Talk Your Book：AI的赢家与输家

今天的《Animal Spirits Talk Your Book》节目由Alger赞助播出。

Today's Animal Spirits Talk Your Book is brought to you by Alger.

访问alger.com了解更多关于ATFV的信息。

Go to alger.com to learn more about ATFV.

那就是Alger三十五ETF。

That's the Alger thirty five ETF.

集中投资组合，ATFV。

Concentrated portfolio, ATFV.

访问Alger.com了解更多。

Alger.com to learn more.

欢迎收听《Animal Spirits》，一档关于市场、生活和投资的节目。

Welcome to animal spirits, a show about markets, life, and investing.

加入Michael Batnick和Ben Carlson，一起聊聊他们在读、在写、在看的内容。

Join Michael Batnick and Ben Carlson as they talk about what they're reading, writing, and watching.

Michael和Ben表达的所有观点仅代表他们个人，不代表Ritholtz Wealth Management的观点。

All opinions expressed by Michael and Ben are solely their own opinion and do not reflect the opinion of Ritholtz Wealth Management.

本播客仅供信息参考，不应作为任何投资决策的依据。

This podcast is for informational purposes only and should not be relied upon for any investment decisions.

Ritholtz Wealth Management 的客户可能持有本期播客中讨论的证券头寸。

Clients of Ritholtz Wealth Management may maintain positions in the securities discussed in this podcast.

欢迎来到《动物精神》，我是 Michael 和 Ben。

Welcome to Animal Spirits with Michael and Ben.

今天的节目我们请到了 Dan Chung。

On today's show, we are joined by Dan Chung.

Dan是Alger的首席执行官、首席投资官兼投资组合经理。

Dan is the CEO and CIO and a portfolio manager at Alger.

我们今天聊的是他们的集中型投资组合。在市场高度集中的情况下，做集中型投资要么特别容易，要么特别难。

And what we talked about today was their concentrated portfolio, which in a concentrated market, having a concentrated portfolio has to be either really easy or really hard.

对吧？

Right?

因为你可以把集中持仓做成一个“隐形指数”，但如果你从现有的集中持仓中撤出一只股票，那就相当于下了一个巨大的赌注。

Because you could do a closet index of a concentrated positions, but if you take away from one of the current concentrated positions, you're taking a huge bet.

明白我的意思吧？

Get what I'm saying?

明白了。

I do.

回想我以前做基金经理尽职调查的时候，那时候是给大学捐赠基金和基金会做评估的，我特别喜欢那些坦诚的经理，因为我觉得你并不总能从他们那儿听到真话。

So back in my days of manager due diligence in the endowments and foundations days, I really liked when a portfolio manager was honest because I feel like you don't always get honesty out of these people.

倒不是他们故意不诚实，而是他们总想让你信任他们。

Not that they're trying to be dishonest, but they're trying to put forth they want you to trust them.

他们希望你以为他们无所不知。

They want you to assume that they know everything.

对吧？

Correct?

所以我挺欣赏那些愿意承认‘哎呀，我也不太确定’或者‘万一……呢’的投资经理。

So I like it when a portfolio manager is willing to go, well gee, I don't know or what if?

而且随着对话深入，丹聊到了Meta，这点我也很喜欢。

And I liked how as our conversation progressed, Dan talked about Meta.

他说，我们持有Meta，但我不确定他们会不会成为这次AI浪潮中的输家之一。

He says, we own Meta, but I don't know if they're gonna be one of the losers from this whole thing.

我喜欢这种自我认知，就是那种，天哪，这是一家经过验证的公司，但如果他们成了被砍掉的那一批呢？

And I like that sort of self awareness of going, man, this is a proven company, but what if they're one of the one the heads in the chopping block here?

我喜欢这种自我认知，就像在说，嘿，这完全有可能。

I I like that self awareness and like, hey, It's possible.

我们持有的某只股票可能会因为AI而成为输家。

One of our holdings could be a loser from AI.

我们现在还不知道。

We don't know yet.

对。

Yeah.

市场显然在说Meta不会是AI的赢家，但市场是对是错，现在还很不好说。

The market is clearly saying that Meta is not going to be an AI winner, and it's very unclear whether the market is very right or very wrong.

没人能预知未来。

Nobody can see the future.

所以，这，这，那，那，我同意。

So it's it's it was a it was a I agree.

那是谈话的亮点，因为一方面，天哪，这是世界上最好的广告平台。

That was a highlight of the conversation because on the one hand, man, it's the greatest advertising platform in the world.

它真的很便宜。

It's really cheap.

我的意思是，十四、十五倍的市盈率。

I mean, fourteen, fifteen times earnings.

那真是大幅折价。

That's a severe discount.

别提它本身了。

Forget about to itself.

对于市场来说。

To the market.

所以你会觉得在这里卖它就像个傻瓜。

So you'd feel like a donkey selling it here.

但另一方面，市场也不傻。

But on the other hand, the market's not stupid.

他们心里有数。

They know that.

大家确实对AI模型带来的增长可持续性有疑问，而且投入成本又是另一回事了。

There's legitimate questions as to the sustainability and the growth rates that they're that they're getting from the AI models, and the spend is a whole another story.

所以，是的，我很享受这部分讨论。

So, yeah, I I enjoyed that part of it.

因为别的经理可能会说，哎，这说不通啊。

Where where because the the another manager could be like, oh, but this doesn't make sense.

你居然要跟马克·扎克伯格对着干。

You're gonna bet against Mark Zuckerberg.

看看他的过往战绩。

Look at his track record.

显然市场判断错了。

Of course, the market has it wrong.

谁知道呢？

Who knows?

是的。

Yes.

我觉得思考AI的赢家和输家特别有意思，看着它就像嘉年华游戏一样，你打水枪，马就上下跑动。

I find it endlessly fascinating to think about the AI winners and losers and watch it's it's like the carnival game where you're shooting the the water gun and the horses are going up and back and forth.

好比喻。

Good analogy.

那感觉真的就是这样，而且作为投资组合经理，你有时候会抓狂，心想：我真不敢相信市场没看到这一点，他们本该是赢家，或者本该是输家，或者不管怎样，我真不敢相信我们没持有的这家公司表现这么好。

That's what it really feels like and and I I think it's as a portfolio manager, you have to at times rip your hair out of your head to be like, I can't believe the market isn't seeing this, that they should be a winner and they should be a loser or whatever it is, and I can't believe this company that we don't own anymore is doing so well.

为什么不是这样呢？这部分既有趣又让人压力大。

Why are how is it not so anyway, that part has to be entertaining and also stressful.

总之，我们和Alger的Dan Chung进行了一次非常愉快的对话，聊了AI的赢家和输家，他们如何看待现有企业，以及那些新兴公司等等。

Anyway, we had a really fun conversation with Dan about all the AI winners and losers, how they're thinking about the incumbents, the the newer companies that are coming up, and more.

以上就是我们和Alger的Dan Chung的访谈。

So here's our talk with Dan Chung from Alger.

丹，欢迎来到节目。

Dan, welcome to the show.

谢谢你，迈克尔。

Thank you, Michael.

你的简历看起来几乎和我的一样令人印象深刻。

Your resume looks almost as impressive as mine.

你在公司待了挺长时间了，但跟我们说说你是谁吧。

You've been with the company for quite a while, but tell us who you are.

我们为什么要跟你聊？

Why are we talking to you?

好的。

Okay.

我是Alger的首席执行官、首席投资官兼高级投资组合经理。

So I'm CEO and Chief Investment Officer and a Senior Portfolio Manager at Alger.

我从1994年就加入了Alger的分析师培训项目，一直待到现在。

I've been here since 1994 when I joined in the Alger Analyst Training Program.

九十年代的时候，我是这里的科技股分析师，也是科技板块的负责人。

In the nineties, I was known as the tech analyst and the head of tech here.

在那之前，我是一名公司律师。

And before that, I had a career as a corporate lawyer.

所以你在这一行干了很久，而且在一个地方待了很长时间，这显然说明了你所在的公司很不错。

So you've been in this business for quite a while at one place for a long time, which speaks volumes to the company that you're at, obviously.

咱们直接来做个对比吧。

Let's go right into the comparisons.

有什么不同呢？

What is different?

我们可以这样来理解。

And let's frame it this way.

现在的技术环境和九十年代中期你刚入行时相比，有什么不一样？你可以从任何角度来回答这个问题。

What is different about the technological environment, and go wherever you wanna go with this question, versus the mid to late nineties when you were starting your career?

对。

Yeah.

这个问题问得特别好。虽然很多人都在做对比，但我认为搞清楚这两个时期的差异其实非常关键。

That's a great question because, a lot of people are making comparisons, but I think the differences between the two periods are actually vital to understand.

最明显的对比当然是，就像九十年代互联网一样，现在人工智能也在蓬勃发展，这背后是由一场技术革命推动的。

The main comparison, of course, is simply that AI, just as in the nineties it was the Internet, is booming, and, of course, it's driven by a technological revolution.

但在我看来，这两者之间的区别其实非常大。

But the differences are very stark in my mind.

最大的不同在于，当前在AI投资热潮中占据主导地位的领军企业，与九十年代互联网热潮中的领军企业截然不同。

The biggest one being actually that the current leaders in the AI investment boom are very different than the leaders in the 1990s internet boom.

尤其是现在的这些公司，不管是微软、亚马逊、谷歌、Meta、甲骨文，还是像Anthropic或OpenAI这样的企业。当然，必须得把英伟达算在内。我们说的是那些基本面极其强劲的公司，有真实的营收，巨大的利润，而且盈利能力极强。它们都是行业领袖，比如微软已经领先三十多年了，亚马逊也领先了二十五年以上，而英伟达本身更是几十年的领导者。

And in particular, the companies today, whether it's Microsoft or Amazon or Google or Meta or Oracle or yes, Anthropic or OpenAI, But in that first group, and I sure, of course, have to include NVIDIA, we are talking about companies with incredibly strong fundamentals, real revenues, massive profits, highly profitable companies that have been leaders, well, like Microsoft for more than thirty years, others like Amazon for twenty five plus years, and NVIDIA itself, a leader for decades.

这些领军企业的基本面非常扎实。

The leaders have strong fundamentals.

它们正在投入资源，去抓住自己这一代人所看到的最大机遇，也就是人工智能。

They are investing in what they see as the largest opportunity of their generation, which is AI.

这一点跟九十年代形成了鲜明对比，那时候的领军企业往往并不盈利，营收也微乎其微。

And that itself is strikingly different than the 1990s where the leaders, were often companies that were not profitable, minimal revenues.

它们确实是创新者，但规模都很小。比如亚马逊上市时，营收才一百万美元，主要在网上卖书。

They were innovators, but they were very small companies, like Amazon when it went public that basically had a million dollars of revenue and was selling books online.

还有雅虎，虽然上市时的营收比那多些，但那会儿还是个初创公司，根基不稳，也没经过市场验证，更不是广告领域的领头羊。

Or Yahoo, which, yes, went public with more revenues than that, but was a fledgling company, not established or proven, and not a leader in advertising.

所以说，现在引领AI浪潮的这批领导者，跟以前完全是两代人。

So it's a very different generation of leaders leading right now in AI.

我觉得这点非常重要。这不仅关系到他们投资项目的质量，更关乎投资者应有的信心——至少从整体来看，这些公司相当成熟、靠谱，基本上知道自己该干什么。反观90年代的那些公司，虽然像亚马逊这样的企业后来证明了自己，非常了不起，但当时它们还得处处证明自己，可以说是使出了浑身解数。

And I think it matters a lot, not only in the quality of what they're investing in and the confidence that investors should have that these companies, as a group at least, are pretty mature, pretty responsible, essentially know what they're doing versus the companies in the 1990s were highly proven, were amazing companies like Amazon, but had to prove themselves and were doing absolutely everything they could.

他们做的一部分事情就是，把八十年代的那套老规矩全扔了，然后自己制定了一套新的玩法。

And and part of what they did is they threw out, you know, they threw out the the cookbook from the nineteen eighties and and implemented their own playbooks.

是的，有些确实奏效了，但就像我们知道的，很多都失败了。

And, yes, some of them work, but as we know, many of them failed.

我觉得这一代领导者很明显都是经过验证的、实打实的领军人物。

I think this generation of leaders clearly these are proven leaders here.

作为一个喜欢关注市场的人，我觉得看着这一切逐步展开挺有意思的。

As someone who enjoys following the markets, I think that this whole watching this whole thing play out is very entertaining.

作为基金经理，我挺好奇你的看法。

I'm curious for you as a portfolio manager.

你觉得在AI这场革命里挑出赢家或输家，是让你觉得有趣刺激，还是真的压力很大？

Do you find trying to pick through the winners and losers of this whole AI revolution, do you find it fun and exciting or is it really stressful for you?

你是怎么看待这种市场环境的呢？

Like, do you how do you view this type of environment?

我觉得自己特别幸运，因为我是少数既在九十年代投资过互联网科技，又……

I feel like I'm incredibly fortunate as one of the few investors that I know of who was both an Internet tech investor in the nineties.

再说一遍，我在九十年代末就负责阿尔格的技术投资，现在还在做，所以能亲眼看到我们现在的AI热潮。

Again, I headed Alger's tech investing by the end of the nineties, who's still investing today and gets to see this in, you know, the AI phenomenon that we have.

另外，当然，我手下还有一些非常优秀、经验丰富的投资组合经理和分析师。

In addition, of course, I've got some really, really talented, very experienced portfolio managers and analysts.

所以我对此机会感到非常兴奋。

And so I'm very excited about the opportunity.

我认为这是一个比互联网本身更大、更强、也更好的机会。

I think it is a bigger, stronger, and actually better opportunity than the Internet itself was.

对基本面自下而上投资的投资者来说，现在真的是个令人兴奋的时代。

And it's it's really an exciting time to be a fundamental bottoms up investor.

好的。

Alright.

我知道你刚才说了自下而上。

I know you just said bottoms up.

我想深入到底层去研究，但咱们先保持在宏观层面吧。我同意你的看法。

I wanna get to the bottoms, but let's stay on top for a I agree with you.

确实如此，你刚才提到的这些公司之间的差异非常大。

Incontrovertible, what you said earlier about the differences in these companies, very different.

但正因为差异这么多，甚至可能比共同点还多，我并不认为这意味着这件事可以无休止地持续下去，也不意味着地平线上没有潜伏的风险。

But just because there are as many differences, maybe more so than similarities, I don't think that's a green light that this thing can go on forever and that there's not any risks lurking on the horizon.

潜在的危险肯定不少。

Surely there's no shortage of potential dangers out there.

如果说有什么是让你现在稍微有点担心的，那就是：是的，这确实是变革性的。

If there were one thing that you see today that has you a little bit nervous that, yes, this is transformative.

是的，这些公司利润高得惊人。

Yes, these companies are insanely profitable.

它们全力以赴，就像之前说的那样。

They are pedal to the metal, like all that stipulated.

你今天看到的一两件事，会不会让你觉得，我不喜欢现在看到的这些？

What is one or two things that you see today that does make you think, I don't like what I'm seeing here?

有几件事。

There's a couple things.

其中一件，和九十年代的情况很像。

So one of them, and it's very similar to the nineties in this regard.

所以想想光纤，对吧？

So let's think about fiber optics, Right?

因为在九十年代，那可是大事。

Because in the nineties, that was the big thing.

铺设光纤，那场大竞赛。

Laying down fiber optics, the big race.

这场竞赛最终是合理的，但过程中有很多公司成了输家。它们砸了巨额资金，却没有合适的管理团队、战略或技术，最后纷纷倒闭。

The race was ultimately justified, but there were many, many, many losers of individual companies that spent a ton of money, didn't have the management teams, didn't have the strategies, didn't have the technology, and ended up failing.

至于人工智能，首先我认为整体来看，大家在这方面的投入是有道理的。比如OpenAI和Anthropic这些头部模型，它们的收入增长非常惊人，两者加起来的增长速度远超我们以往见过的任何情况，这也印证了投入的合理性。

With AI, first of all, I think the overall, quote, spending of this across AI, in some sense, is going to be justified, and we're already seeing it justified by the tremendous revenue growth of, for example, the leading models like OpenAI and Anthropic, which together are growing faster than anything we've ever seen.

所以很明显，这意味着市场机会很大。

So clearly, that means the opportunity is big.

但我同样认为，即将发生的颠覆性影响也非常清晰。这不仅限于互联网公司、科技公司或软件公司，而是会波及许多行业。因此，面临颠覆的风险无疑也在增加。

But I think what is also equally clear is that the disruption that is going to occur, just in internet companies, not just in tech companies, not just in software companies, but across many industries, I think the threat of disruption is, of course, also increasing.

而且，肯定会有公司被颠覆的速度远超以往，主要是受到AI竞争对手的冲击，或者是那些采用AI并运用得比其他人更有效的对手。

And there will be companies that are probably disrupted faster than we've ever seen before by basically AI competitors or competitors who adopt AI and use it more effectively than others.

所以我觉得这是我最大的担忧之一。

So I think that's one of the biggest concerns I have.

从市场角度来看，真正让我担心的是，我们有时对市场缺乏细微的理解，或者说对基本规律认识不足：一百个竞争对手去追逐一个大市场是很正常的，其中三四十家失败也是自然现象。

From a market perspective, what actually does concern me is sometimes the lack of nuance that we're seeing in market or just fundamental understanding that it's natural to have a 100 competitors chase a big market, and it's also natural that 30 to 40 of those will fail.

你知道，中间大部分公司会过得还行，而最终大概只有十家能成为顶部的赢家。

You know, a whole bunch in the middle will be okay, and you'll have maybe 10 winners at the top.

因为我们在市场上看到的波动，几乎就像是，你知道，现在大家对资本支出有很多担忧，特别是关于开源模型和那些封闭的领先模型之间的比较。

Because what we're seeing in the market around the volatility is it's almost like, you know, right now, there's a lot of consternation about the capital spending and in particular about open source models versus the leading models, which are closed.

这导致了几乎情绪化的膝跳反应，让市场变得非常不稳定。

And it's it's causing very kind of almost emotional knee jerk reactions, which is making the market really volatile.

所以这本身就是一个危险的信号。

And so that's that is a dangerous sign in and of itself.

对吧？

Right?

作为有经验的市场投资者，我们知道当你看到基本面被忽视，市场仅仅因为情绪或某家公司的一条新闻就出现极度波动时，你确实需要担心，这是不是意味着市场在某个方向上过于泡沫化了？

As experienced market investors, we know when you see fundamentals being somewhat ignored and the market really reacting in a hyper volatile way simply around sentiment or a news flash coming from one company, you do have to worry, like, is that a sign of, you know, excess frothiness on in in either direction?

在这样一个世界里，你怎么看待风险管理？

How do you think about risk management in a world like that?

因为你说得对。

Because you're right.

整个Leopold在situational awareness的情况，嘿。

The whole Leopold situation at at situational awareness was, hey.

这个人把未来十年的规划都列出来了，看起来方向是对的，但市场预期被提前了，风险也过大。

This guy laid out the next 10, and he seems like he's on the right path, but expectations got pulled forward and too much risk.

就连萨姆·奥特曼在最近的一次采访中也提到，技术确实已经到位，但普及速度比预期的要慢，因为惯性太大了。

And even Sam Altman on a recent interview said, listen, this stuff is actually like, the technology is there, but the adoption is slower than I would have expected because there's so much inertia.

所以我觉得这类事情的时间点很难把握，你可能判断对了，但时间跨度不对，或者幅度不对。

So I feel like the timing of this stuff, you can be right but be on the wrong time horizon or, you know, right but by the wrong magnitude.

那么在这种情况下的风险管理该怎么考虑呢？毕竟基本面没错，但预期和市场从众心理之间存在错位，诸如此类的问题。

So how do you even think about risk management in a situation like this where, hey, listen, fundamentally, was right but the expectation there's a mismatch between the expectations and the herd mentality and all these things.

即便你判断这项技术具有变革性，市场依然充满挑战。

It's it's it makes for a challenging environment even though you could be right on the fact that this technology is gonna be transformational.

确实如此。

Absolutely.

所以在阿尔杰，我们管理风险的首要方式是审视公司的基本面。

So, the way we manage risk at Alger is, first of all, to look at the fundamentals of the companies.

幸运的是，目前表现优异的大多是那些经验丰富的老牌企业。

We're fortunate that many of the companies that are doing really well right now are well seasoned companies.

所以不只是英伟达，大多数半导体公司都成立很久了。

So not just NVIDIA, but most of the semiconductor companies have been around forever.

美光、西部数据，还有那些半导体设备公司也是。

Micron, Western Digital, the semiconductor equipment companies.

大多数软件公司也一样，不管它们是AI赢家还是输家，比如甲骨文、ServiceNow、Salesforce，这些都是老牌大公司。

Same thing with most of the software companies, whether they're going to be AI winners or losers, Oracle, ServiceNow, salesforce.com, these are big established companies.

所以我们很谨慎，会做情景分析，看看它们的风险在哪，可能的结果是什么，当然也包括牛市情境下这些公司的乐观预期和上行空间。

So we're careful about looking at a scenario based analysis of what are their risks, what are their likely outcomes, and of course, maybe what are their bull case, the optimistic upside for these companies.

我们在阿尔杰管理风险的方式，主要是根据不同情景来评估这些股票的风险收益比。

And the way we manage risk at Alger is to consider essentially the risk reward in their stocks based on different scenarios.

比如你可以做一个看跌的假设，认为公司只会温和增长，但同时也要算出，基于自由现金流和收入，在较低的估值倍数下它值多少钱。

So you could take a bearish case of a company, think that it'll be moderate growth, but still also say, but it's gonna be worth X and Y based on its free cash flow and revenues and on a lower multiple.

你也可以从另一个角度看同一家公司：如果它在人工智能领域取得成功，增长率提升，利润率改善，那么它会获得更高的估值倍数。

And you can also look at the same company and say, if it succeeds in AI, increases its growth rate, margins improve, it'll get a higher multiple.

所以我们在投资组合中非常谨慎地控制仓位大小，始终围绕我们从基本面和目标价中看到的风险收益比来做决策。

So we're very careful about position size in our portfolios around risk reward as we see it from the fundamentals and the price targets that they imply.

最难的部分，就像我们之前提到的，是市场波动太大，我们经常看到价格变化得非常快，甚至比基本面变化的速度还要快，软件股就是最好的例子。

The toughest part as we alluded to in the earlier question is with so much volatility in the market, we're we're sometimes seeing these price prices change really very rapidly and a little bit more rapidly than they should in the sense of the fundamentals aren't changing that quickly, and the software stocks are the best example of that.

而且很多股票今年最低点时下跌了50%。

And many of them were 50% down at their lows this year.

很多股票现在已经从低点反弹了30%。

Many of them have now rallied 30% from those lows.

但这其中并没有多少是真正基于基本面的。

Not that much of that is really based on the fundamentals.

你提到的这些股票篮子，比如AI赢家、软件输家，或者当天的热门主题，你会看到这些股票一起波动，但那是短期现象。

You mentioned these baskets of stocks, the AI winners, the software losers, whatever the theme of the day is, and you see all these stocks move together, but that's short term stuff.

我们知道，长期来看，尽管它们每天一起波动，但如果你把时间拉长到三个月、六个月、一年、三年，市场会很好地区分出赢家和输家。

We know that over time, even though they move together on a day to day basis, if you zoom out a little bit, three months, six months, one year, three year, the market does a good job of separating the winners from the losers.

不是每天，也不是每周或每月，但最终市场会理清，哪些公司表现好。

Not every day, not every week or month, but eventually the market sorts, alright, these are the companies that are doing well.

这些公司才是真正面临压力的。

These are the companies that are under pressure.

你们在 Alger 的做法是这样的，我们说的是 Alger 35 ETF，它的代码是 ATFV。

And what you do at Alger, and we're talking about the Alger thirty five ETF, the ticker for this is ATFV.

这可不是那种挂羊头卖狗肉的指数基金。

This is not a closet index fund.

你们是真正下重注的。

You guys are really going for it.

比如，这里的数据有点旧了，但我们看的是截至4月30日第一季度的数据。

So for example, this bit data is a bit stale, but we're looking at data as of as of the end of the first end of April, April 30.

再说一遍，我知道这话题有点过时了，毕竟现在都八月了，但为了聊得顺畅，咱们还是提一下。

And, again, I know it's stale because we're in August, but just for the sake of of conversation.

英伟达，这是一家大公司，当时占标普500指数的7.8%。

NVIDIA, which is a large company, 7.8% of the S and P 500 at the time.

它的预期回报是14.3%。

You had a 14.3% waiting.

所以在英伟达上的主动持仓占比是6.5%。

So a 6.5% active share in NVIDIA.

Nebius，它甚至都不在指数里。

Nebius, not even in the index.

你们持有5%的仓位。

You guys had a 5% position in.

西部数据，截至4月30日，权重仅为0.2%。

Western Dig, a 0.2% weighting again as of the end of April 30.

标普500里它只占24个基点，而你们持有了4.8%。

24 basis points in the S and P, you guys were at 4.8%.

所以截至那个日期，前十大持仓占比达到了64%。

So the top 10 holdings as of that date were 64%.

大家担心标普500指数的集中度太高了。

People are worried that the S and P 500 was concentrated.

你们真的是重仓押注啊。

You guys are really leaning into it.

那这种信心是从哪儿来的呢？

So where does that conviction come from?

是的。

Yeah.

阿尔杰35号是阿尔杰投资组合里的核心标的，这份信心主要源于几个方面。

Alger thirty five is a best ideas across Alger portfolio, and the conviction really comes from a couple things.

首先，我们做这件事已经六十二年了。

One is we've been doing this for sixty two years.

我自己从事这行也有三十多年了。

I've been doing it for thirty plus.

连我的投资组合经理都比我年轻不少，但他们都有二三十年从业经验了。

Even my portfolio managers who are quite a bit younger than me have been doing it for twenty five, you know, twenty years.

我们团队非常优秀。

We have a great team.

我们在高增长、动态变化以及行业颠覆方面有着丰富的经验。

We have a lot of experience in high growth, in dynamic change, in disruption in industries.

说实话，这类事件正是阿尔杰（Alger）重点布局的时候，因为我们深知此时机会最大。

I mean, these are actually the kinds of events where Alger really leans in and understands that the opportunities are the greatest.

首先，我们有经验丰富的分析师和基金经理，他们经历过像这样的行业变革。

So first of all, it is expert analysts and portfolio managers that have experienced in industries that go through revolutions like this one.

其次，我们对公司进行细致的基本面建模分析。

Second, careful detailed fundamental modeling of companies.

你提到的每一家公司，以及我们持有的所有公司，都有详细的财务模型、损益表和现金流预测；同时我们与企业管理层保持密切沟通，并在公司内外开展大量研究，以了解其竞争地位、管理团队的质量、产品的质量，当然还有其股票相对于竞争对手的价值。

Every single one of them that you mentioned and all that we own, detailed financial models, p and l's, cash flows, and also close contact with the management team, but also a lot of research around and outside of the company itself, of course, to understand its competitive positioning, the quality of its management, the quality of its products, and ultimately, of course, the value of of its stock relative to competitors.

所以这是一项辛苦的工作。

And so it's hard work.

市面上有很多优质的公司。

There's a lot of good companies out there.

在阿尔格35基金里，我们重点聚焦那些我们认为护城河最深、当下机会最大、且长期前景也最好的公司。

Alger thirty five, we're focusing in on what we think are the best companies with the biggest moats, the best opportunities now, but also the best opportunities longer term.

你提到了一些我们认定为当前赢家、而且很可能也是长期赢家的公司。

And you've mentioned some of the companies that we've identified as winning now and and winning probably long term.

Nebius是我们非常引以为豪的一个标的。

Nebius is one we're really proud of.

我是说，这家公司我们刚投的时候还是个小盘股。

I mean, this company, when we first invested in it, was a small cap.

现在它的市值已经超过五百亿了。

It's now a 50,000,000,000 plus market cap.

那时候我们就看出来它做AI数据中心特别厉害，当时这玩意儿还没成风口呢。

What we identified there was excellent AI data centers even before AI data centers was sort of a fad.

我是说，现在好像人人都觉得好吧，其实每个州好像都在拦着建AI数据中心，尽管大家都想建。

I mean, now it's everybody seems to be well, actually, every state seems to be stopping AI data centers even as everybody wants to build an AI data center.

但也许就像我们几年前刚发现的那样，不仅是因为他们的数据中心能力，更因为他们的软件能力让他们成为了一个开发AI的平台。

But maybe as we just identified a couple years ago, not just because of the data center capabilities, but because of their software capabilities that make them a platform for developing AI upon.

当然，在过去几年里，这确实变成了现实，因为他们赢得了大量在其平台上提供帮助的合同。

And, of course, in the last years or so, it's really come true as they've won a tremendous amount of contracts for helping on their platform.

你能解释一下像这样的公司吗？看起来在最近十二到十五个月里，它简直是一路飙升。

Can you explain how a company like that it seems like in the last, I don't know, twelve to fifteen months, just went absolutely vertical.

怎么感觉好像有些公司是后来才被大家发现的？

How does it almost seems like there are these companies that are being discovered late to the game.

AI的ChatGPT时刻是在2022年，但这并不意味着一切都在那之后才发生。

It's not like, you know, AI's the ChatGPT moment was in 2022.

为什么现在有些股票看起来像是突然进入了快速追赶期呢？

What is it that it seems like there's a the catch up period happened so quickly with some of these stocks now?

也就是说，它们之前是怎么被市场忽略的呢？

Like, how are they missed essentially?

关于AI，我们现在更多是从长期投资的角度来看待。

So with AI right now, we have, sort of a long term investing.

这不算一个具体的计划，因为它没有计划那么细致，但它确实是一个宏观的战略，旨在理清像这样的建设过程中各项事务发生的先后顺序。

It's not a plan because it's not as detailed as a plan, but it's certainly an overarching strategy of understanding the sequence in which things will happen in a build out like this.

我们在两三年之前就明白了，虽然英伟达的半导体会领先，但AI数据中心将成为关注的焦点。

And we understood two and a half, three years ago that, yes, semiconductors NVIDIA was going to lead, but the AI data center was going to be a center of focus.

所以我们不仅投资了领先的半导体公司，还投资了提供数据中心技术的供应商。

And so we invested in not only semiconductors that were leading it, but also the providers of data center technology.

比如，甚至包括数据中心的液冷技术，像Vertiv这样的公司；坦率地说，还包括非科技领域的能源股，我们知道它们会提供电力，比如Talon和GE Vernova。

And also, for example, even things like liquid cooling into data centers, so companies like Vertiv, and frankly, even outside of tech in energy stocks that we knew would be supplying electricity, companies like Talon and GE Vernova.

识别像Nebius这样的公司，真的需要你是一个基本面投资者。

The identification of a company like Nebius is really something you have to be a fundamental investor.

因为第一个问题是，我们之前有计划。

Because the first question is we had the plan.

我们说，好的，数据中心会增长。

We say, okay, data centers are gonna grow.

我们跟云服务商确认过，数据中心的需求是存在的，而且AI数据中心跟其他数据中心不一样。

We confirmed with cloud providers that the need for data centers and that AI data centers are not like other data centers.

它们的架构会有所不同。

They're gonna be architected differently.

然后我们就做了很基础的研究，说，好的，世界上有哪些数据中心提供商？

Then we just did the very basic research of saying, okay, who are all the data center providers in the world?

不管大小，我们都不在乎。

The big or small, we didn't care.

我们走遍了全球的整个领域。

We went across the whole landscape globally.

接着我们就想，哪些公司看起来最能为AI数据中心做好计算方面的准备？

Then we said, okay, which ones seem best prepared for AI data center computing?

他们拥有相应的资源。

They have the resources.

他们也积累了一些经验。

They have some experience.

其实我们第一次看到Nebius的时候，觉得它当时规模还很小，但我们心里还是忍不住感叹：哇，真不错。

And actually, when we first saw Nebius, we're a very small company, but we were like, wow.

这些人的眼光非常超前。

These guys are way ahead of the curve.

他们之所以之前没被发掘， partly 是因为这家公司的股票在荷兰交易，公司总部也设在那里。

Now part of the reason they were undiscovered is that the company stock trades in The Netherlands, and the company is based there.

而且实际上，他们是从前身是一家俄罗斯互联网公司叫Yandex的。

And actually, they came out of a former Russian Internet company called Yandex.

所以他们没有出现在美国投资者的视野里。

And so they weren't on the, if you will, The US radar screen.

我觉得，在欧洲这么个不起眼的地方，藏着这么一家小盘股公司，它在AI和互联网领域都积累了不少经验。

And I would say, know, buried there in Europe was this small cap company with a lot of AI experience and a lot of Internet experience.

所以我们把范围缩小到这家公司和其他几家公司，然后去跟公司见了面，搭建了我们的模型，也了解了其他方面的价值。

And so we narrowed it down to that and other lists, and then we met with the company, built our models, understood other value.

Nebius还有几个很有意思的价值点，比如他们在一家叫Clickhouse的公司里有很大的股份，这是一家专门处理非结构化数据的AI数据库公司。

There's other aspects to value in Nebius that actually we thought very interesting, which includes they also have a very big position in a company called Clickhouse, which is AI database for unstructured data.

也就是像视频、照片这类非结构化数据。

So videos, photos, unstructured data.

而且他们居然还拥有一项自动驾驶技术的资产，这家公司之前在俄罗斯，后来因为俄乌战争在俄罗斯被关停。

And they also have actually oddly enough an asset in an autonomous driving technology that was once in Russia and then got shut down in Russia after the Ukraine Russia war.

那是Yandex吗？

Was that Yandex?

对。

Yeah.

就是Yandex。

It was a Yandex.

它们全都是雅可斯旗下的资产。

They were all Yandex properties.

有趣的是，随着这家小公司变得愈发复杂——这也是基本面分析师驱动的研究所体现的价值所在——我们作为投资者的兴趣反而更浓了。

The funny thing is as this little company became more complicated, and this is the value I think of fundamental analyst driven research, we got more interested as investors.

我们不禁感叹，哇，这家小公司竟然拥有这么多真正有趣的资产。

We're saying, wow, this little company has a lot of really interesting assets.

我之所以这么说，是因为我们是基本面投资者；如果你是纯粹的量化投资者，那这里根本没什么可看的。

And I say that as fundamental investors because if you were a purely quant investor, there would have been nothing to see here.

我的意思是，收入很低，也没有利润。

I mean, low revenues, no profits.

你可能没注意到，它从Yandex剥离出来的资产其实挺有意思的。

You wouldn't have probably seen that it had assets coming out of Yandex that were quite interesting.

你知道，那些只是持股头寸而已。

You know, these these they were just, you know, shareholding positions.

当我们与管理层会面，了解他们的历史以及那种令人惊叹的企业文化后，我们的兴趣越来越浓了。

You know, we got more and more interested when we met with the management and their history, and an incredible culture, by the way.

我的意思是，我可能说得太久了，但我们确实特别欣赏管理层和CEO。

I mean, I'm maybe I'm going on too long about it, but we really do admire the management and the CEO here in particular.

Arcadia太棒了。

Arcadia is amazing.

但这些家伙不得不在半夜离开俄罗斯，因为担心被扣押，他们确实做到了。

But these guys had to leave Russia in the middle of the night for fear of basically being seized, and they did that.

人们逃散了，遍布世界各地，但他们作为一家公司保持了团结，并且长期努力工作，就为了坚持下去，经营公司，打造这些产品，而且他们恰好有远见，就是这样。

The people fled, went all over the world, but they kept themselves together, a core of it as a company, And they worked really hard for for a long time to just stay together, be a company, build these products, and they happened to, of course, have the vision, and that that that's it.

向Arcadia和他的整个管理团队致敬，特别是CTO，他们有远见，知道AI数据中心需要不同于典型数据中心的平台，而我们想参与其中。

Hats off to Arcadia and his entire management team, his CTO also, in particular, that they had the vision like AI data centers are gonna require a different kind of of platform than your typical data centers, and we wanna be there.

他们就在那里，他们确实在那里。

And they were there they were.

新云在循环融资的案例中经常被提及。

The neo clouds are often cited in the circular financing bear case.

当你听到这个，我肯定你会觉得烦。

When you hear this, I'm sure it drives you crazy.

那些看空的人到底没理解这些关系的哪些方面，为什么这看起来一点都不像互联网泡沫的崩溃？

What is it that the bears are not understanding about these relationships and why it's not looking anything like the .com implosion?

所以我想说，对于一些领先的新云公司，比如Coralweave和Nebius，这并不是循环融资。

So I would say that in the case of some of the leading neo clouds like Coralweave and Nebius, this is not circular financing.

我是说，这些公司首先是在建设真正的资产。

This is I mean, these companies, first of all, they are building real assets.

他们在建设AI数据中心的最前沿。

They're building the leading edge of AI data centers.

对吧？

Right?

他们在建设这些数据中心，并配备，你知道的，NVIDIA最前沿的服务器芯片，似乎，你知道的，Vera Rubin现在要出来了，已经在那里了。

They are building them and equipping them with, you know, NVIDIA's leading edge server chips seem to be you know, Vera Rubin is coming out now, is there already.

所以，我是说，那是一项资产。

So, I mean, that's an asset.

那是他们正在建设的真实资产。

It's a real asset that they're building.

你想叫它办公楼也行。

Call it an office building if you want.

这种循环融资，我1999年见过类似的，当时是风投给你十亿美元，给一家零收入的公司定两千亿美元的市值，就因为你是一家带.com后缀的互联网初创公司。

The circular financing, which again, I remember from 1999, what it looked like was VC gives you a billion dollars and gives you a $2,000,000,000 market cap for a company that has no revenues, but you're an internet startup that has .com in its name.

对吧？

Right?

就是那种某某点com的名字。

You know, something.com.

当然，你们也在试图复制亚马逊、eBay或雅虎的成功。

And you're, of course, trying to follow on the success of Amazon or an eBay or Yahoo.

那接下来你们怎么做呢？

And then what do you do?

你们手里没有任何资产。

You don't have any assets.

你们花掉十亿里的五亿，投在雅虎和其他服务上做广告，目的是什么呢？

You spend of that billion, you spend 500,000,000 on on advertising, you know, on Yahoo and other services to try to do what?

吸引大量订阅用户和订阅量，以及所谓的“增长”。

Get lots of subscribers and subscriptions and and, quote, growth.

对吧？

Right?

因为你在上市时把自己包装成一家收入不多，但看看我签下了多少订阅用户的公司。

Because you sell yourself as a company going public with not much revenue, but look at all the subscribers I've signed up.

对吧？

Right?

看看我网站上所有的用户和用户数据指标。

Look at all the all the users the users, date metrics that I have for my website.

但这些都是靠广告吹起来的。

But you've pumped it up by advertising.

我的意思是，广告当然不是在建立资产。

My point is the advertising, of course, is not building an asset.

钱花出去就没了。

You spend it and it's gone.

你知道，我记得在1998年和1999年，你知道，闻所未闻的互联网公司在超级碗上买广告。

You know, I remember in 1998 and 1999, you know, unheard of Internet companies buying Super Bowl hats.

那是循环广告，一种循环融资，最终毫无结果。

That's circular advertising, sort of circular financing that's going nowhere.

所以那是顶峰吗？

So is that the top?

如果我们看到Nebius在情人节那天出现在超级碗上，你会卖掉你所有的股份吗？

If we see Nebius during the Super Bowl on on Valentine's Day this year, by way, will will you sell your entire stake?

嗯，不会。

Well, no.

所以这就是区别。

So there's the difference.

我觉得Nebius，而且关键在于，

I think Nebius and here's the thing.

会有一些比特币矿工现在转型为AI数据中心，其中一些会失败。

There are going to be I I think some of these Bitcoin miners that are now pivoting to be, I'm gonna be an AI data center, they're gonna be an some of them are going to be losers.

他们太晚了。

They're too late.

那个反数据中心的事情会挡他们的路。

The the anti data center thing is gonna get in their way.

他们不是第一个排队买英伟达芯片或内存的。

They're not they were not first in line to sign up for NVIDIA chips or memory.

你知道，我觉得他们会遇到很多物流问题。

You know, they're gonna have a lot of logistical problems, I think.

而且你知道，在任何市场里，如果市场总值是X，它不会按市场份额来分。

And as you know, in any market, you know, if the value of the market is x, it doesn't get split up by market share.

你知道，如果你有1%，你就拿1%。

You know, if you have 1%, you get 1%.

你有10%，你就拿10%。

You have 10%, you get 10%.

不。

No.

赢家通常是那些占据最大市场份额的公司，它们的估值倍数也会高于其市场份额。

The the winner is the one who gets the biggest market share tends to get a multiple of its market share and value.

对吧？

Right?

它们能获得超额的估值溢价。

They get sort of outsized value.

在制药行业，大家很清楚一个规律：如果有三款药，排名第一的药通常能拿下50%的市场份额，排名第二的大概占25%到30%，剩下的公司就只能去争抢残羹冷炙了。

In pharmaceuticals, it's it's well known that, like, very typically, if you have three drugs, you know, the number one drug gets 50% of the market, the number two drug gets, like, 25 or 30, and then everybody else gets know, is fighting for the scraps.

所以你看，排在第一或者第二的位置价值要大得多。

And so, you know, that number one or number two position is much more valuable.

回头看过去两三年，赢家输家的变化挺有意思的。一会儿说谷歌不行了，一会儿又说谷歌领先了；接着微软又领先了，其他家就落后了。看着这些老牌巨头像是在赛马一样，这边刚宣布谁谁谁要完蛋了，那边又觉得肯定是自己最后能赢。而且看起来大部分公司好像是一起All in的。

It's interesting to see how much the change has been over the last two to three years of who those winners and losers are, and it seems like, oh, Google's out in first Google's dead, then Google's out in front, then Microsoft's out in front, and they're lagging, and just to see these incumbents kind of have this horse race where, you know, people are declaring them dead and they know they'll be the leader, and how much of this can be where because it seems like most of these companies kind of just gone all in together.

感觉就像他们手拉手一起跳桥似的，决定要把各自的自由现金流都砸进去。

It's like they've held hands and jumped off the bridge and decided we're gonna listen, we're gonna plow through our free cash flow all together.

大家都这么干。

Everyone's gonna do it.

如果一个人去做，那我们所有人都会去做。

If one of us is gonna do it, we're all going to do it.

是啊。

Yeah.

这个机会对这些公司来说有多大空间？比如说像Mag七或者前十名，就算不是明显的赢家，至少它们都能活下来，过得不错吧？

How much room is there with this opportunity for all of these companies in say like the Mag seven I guess or that, you know, the top 10 to come out, if not the clear winners, at least they're all gonna survive and be okay?

还是你觉得那批大公司里会有人真的输掉？

Or do you think that there's going to be real losers from that big group of names?

首先，我很喜欢你总结的谷歌那部分，完全正确。

So first of all, I I love how you recapped Google because that's exactly right.

而且这是我们风险评估的一个好例子，因为我们从2006年谷歌IPO时就一直持有它，很了解这家公司，也明白为什么人们觉得它面临AI的威胁。

And it's a good example of our risk reward, our scenario because when we saw we've been in Google since 2006 in its IPO, know the company well, understand why people think it was at risk from AI, I.

简单来说就是，哦，好吧，我要用ChatGPT了。

Simplistically, oh, well, I'm gonna go to ChatGPT.

我再也不搜索了。

I'll never do a search again.

但这种想法大错特错，因为它完全低估了“我们”是谁。

But this is so wrong because it completely under understands who is the we.

这里的“我们”指的是数十亿普通用户，他们可能并不特别在意ChatGPT本身。

The we is billions of individual people who probably don't really care about ChatGPT per se.

他们在意的是，长期以来一直使用并喜爱Gmail和谷歌搜索。

What they do care about is that they've been using and loving Gmail and Google Search for a long time.

他们喜欢YouTube，也喜欢谷歌的客户服务。

They love YouTube and their customers.

他们把自己和这家公司联系在一起。

They're they associate themselves with this company.

所以谷歌并不需要成为第一个发明任何东西的公司。

So Google didn't have to be, you know, the first to invent anything.

我们一直认为，当股价因负面情绪而不断下跌时，是市场搞错了。而谷歌现在正在证明这一点。让普通人接触AI最简单的方式，就是拥有领先的搜索产品，然后用AI来增强它，提供越来越好的搜索结果。

And we we always thought that as the stock went down and down and down on the negative sentiment, that people had got it wrong, that and and Google's proving it right now, which is the easiest way for introduce an ordinary person to AI is you have the leading search product, and then you would power it with AI, and you give them even better and better search results.

这正是谷歌现在所做到的。

So that's exactly what's happened with Google.

更重要的是，人们忽略了谷歌在基础技术上的投资，比如他们自己的半导体技术，还有云计算，以及如何运营超大规模云环境，这些现在都完全得到了回报。

More importantly, people miss that Google's investments in fundamental technologies, so their own semiconductor technology as well as in the cloud and run and and how to run a hyperscaler cloud environment is totally paying off.

对吧？

Right?

谷歌云上个季度增长了100%。

Google Cloud is growing 100% last quarter.

这些都是非常大的业务，而且增长非常快。

These are very big businesses growing super fast.

这可以说是证明了谷歌的技术卓越性非常高。

It's it's a, you know, it's a confirmation that the technological excellence at Google is is very, very high.

而且，我敢说。

And, you know, I'll I'll hazard.

我确实觉得关于Meta的问题更多一些。

I I I do think there's more question about Meta.

我得说，我对Meta的担忧有点奇怪，因为我们很早就把它认定为人工智能短期内的赢家，从某种意义上说，他们一直在用人工智能来优化数字广告定向投放。

I have to say my concern about Meta is a little odd in that we identified it early as, frankly, a winner from AI short term in the sense that they've been using AI to power their digital ads targeting for a while.

Reels 火得不行。

Reels is on fire.

问题是这能持续多久？

The question is how much longer can that go on?

而且他们只有一项业务。

And they have only one business.

不像谷歌，他们基本上只有广告业务。

Unlike Google, they've only got basically an advertising business.

对。

Right.

他们试过的其他东西都没怎么成功，

Everything else they've tried hasn't really worked very well,

对吧？

right?

没错。

Exactly.