a case study). It fell earlier on memory-related news, then went on to make new highs — used to illustrate that “a short-term negative isn’t the long-term trend.”
• Credo, Astera Labs (ALAB) — case studies. Both are “crushed early in the year, then rebounded hard” high-beta names, used to show that fundamentals didn’t change; it was the flip-flopping market narrative that drove the big swings. [High beta falls hardest in a risk-off — but falling hard ≠ broken fundamentals.]

China

• Kimi K3 (Moonshot AI) / DeepSeek — neutral, rebutting the “AI threat” thesis. The market misreads low-cost models as a compute negative; the host argues the parameter counts are still enormous and don’t dent AI-server demand — the same “efficiency = bearish” misread as DeepSeek.

Our own take (the Realpha layer)

After listening, we’d add three framework-level reads — not stock calls, but a lens for how to think:

1. First, learn to tell “multiple compression” from “earnings compression.” This is the most actionable lesson of the episode. When a stock’s results improve, price-hike signals keep coming, and demand is visibly stronger — yet the shares get dragged down with the market’s multiple — that’s usually a flows-driven mispricing. Conversely, if guidance is cut, customers pull orders, and inventories balloon, that’s fundamentals turning bad. The first is opportunity, the second is a trap — but on the tape, both “look the same shade of red.” Fail to tell them apart and you’ll bail during the mispricing and hold through the real deterioration.

2. “Earnings stocks vs. story stocks” behave worlds apart in a risk-off. The host clearly leaned toward earnings-backed names this episode. That squares with something we’ve validated ourselves: when valuation multiples get systematically compressed, stocks with real earnings and a valuation anchor hold up far better than pure-narrative names. In our own factor testing, the “valuation anchor” (cheap-multiple) signal is one of the few that survives scrutiny — but only at the cheap end; chase it up to expensive levels and it hurts just the same. In other words: earnings stocks aren’t a cure-all — cheap earnings stocks are.

3. Treat “publicly shorting a single name” with respect. The host was noticeably more careful this episode about commenting on individual stocks, especially bearish views. That’s not timidity; it’s risk management against being used as exit liquidity or drawing retaliatory attacks. We agree, and it’s one reason we sell only a judgment framework to the public and never call individual buys and sells.

What we can take from it: three portable lessons

1. Killed multiple ≠ killed fundamentals. A flows-driven overshoot and a fundamentals breakdown look identical on the tape, but one is an opportunity and the other a trap. Learning to see “is it the earnings or the multiple getting killed?” is the most valuable takeaway here.

2. Efficiency gains aren’t demand destruction (Jevons Paradox). Every time a “cheaper technology” arrives, the market reflexively reads it as bearish. History keeps showing the opposite: getting cheaper usually grows the pie rather than shrinking it. A headline is just the starting point; the analysis is two or three hops downstream.

3. Discipline is written first, executed later. “Down 15%, flip to defense” works precisely because it was set while calm and followed while panicking. Figure out how you lose before how you win — hard-code the rule, and emotion can’t drag you off it.

Go listen to the original

However detailed our notes, they can’t compress the host’s live delivery, pacing, and off-the-cuff analogies. We strongly recommend listening to the full episode:

• Gooaye (available on Apple Podcasts / Spotify / Firstory)
• Search for the show name “股癌 Gooaye” to find the latest episode

---

This article is a listening note and educational summary of a public podcast. The content reflects the host’s personal opinions, does not represent the views of this site, and does not constitute any buy/sell recommendation, price target, or investment solicitation for any individual security. All references to specific stocks and their moves are as stated by the show’s host and guests. Investing carries risk — leveraged instruments such as index futures carry especially high risk — and any decision should be made on your own judgment or in consultation with a qualified professional.

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.