In short: The earnings-bubble exhibit: "having memory companies having operating margins of 80% is not normal and trust me I firmly believe that once China ramps, gets to their ambitions in the memory markets, those margins will go back to below the average." Also the stock Korean retail bought on leverage — "your Korean listeners that got wiped out because they were on leverage or buying levered ETFs on Samsung and Hynix."
SK Hynix is a Korean maker of memory chips, including the high-speed memory that sits next to AI processors. Demand from AI has let memory makers earn operating margins around 80% — meaning about 80 cents of every sales dollar is profit before interest and tax. Niles says that is "not normal," and calls it an earnings bubble: the stocks don't look expensive only because today's profits are unusually high.
Memory is a commodity business that has always gone through boom and bust. China has made memory a national-security priority, its two big memory makers (CXMT and YMTC) are raising money in public markets, and it has a huge home market to sell into. Once that supply arrives, he "firmly believe[s]" memory margins will fall back "below the average."
It is also the stock at the center of a painful lesson: Korean individual investors who bought it and Samsung with borrowed money and leveraged funds were wiped out in the June–July sell-off.
34:50But you've got some very high valuation levels if you look at it relative to GDP or normalized earnings because these companies are all over-earning to some degree. We don't have a multiple bubble to some degree but we do have an earnings bubble, right? Having memory companies having operating margins of 80% is not normal and trust me I firmly believe that once China ramps, gets to their ambitions in the memory markets, those margins will go back to below the average and so you do have a lot of risks in the market
In short: Referenced twice, both times as a casualty. It is the stock whose implosion "absolutely destroyed" leveraged Korean retail investors, and it is the incumbent YMTC has publicly targeted: "they plan to be bigger than Samsung or SK Hynix in the NAND business by the end of next year." Memory valuations, he notes, only look cheap "if… this is no longer cyclical."
SK Hynix is one of the two Korean memory giants, and it appears in this conversation twice, both times on the receiving end.
First as a casualty of leverage: when SK Hynix and Samsung "imploded," Korean retail investors who had bought on margin and then bought leveraged funds on top of that margin "got absolutely destroyed." That is the concrete story behind his closing rule against using borrowed money.
Second as the incumbent in China's crosshairs — YMTC has publicly targeted passing it in NAND within roughly a year. His broader point on memory valuations is that they only look cheap on the assumption the industry is no longer cyclical, and a state-funded new entrant is precisely what makes an industry cyclical again.
7:07And so you saw that with Korean retail investors who got absolutely destroyed when SK Hynix and Samsung imploded, but you also saw that with funds such as situational awareness. And there were some other funds that I heard that were in trouble as well. But luckily situational awareness being pushed into that for sale with Citadel helped solve some of those issues because the market obviously ripped the next day, especially in a lot of those positions that were getting unwound.
In short: Referenced — leads in HBM, just filed for a ~$29B US listing, and is adding supply (some read early 2H26 price softening as a result). One of the three-player oligopoly.
Nothing matches this filter.
Verbatim excerpts from the public transcripts (auto-pulled at each mention's timestamp, lightly cleaned). Timestamps link into the video; "source page" opens that commentator's full analysis at this row. Click a mention's header line to expand it (one open at a time). For personal study — not investment advice.