# Gooaye EP715: Take Away the Buyer in a Hurry and the Curve Dents Source: Realpha Blog (blog.getrealpha.com) Original article and charts: https://blog.getrealpha.com/en/blog/gooaye-2026-11-18-ep715/ > Notes on Gooaye EP715 (18 November 2026): the host's US account back below 20% for the year, his read that this Nvidia quarter still belongs to the departed CEO, a research house cutting its own estimate by $1.6 trillion, and the one check he leaves listeners — what happens to this company if the rush orders disappear. Educational listening notes only — no investment advice, no price targets, and the judgements here may be wrong. Published: 2026-11-19 Locale: en Tags: Gooaye, podcast notes, AI supply chain, rush orders, investing discipline TL;DR: The host's point is that urgency is receding: with one aggressive buyer gone, the scramble for parts and the price hikes ease off and orders slide back, while he still expects the three-to-five-year volume to hold. The check he leaves listeners is whether a company takes a financial hit once the rush orders disappear. ![A container yard at night, a long line of trucks standing still, an empty lit crane cab under the floodlights, containers still stacked high, the quay running deep into the distance](/covers/gooaye-2026-11-18-ep715-cover.png) > The wild goose advances to the rock; it eats and drinks at ease. Auspicious. > > —— Book of Changes, Hexagram Jian, second line (pre-Qin; my own translation) In Gooaye EP715, released 18 November 2026, host Hsieh Meng-kung says his US account is back below 20% for the year, and that this time he cut the positions whose chart patterns had broken instead of holding and waiting for support the way he did in July. He relays that Jensen Huang has publicly said he is willing to come back and run Nvidia, that the board answered that its decision is made, and he frames this quarter as still belonging to Huang — a new manager's decisions take several quarters to show up in the numbers, and the share price moves first. His scenario is that urgency is receding: with one aggressive buyer gone, the scramble for parts and the price hikes recede and the rising curve dents, while he expects the three-to-five-year volume to hold. The scenario rests on the board not changing its position, and he says plainly that he cannot read the current market. ![A gray dashed straight line shows total volume rising steadily over the long term, while a solid line dips into a trough midway before returning to the same endpoint height.](/figures/total-volume-vs-timing-dip-en.svg) ## I was only tracking total demand, never who was pushing it My own arithmetic went like this: AI demand holds up, data centres take years to build, so the sharp drops in between are noise and the thing to do is hold. That frame has talked me out of scaring myself out of good positions more than once. The frame has exactly one variable — the total. If the total hadn't changed, I treated nothing as having changed. The line above is about a wild goose advancing in stages, pausing on a rock, eating calmly, and that being the auspicious reading. I used to take it as "there's no hurry." I skipped the word doing the work: gradual. Who sets that bird's pace turns out to be a harder question than where it is flying. This episode pulled my frame apart. The host admits July already taught him a lesson: you assume there is support at the monthly moving average, then at the quarterly one, and there isn't — it just keeps going down. Back then he held on, because he was far enough ahead not to care. This time he cut the downside strictly, trimming whatever had broken its pattern or printed a top, and then he says that even with those cuts, these past two days have been deeply uncomfortable. That is where I stopped. His position control is tighter than mine and the outcome still hurt, so my "the total hasn't changed, therefore hold" idea was missing something other than discipline. ## The buyer in a hurry left, and the timing moved His read on the earnings call supplied the missing piece. He had expected Nvidia to guide down, not out of pessimism but because the company can see how high the sell side has gone and would want to pull the number back itself. What he got instead was no cut, plus an enlarged buyback. His comment on that carries the most information in the episode: if you genuinely believe money poured into AI will earn its return, you keep pouring. When a company answers with a buyback, it is telling the market it now rates buying its own shares above expanding outward — and that is a foot moving toward the brake. ![A cash block on the left splits into two thick arrows: the upper arrow leads to expansion, while the lower arrow leads to buybacks and is labeled as hitting the brakes.](/figures/two-exits-for-the-cash-en.svg) Then he opens up the human side. The incoming manager is not the founder. The upbeat lines about preserving the company's lead will still get said, but the aggressive route — financing customers freely, flooring the accelerator on the supply chain's iteration cadence — takes a particular kind of nerve. What a parachuted-in or finance-trained manager can do is cut cost, restructure, and make the statements look better; that is the supply-chain-management route. Even if the new person wants to pick up the predecessor's playbook, he says, it takes time. So his conclusion: stretch it to three or five years and the volume may not change much, but the feeling of "I have to secure parts this month" goes away. The scramble for memory, for optical components, the cornering of other people's capacity — that eases once the aggressive player is gone. His word for it is push out. The orders are still there; the timing moves back. He adds a note about lag that I found the most usable line in the hour. Plenty of headlines called this "the first earnings report after Jensen Huang left," and he says that is wrong — the next several quarters of numbers still sit inside the predecessor's range. His example is Lip-Bu Tan taking over Intel: the early results that appear under a new chief executive are the tree the previous one planted, breaking ground on the new one's watch. Fundamentals take quarters to show a new manager; the share price will have finished moving long before that. ![Two staggered color bands span a quarterly timeline, with the predecessor's band covering the first few quarters and overlapping the successor's band, while a dashed stock-price marker sits at the far left.](/figures/whose-numbers-are-these-en.svg) A research house cut its own figure in the same window — a week or two after its previous version, he says, taking $1.6 trillion out, something over 30%. He hedged that recollection himself. His read: even after a cut that size, the number still sits at the high end of the market's range, so this is no end-of-the-world scenario, just a figure that had been pitched too high. He thinks the cut to the total is survivable; what the news really raises is the probability of push out. ![Two bars stand side by side, with the right bar more than 30% shorter than the left; the removed section is labeled as a 1.6 trillion cut, and a lower dashed block on the right represents other estimates.](/figures/cut-but-still-the-high-bar-en.svg) Which leaves the homework he hands listeners as a question: if the rush orders disappear, does this company take a financial hit? If yes, be careful. If the rush orders disappear while long-run demand holds, that is the part of AI he considers relatively touchable. ## I ask about rush-order dependence first, and admit what I can't see You may have spent this week in the same confusion I did: the companies I hold haven't deteriorated, the guidance wasn't cut, and the shares fell hard anyway. What this episode gave me is to split "demand" into two layers, volume and timing. Volume underwrites what a company is worth in three years; timing decides what the next four quarters of orders look like. In a sell-off what the market reprices is timing, and the share price moves on timing first. The question I now rank holdings on is this: how much of this company's revenue over the past two years was padded by other people rushing to lock up supply. ![Each company's revenue bar has two layers: baseline demand below and rush orders above; the upper layer is much thicker for the company on the right, creating a much larger drop when it disappears.](/figures/how-much-was-rushed-en.svg) He also left a conditional on the day's tape that I wrote down. The index was sold hard in the morning and recovered, and his read is that this market has lost its confidence. Going long where confidence is gone strikes him as worth doing, with the exit written down in advance: when it reaches the big moving average, you sell. He is not calling a reversal, and he does not expect the old highs back soon. ![A price line plunges to a low and rebounds in a V shape, but only reaches below the major moving average; the prior high is marked by a higher dashed line, and both reference lines remain above the price.](/figures/exit-line-drawn-first-en.svg) The second question you may be sitting with: I'm underwater, do I cut or hold? He doesn't answer it, but he gives two things. One is his trigger for admitting it — a broken pattern, a chart that can't recover, not a percentage drawdown. The other is the mirror of averaging down: people who lose and decide to win it back with 200% or 250% exposure in one shot. His verdict is that doing so puts you at some position on the slaughter line, and your turn comes eventually. That one left me cold, because the case for averaging down always writes up beautifully, and a beautiful case is usually a reluctance to take the loss. The third item runs against intuition and I copied it out. He normally pays little attention to broker-flow data, finding it flaky, and in a tape like this one he watches it closely — because he wants to know who is about to be margin-called. Where the biggest margin buyers got in, which branch they trade through, when they get forced out, how many shares are left, when the last of them goes. In a strong bull market, he says, get the theme right and it doesn't matter at all who is buying. The same tool is worth that different an amount depending on the tape, which had not occurred to me. Two things I'm not confident about. First, the whole scenario hangs on the board's position, which he treats as a variable he'll keep tracking rather than a settled fact. Second, he is blunt about the limit of his own read: he called the satellite group well and it has been one of the relatively strong areas, and he asks what good that did — everyone went down together, you just went down less. What's left, he says, is finding a way to survive in the cracks. When someone who has worked the scenario out this finely says he can't read the market, I'm not going to pretend I can. There's one stretch with nothing to do with the tape that I liked most. He went to an orchestral concert of the soundtrack to an independent game, worried on the way in that a hall full of men listening to game music would feel awkward, and came out moved enough to play the game again that night. His reason for telling the story connects to everything before it: when you know you can't change what's in front of you, find things that make you happy and keep your own state somewhere high, because a lot of opportunities show up inside the despair and the problem is living through the trough. ## Worth a look - Gooaye EP715 (18 November 2026), the source for these notes - Nvidia's quarterly filings and earnings-call transcripts on its investor relations site, to check the description of the quarter and the buyback - The handover dates of Intel's recent chief executives and the quarters that followed, to check the "whose results are these" lag yourself - Margin balance and short-to-margin statistics from the Taiwan Stock Exchange and the Taipei Exchange, for the leverage picture - Earnings presentations from memory and optical-component suppliers, to track whether the rush orders and price hikes actually recede ## The one thing to take with you **How fast something happens depends on who is pushing it from behind.** The step worth paying for in this episode is that one: the size of demand and the urgency of demand are two variables, and most of us only did the arithmetic on the first. Once the person shouting "I need it next month" walks out, the thing still gets built, the date slides, and every link that lives on hurry feels it first. Here's something I've tried that has nothing to do with investing. Pick one thing you assume will just happen — the renovation at home you've discussed for a year, the project your team has penciled in for next year, the follow-up appointment an older relative promised to make. Write down who is pushing it right now, and then write one line: if that person stepped out today, how long would it slip? The one where you can name a number of months, you know what's carrying it. The one where you can't, you only assumed it was moving.