7月瑞士宝胜对话中英文翻译版20260720

Recorded 2026-07| English transcript with Chinese translation Opening /开场 Richard Tang Hello everyone, welcome to the Julius Baer podcast. This is Richard Tang, the China strategist and head of research HongKong for Bank Julius Baer. It's great to have Hong Hao with us again to discuss the Outlook for the Chinese market. Hao is arenowned economist and he's also the managing partner and CIO of Lotus Asset Management. 大家好，欢迎收听瑞士宝盛的播客节目。我是Richard Tang，瑞士宝盛中国策略师兼香港研究主管。很高兴再次邀请洪灏参加节目，讨论中国市场前景。洪灏是一位知名经济学家，同时担任Lotus Asset Management管理合伙人兼首席投资官。洪灏，你好，非常感谢你今天接受我们的访谈。 Hong Hao Hey Richard, great to be here.你好，Richard。很高兴再次参加节目。 Q1. How should investors view the correction in AI infrastructure stocks? /如何看待AI基础设施股票的回调？ Richard Tang Hi Hao, thank you very much for the time speaking with us today. So how during the past few weeks, we've seen a verysignificant pullback in all these AI infrastructure stocks, basically around global markets. We not only talk about the us, butalso different markets in Asia, Korea, Taiwan, Japan, for example. And in China in particular, many popular AI names havefallen, say, 20-30% or even more from their highs. The driver of this is, of course, first there's this elevated volatility in theKorean memory sector. And then there were also recently some marginal shifts in the AI narrative as well. We alreadydiscussed in the last episode about the crowding risk in AI stocks, and I'm sure the audience is not entirely surprised by thecorrection that's going on right now. But I think what's more interesting is that we've seen, and we've got so many questionsabout whether we should buy the dip now or should we wait later. And at Julius bar, we still see the structural uptrend in AItoplay out over multiple years. In the near term, the nearest potential catalyst will be when the us hyers scalers share an updateon their capex guidance by the end of this month, when they're due to report their Q2 results. Now if these numbers arereaffirmed or even raised. I think this should help market to resume its primary uptrend after the current correction. How Ithink last time we spoke you changed to lean towards a more conservative stance for AI. What about now? After the recentcorrection, are we close to the primary uptrend again or you think we should wait a little longer? 过去几周，全球AI基础设施股票出现了非常明显的回调，不仅美国市场如此，亚洲的韩国、中国台湾和日本市场也受到影响。中国许多热门AI股票从高点下跌了20%至30%，甚至更多。原因包括韩国存储器行业波动加剧，以及市场对AI故事的边际变化。上一期节目我们讨论过AI股票的拥挤风险，因此这次调整并不完全出乎意料。现在很多投资者关心的是：应该逢低买入，还是继续等待？瑞士宝盛仍认为，AI的结构性上升趋势将持续多年。近期最重要的催化剂，是美国超大规模云服务商在本月底公布季度业绩时更新资本开支指引。如果指引得到确认甚至上调，市场可能在本轮调整后恢复主要上升趋势。上次访谈时你对AI的态度转为更加谨慎；经历这次回调后，现在是否接近重新进入主要上升趋势的时点？ Hong Hao I think last time we adopted a very cautious stance on the AI names and many of the Chinese. AI names is down like 50% ofsome even 60%. So there's a very significant correction over a space of less than a month. So I think right now there is atechnical rebound going on in the KOSPI because KOSPI has reached a very significant technical level and also thedeleveraging process initiated by the Korean regulators may be halfway through. There seems to be a little bit of stabilizationefforts coming in as well. So I think overall we commend the Korean regulators efforts to the dele and bringing the speculativeaccess in that market, but I think the job is far from being complete as long as the leveraging process continues. Then therewill be continuing volatility and also some downside from here. So I think right now what's happening is a technical reboundbecause stocks are very oversold in a very tight space and therefore it all goes well for a technical rebound. But whetherpeople want to trade such a volatile rebound or not, that's a very different story. 上次我们对AI股票以及许多中国AI股票采取了非常谨慎的态度，其中一些股票下跌了50%，甚至60%。不到一个月内出现如此大幅的调整，说明市场波动非常剧烈。当前韩国股市出现技术性反弹，因为KOSPI已经到达重要技 术位置，韩国监管部门推动的去杠杆过程可能已进行到一半，同时市场也出现了一些稳定措施。我们总体上认可韩国监管部门减少市场投机过度的努力，但只要去杠杆继续，工作就远未完成，市场仍会有持续波动和下行压力。眼下更像是股票在短时间内严重超卖后的技术性反弹。是否参与这样高波动的反弹，则是另一回事。 Richard Tang How about the fundamental earnings? You still feel confident that that's fine? 基本面盈利怎么样？你仍然有信心吗？ Hong Hao Yeah, I think fundament is sound. There's a joke going on if the Korean bubble burst. Right now is the cheapest bubble toburst in history, right? Which is true. Korean market is trading at the lowest valuation multiple in history is trading at about six,seven times and therefore I wouldn't say that the market is terribly expensive, but because stock price overh hyped overreflected the earnings fundamentals. So therefore there is a technical correction. So I think going forward we still expect verystrong earnings growth in the semiconductor sector, but then at the same time because capex will come in waves. So the firstwave is services is this year. So Max ev's put in one trillion US dollars into the AI sectoror and then all the upstreamsemiconductor cheap names starting in benefit. But then once it gets to 2027 then investors will be looking for signs of cashflow return from these investments, w