← 返回任务列表

Special Guest Bill Ackman on Markets, AI and Concentrated Investing

375 段 · 3 位说话人 · 原片 33:41
M1
M10:00

本期节目中表达的观点仅代表 trust manager 本人,并不构成买入、卖出或持有任何投资的建议,也不一定代表 Hargreaves Lansdown 的观点。

The views expressed during this episode were those of the trust manager and not a recommendation to buy, sell, or hold any investments, nor do they necessarily reflect the views of Hargreaves Lansdown.
F1
F10:10

大家好,欢迎收听 Hargreaves Lansdown 的 Switch Your Money On 播客。

Hello, and welcome to the Switch Your Money On podcast from Hargreaves Lansdown.

我是 Anna Macdonald。

I'm Anna Macdonald.
M1
M10:16

我是 Matt Britzman。

And I'm Matt Britzman.
F1
F10:17

今天是六月十一日,星期四,我和 Matt 很高兴请到了 Pershing Square Capital Management 的创始人兼 CEO,Bill Ackman。

Today, which is Thursday, the 11th of June, Matt and I are joined by Bill Ackman, founder and CEO of Pershing Square Capital Management.

他是过去几十年里最知名的主动型投资人之一。

One of the best-known active investors of the last couple of decades.

Pershing Square 采取的是一种非常聚焦的策略,通常只持有大概八到十二个高确信度的投资,主要是在美国的大型公司里。

Pershing Square runs a very focused strategy, typically around 8 to 12 high-conviction investments, mainly in large US companies.

这很明确,不是那种分散化、类似指数的投资方式。

It's very deliberately not a diversified index-like approach.

在英国,很多投资者应该都熟悉上市载体 Pershing Square Holdings,这是一只市值七十亿英镑的 investment trust,可以让投资者接触到这套策略。

In the UK, many investors will be familiar with the listed vehicle Pershing Square Holdings, a £7 billion market cap investment trust which gives access to that strategy.
M1
M10:54

对,Anna,所以这只 trust 的一个重要特点就是所谓的 permanent capital,这也是我们接下来想和 Bill 聊的,看看他为什么喜欢这种结构。

Yep, Anna, so one of the important features of the trust is this idea of permanent capital, and it's something that we're going to talk to Bill about in terms of why he likes that structure.

不过先补充一点背景,和开放式基金不同,开放式基金里投资者每天都可能赎回;而 closed-end investment trust 不会这样,通常这会让管理人能够采取更长期的投资方式。

But just for a bit of kind of background, unlike an open-ended fund where investors are redeeming day to day, that doesn't happen with a closed-end investment trust, and typically that can give a manager the ability to take a longer-term approach.

这个点我们也希望等会儿能请 Bill 展开讲讲。

Again, hopefully something that we'll get Bill to talk about.

我们还会和 Bill 更广泛地聊聊市场。

We're also going to talk to Bill about markets more broadly.

我们会谈到 AI,大概也会聊到一些 SpaceX、OpenAI、Anthropic,还有我们现在看到的一些大型 IPO。

We're going to cover AI. There'll likely be some chatter on SpaceX, OpenAI, Anthropic, some of the big IPOs that we're seeing.

所以我觉得这一期应该会很精彩。

So I think this is going to be a good one.
F1
F11:31

Bill,早上好,欢迎来到我们的播客。

Good morning, Bill, and welcome to our podcast.

我是 Anna Macdonald,Investment Strategy Director。

I'm Anna Macdonald, Investment Strategy Director.
M1
M11:37

我是 Matt Britzman,Senior Equity Analyst。

And I'm Matt Britzman, Senior Equity Analyst.

你好,Bill。

Hi, Bill.

感谢你来参加我们的节目。

Thanks for joining us.
M2
M21:40

当然。

Of course.
F1
F11:41

在我们聊具体个股之前,能不能先谈谈,为什么拥有一个 closed-end vehicle 对你来说这么重要?

Before we get into the individual stocks, can we start on why having a closed-end vehicle is so important to you?
M2
M21:51

你看,现在的市场已经变得越来越短期化了。

Well, look, the markets have become increasingly short-term.

越来越多的资金掌握在一些人所谓的 pod shops 手里,也就是那种把资金分配给大量小团队的机构,这些团队的报酬是根据相对短期的业绩来决定的,而且他们对所谓 risk limits 卡得非常严,也就是说,只要出现任何程度的下行波动,他们基本上就会被迫卖出这些名字。

More and more capital is controlled by what some people call pod shops, firms where capital is deployed among a large number of small teams, where the teams are compensated based on relatively short-term performance, and where they're very tight with what they call risk limits, meaning if there's any degree of downside volatility, they're sort of forced out of names.

现在的市场还有一个特点,就是 index fund 持有的公司比例越来越高。

The market's also characterized by index funds that are owning an increasing percentage of companies.

而这部分资金会把资本从流通盘里拿走,这就意味着边际上的买家和卖家会有更大的影响力。

And that money takes capital out of the float, which means the marginal buyer and seller has bigger influence.

结果就是,波动会大得多。

The result of that is a lot more volatility.

而在这种环境下,你会经历一些阶段,明明质量非常高的企业,却在以非常深度折价的价格交易。

And in that kind of environment, you can go through periods of time where very high-quality businesses are trading at very deeply discounted prices.

但如果你的资本是短期的,本质上,这些投资者持有的就是短期资本。

But if your capital is short-term, effectively, those investors have short-term capital.

一旦情况比预期稍微差一点,他们就会被迫卖出。

They're forced to sell when things are even slightly worse than anticipated.

这就创造了一个机会,前提是你的资本基础能够承受这种短期波动。

And that creates an opportunity as long as your capital base can sustain that short-term volatility.

这也是我们优势所在。

And that's what gives us the advantage.

所以这就是为什么我们实际上已经转向了一种模式,让我们几乎百分之百的资本都放在这些永久性结构里。

And that's why we've really gone to a model where nearly 100% of our capital is in these permanent structures.

这样一来,我们就能做长期投资者。

So, it enables us to be a long-term investor.

你想想看,什么叫公司?如果你选的是一家好公司,它本质上就是一个无限存续的实体。

And if you think about what is a company, a company is, if you pick a good one, it's an infinite-life entity.

而这门生意的价值,就是它未来能产生的现金流折现到今天的现值。

And the value of that business is the present value of the cash you can generate from it.

如果你只持有这家公司一个月或者两个月,就会错过很多重要机会。

To own that business for a month or 2 is going to miss out on significant opportunities.
F1
F13:19

好。

OK.
M2
M23:20

因为你被迫只能在这么短的时间里持有这家公司。

You're forced to own the business for such a short period of time.
F1
F13:23

所以这是不是意味着,就你的投资方式来说——因为你的组合是高度集中的——你会坐在那里想,如果市场出现波动,比如说发生抛售,我们手上已经准备好了几个点子,像储备阵容一样,到时候就能拿出来直接上场?

So does that mean that you are, in terms of the— you have a concentrated portfolio, that you are sitting and thinking, right, if the market moves, you know, for example, if there's a sell-off, we are ready to go with a couple of ideas in the sort of in the stable that we can bring them out and bring them onto the field?

你是这么想的吗?

Is that how you think about it?
M2
M23:41

我们把它叫作一个 library,不叫 stable。

We call it a library as opposed to a stable.

不过对,没错,我们建立了一个由很多高质量公司组成的 library,这些都是我们很想持有的公司。

But yes, we've built a library of very high-quality companies that we'd love to own.

实际上我们现在正处在一个阶段,这类公司——也就是我们喜欢的那种公司——正能以非常有吸引力的价格买到。

And actually we're going through a period right now where those kind of businesses are available at very attractive prices— the kind of companies we like.
F1
F13:54

好,那比如说过去五年里,你觉得市场的这种表现方式——随着被动基金越来越多,还有越来越多的高频交易——是不是在某种程度上给你们提供了这些机会?

Okay, so do you think over the last 5 years, for example, that the way that the market's behaved, with increasing passive funds and more and more high-frequency trading, that that is somehow providing these opportunities for you?

但到了某个阶段,你觉得这种平衡会不会发生转变?

But at some stage, do you think that balance shifts?

还是说,你觉得这会成为未来持续存在的一个特征?

Or do you think this is going to be a feature going forwards?

如果我们拿能不能跑赢指数来衡量自己的话,那这种主动管理风格到底会在什么时候重新体现出跑赢指数的表现?

At what point does this active management style really come back into delivering index-beating performance if this is what we're going to measure ourselves against?
M2
M24:24

我觉得,绝大多数主动管理人不可能跑赢市场,但其中少数人是有可能做到的。

I think it's not possible for the substantial majority of active managers to beat the market, but it is possible for a minority of them to.

我不知道这算不算一种趋势性的事情,但我确实认为,只要你的客户不离开,机会集合就会变得更大。

I don't know if it's a trend-based thing, but I do think the opportunity set becomes greater as long as your clients don't leave.

我们的表现一定会和市场不一样,因为我们的持仓非常集中。

Our performance is going to be different from the market because we're highly concentrated.

而这在过去二十二年里所意味着的,实际上是非常巨大的。

What that's meant over the last 22 years is, is actually massive.

超额收益,你知道,接近每年六百到七百个基点的超额收益。

Outperformance, you know, approaching 600, 700 basis points per annum of outperformance.

但我们也会经历一些阶段,在那些阶段里我们会大幅跑赢指数。

But we go through periods where we dramatically outperform the index.

你看,哪怕就是过去这几个月,就是这样的一个阶段。

You know, just even the last, last few months is a period like that.

今年 S&P 大概涨了百分之七到百分之八左右。

S&P's up something like 7 or 8% for the year.

我得查一下。

I have to check.
F1
F15:05

我觉得大概——我今天早上看了一眼,我觉得大概是百分之六点五,而你们这边,嗯,你们还是负收益。

I think it's about— I looked this morning, I think it's about 6.5%, and you're like, you know, you're in negative territory.
M2
M25:11

你想象一下,如果在恰恰最应该部署资金的时候,你所有客户都把钱撤走了,会是什么样的世界。

And imagine a world in which all your clients pulled your money at precisely the time that you should be deploying it.

我们现在就在一个价格非常有吸引力的市场里部署资本。

We're deploying capital in a very attractively priced market.

市场现在兴奋的,我会说,是那种闪闪发亮的新东西,也就是半导体这一块,你知道,整个跟 AI 基础设施相关的投资机会。

The market's excited about, I would say, the shiny new thing, which is sort of semiconductors, you know, the whole sort of AI infrastructure play.

然后呢,你知道,可能还有 SpaceX。

And then, you know, perhaps SpaceX.

我觉得有很多钱正在从市场里流出来,因为——我其实听到一个挺有意思的轶事,我觉得很可能是真的。

I think a lot of money is coming out of the market as— I actually heard an interesting anecdote, which I think is likely true.

你知道,SpaceX 的 IPO,我想基本上就是今天进行,我觉得这只股票开始交易——我想是明天。

You know, the SpaceX IPO, which happens, I think, basically today, I think the stock starts trading— I think it's tomorrow.
F1
F15:43

我觉得今天已经截止了。

I think it's closed today.

然后它就开始——我想是明天开始交易。

And then it starts— I think it starts tomorrow.
M2
M25:48

你想象一下,如果你是一个 endowment,手里投了 SpaceX,现在它会从你的私募投资组合转到你的公开市场投资组合里。

Imagine you're an endowment that has an investment in SpaceX, which now is going to go from your private portfolio to your public portfolio.

而且因为它实在太成功了,它会让你在公开市场这部分的配置,如果这么说的话,变成超配。

And because of its tremendous success, it's going to give you overweight, if you will, your allocation to publics.

那你就会被迫卖出一点,好把整体重新调回到原来的配置范围内。

You're forced to sell to bring things sort of back into line.
F1
F16:06

对,这很有意思。

Yeah, that's interesting.

而且其实,从指数的角度来看,S&P 的做法挺有意思的,他们决定不会快速纳入 SpaceX 或其他大型 IPO。

And actually, when we think about it from an index point of view, the S&P, it's quite interesting that they decided they wouldn't fast-track SpaceX or other mega-IPOs.

如果它们的自由流通股比例还是那么低,他们就不会快速纳入。

They won't be fast-tracking them if they've still got that very low free float.

那么低的自由流通股比例,很可能会导致股价朝那个方向出现相当剧烈的波动。

That very low free float will lead to probably some quite dramatic share price movements in that direction.
M2
M26:26

这点我要给 S&P 点个赞。

I give credit to S&P.

你知道,我觉得某种程度上,Nasdaq 是向营销机会妥协了,为了从 New York Stock Exchange 手里抢到这个业务。

You know, I think sort of Nasdaq caved to the marketing opportunity, winning the business from the New York Stock Exchange.

我觉得 S&P 想得更长远一些。

I think S&P is thinking kind of longer term.
F1
F16:35

对,没有,我听到这个消息的时候还挺高兴的。

Yeah, no, I was quite pleased to hear about that.

所以投资者应该这样去想:这是一个高度集中的投资工具,它清楚自己想在更长期实现什么目标;不过,拿它去跟指数做基准比较,也许不是一个正确的思路。

So investors should be thinking about it as this is a concentrated vehicle that knows what it wants to achieve over the longer term, but perhaps benchmarking it to something like an index is not the right way to think about it.
M2
M26:47

不,我觉得拿它跟指数做基准比较是对的,但我觉得你做这个比较的时间跨度应该是几年,而不是几个月。

No, I think benchmarking it to an index is the right way to think about it, but I think the time horizon with which you should do the comparison is several years as opposed to several months.
F1
F16:56

好的,很棒。

Okay, great.
M1
M16:57

Bill,你刚才提到,你管理的是这样一个集中持仓的投资组合——你能不能简单讲讲,你当初是怎么形成这个思路的,比如让投资组合里只放八到十二只股票,以及你为什么偏好这种集中度?

Bill, you mentioned that this idea that you run this concentrated portfolio— can you just touch on how you came to the idea of having, you know, 8 to 12 stocks in the portfolio and why you prefer that concentration?
M2
M27:11

当然。

Sure.

其实,我们这个行业里很多人都会对我们说,Bill,你知道,如果是替我自己管钱,我会完全按照你的方式来做,但我的客户不允许我这么做。

So actually, many people in our industry say to us, Bill, you know, I would manage money exactly the way you do for myself, but my clients don't let me.

而集中持仓的好处,我觉得其实非常明显,就是你可以在一组你觉得有意思的想法里,只持有最好的那些。

And the benefit of concentration, I think, is pretty obvious, which is that you can select among your group of interesting ideas to own only the best ones.

既然你可以多持有第 1、第 2 或第 3 个想法,为什么还要去多持有第 20 个呢?

Why own more of idea 20 when you can own more of idea 1, 2, or 3?

而人们之所以采用没那么集中的方式,你知道,是因为分散化会给你一个更平滑的结果。

And the reason why people operate in a less concentrated fashion, you know, diversification will give you a smoother outcome.

在资产管理行业里,让你丢饭碗的方式,就是做出和别人相反、而且表现不好的业绩。

The way to get fired in asset management is to have performance that's adverse to others.

而如果你采取这种方式,你注定会有差异化的表现,也就是说,偶尔会出现你跑输别人的阶段。

And you're guaranteed to have differentiated performance, i.e., occasional periods where you underperform others.

如果你的资金可能一夜之间就消失,那不去紧贴指数,反而会是一个谨慎得多的策略。

And if your money can disappear overnight, then it's a much more prudent strategy to not hug the index.
F1
F18:02

对。

Yeah.
M2
M28:02

我之前讲过,有多少资本是按照指数来被动管理的。

I talked before about how much capital is passively managed according to the index.

那说的是合同意义上的被动管理资本,也就是实际放在 index fund 里的钱。

That's the contractually passively managed capital, the money actually in index funds.

但绝大多数其他投资人其实都是紧贴着指数走,只有在边际上才会去配置资本。

But the vast majority of other investors hug the index, and only on the margin do they deploy capital.

因为如果股市跌了,你也跌了,那你被炒掉的可能性就更小,对吧?

Because if the stock market is down and you're down, you're less likely to be fired, right?

如果股市涨了,而你的表现跑输大盘,那大家通常就会把钱转到别处,除非他们采取的方式是卖掉 closed-end fund 的份额,而不是卖底层资产——这样就不会逼着你去卖掉底层那些公司。

If the stock market's up and you're underperforming, then people generally take their money elsewhere, unless their method of doing so is selling an interest in a closed-end fund, as opposed to the underlying— forcing you to sell the underlying companies.
M1
M18:39

对。

Yeah.

Bill,我们稍微换个话题。

Bill, let's shift a little bit.

一直以来,大家对你的印象都是 activist investor,对吧,就是去找那些你有机会推动改变的公司。

You have a reputation historically of being an activist investor, right, looking for companies where you can potentially instill change.

你在这方面的策略,这些年有变化吗?

Has your strategy on that shifted over time?

比如说,你现在是不是更倾向于找那种你不用跑去逼董事会卖掉半个公司,或者给投资人写公开信之类的公司?

You know, are you now more looking for companies where you don't have to go and force the board to try and sell half the company, or write open letters to investors, etc.?

我知道你最近上过 Liquidity 和 All In。

I know you were on the Liquidity and All In recently.

你提到过一家公司,当年你拿了百分之十的股份,结果 CEO 连电话都不回你。

You talked about a company where, back in the day, you took a 10% stake and couldn't get the CEO to call you back.

你现在是不是已经把这种思路转向了,少一点对抗,多去找那些不需要你采取 activist approach 的企业?

Have you shifted that mindset now to less fighting and more kind of finding the businesses where you don't have to take that activist approach?
M2
M29:19

其实我们一直都在找那种不用我们做太多事,也能得到成功结果的公司。

We were always looking for businesses where we didn't have to take— do less work and get a successful outcome.

你知道,显然我们最愿意做的,是去支持那些最优秀、可持续增长的公司,嗯,那种管理层本来就在把所有事情都做对的公司。

You know, obviously we'd love to help the greatest durable growth companies, which are, uh, management's already doing all the right things.

那一直都是我们的偏好。

That's always been our preference.

一般来说,或者说历史上一直都是,想找到一家质量非常高、经营得非常好、同时估值又特别有吸引力的公司,是很难的。

It's generally hard, or it has been historically hard, to find a very high-quality business run extremely well, trading at a very attractive valuation.

不过我想说的是,就像我前面提到的,市场里这种外来的短期波动,让这种机会变得更常见了。

However, what I would say is, you know, as I spoke before, that introduced short-term volatility in markets has made that a much more common occurrence.

所以我会说,尤其是最近,我们可以建立一个由非常高质量公司组成的投资组合,而且我们对它未来回报的估计,大概能到百分之二十几的中高段。

So, we can build a portfolio, I would say more recently, of very high-quality businesses where our go-forward estimate of return is something in the mid to high 20s.

就是说,今天我们就能做到。

You know, we can do that today.

而且基本上,这些企业本身就已经经营得很对了。

And basically, these businesses are already operating correctly.

所以你永远都是先选这种。

So, you always choose that first.
F1
F110:04

对。

Yeah.
M2
M210:04

第二点我想说的是,你知道,我们二十多年前刚进入这个行业的时候,还没有业绩记录,也没有什么名声。

The second point I'd make is, you know, when we entered this business 20-odd years ago, we did not have a track record or reputation.

我们那时候也还没有做出过成功的 activist investment。

We had not had a successful activist investment.

但二十年做下来之后,大家就知道我们是谁了。

After you've done it for 20 years, people know who we are.
F1
F110:17

对。

Yeah.
M2
M210:18

而大家对我们的反应——其实只有在你进不了董事会,或者至少他们不尊重你说的话的时候,才需要 activism。

And the response to us— activism is only needed if you can't get into the boardroom, or at least they don't respect what you have to say.

而这一点真的已经变了。

And that's really changed.

所以现在的结果是,我们依然会投资那些还有改进空间的公司,但我们可以关起门来谈这些事,而不是跑去 conference hotel 的 ballroom 里开大会。

And so the result today is we still make investments in companies where there are opportunities for the businesses to be improved, but we can have those conversations behind closed doors as opposed to hosting people in a conference hotel ballroom.
F1
F110:39

对,我是说,当然,自从我开始投资以来我确实注意到,因为你提到的那些因素,比如 pension funds,还有 high-frequency trading,你现在确实会看到一些非常惊人的波动,发生在原本体量非常大的公司身上,而这种情况以前是不会发生的。

Yeah, I mean, certainly I've noticed since I started investing that you do get unusually— because of the aspects that you mentioned about pension funds and about high-frequency trading— you can actually get these incredible moves in what were very large companies that just wouldn't have happened.

以前不会有这么大幅度的波动。

You wouldn't have had that magnitude of the move before.

所以就像你说的,你确实会有那些进场的机会,而且你也会看到持仓价格很快上涨,或者很快下跌。

So you do have, as you say, those options to get in, you know, and to see pretty quick price appreciation or price depreciation in holdings.

我刚才在想你说的,这些公司本身其实已经管理得非常好了。

I was wondering about what you said about these companies are already very well managed.

而且考虑到你很爱发 Twitter,我们也经常会听到你的观点。

And interested, given that you're a prolific tweeter, we, you know, we hear about your views.

你觉得你投资的这些公司——我是说其中一些大公司——是不是也不得不让自己变得更政治化一些?

Do you think the companies that you're investing— I mean, some of these big companies, do you think they've had to become more political in themselves?

我的意思是,你觉得它们是不是必须得应对我们在美国看到的这种情况,或者其他国家也有,但尤其是在美国,才能取得领先?

I mean, do you think they have to deal with what we see in the US, or in other countries, but particularly in the US, to get ahead?
M2
M211:30

嗯,我会说,我们现在有一位更倾向于干预的总统。

Well, I would say we have a more interventionist president.

是在 capital markets 这件事上。

In the capital markets.

你看,President Trump,还有 Biden,举个例子,如果你看 Intel 那个案子,为了某种程度上帮助美国的竞争地位,他给了 Intel Corporation 一笔补助,去帮助一家当时有点挣扎的——你可以把它看作是一家 chipmaker。

You know, you've seen President Trump, you know, Biden, for example, if you look at the Intel case, you know, in order to kind of help the US competitive position, made a grant to the Intel Corporation to help what was kind of a struggling— you think of it as a chipmaker.

而 Trump 的回应是,好吧,那作为这笔纳税人投资的交换,我要 ownership。

And Trump's response was, well, I want ownership in exchange for that taxpayer investment.

所以你知道,如今你可不想站到总统的对立面。

So, you know, you don't want to get adverse to the president these days.
F1
F112:01

对。

Right.
M2
M212:01

所以我觉得,从这个角度看,美国企业界肯定是在关注政治的。

And so I think corporate America is definitely paying attention to politics from that perspective.
F1
F112:06

嗯。

Yeah.

你对现在正在发生的这些事感到意外吗?

Are you surprised by what's going on at the moment?

很难知道 Trump 在进入 Iran 的时候,心里到底预期的是什么。

It's hard to know what Trump expected when he went into Iran.

你觉得在 midterms 之前会不会有很大变化?还有他会怎么回应 affordability 这个问题?

Do you think that there's going to be much change ahead of the midterms and how he's going to respond to affordability?

那本来应该是今年最重要的关键词。

That was meant to be the big watchword of this year.
M2
M212:22

你看,我觉得这要归功于总统,他做了一件我认为会带来一代人层面积极影响的事,代价就是你知道,要承担短期的政治风险,对吧?

Look, I think it's a credit to the president that he did something that I think will have a generationally positive impact at the cost of you know, short-term political risk, right?

他在进入 Iran 之前就知道,带来的影响会是 oil prices 更高,gasoline prices 也更高。

He knew going into Iran that the impact would be higher oil prices and higher gasoline prices.

而这恰恰是那种你非常直观就能感受到的事,你知道,你每隔几天去加一次油,价格一下子动个百分之二十、百分之三十。

And it's the one thing that you can very clearly see, you know, when you fill up the tank every few days and the price, you know, moves 20, 30%.

这东西是会直接打到消费者钱包上的。

That's something that hits the pocketbook of the consumer.

而且从政治上讲,这一直都是人们会有反应的事情。

And politically, it's always something that people respond to.

不过,对美国来说,比短期能源价格更大的风险,是一个拥有核武的 Iran。

However, a bigger risk to the United States than short-term energy prices is a nuclear Iran.

所以我觉得,他做的是一个着眼长期的政治决定,同时接受了短期的政治后果。

And so I think he made the long-term political decision with the short-term political consequences.

这一点我得给他肯定。

I give him credit for that.

我觉得,就 Iran 这场战争而言,我们现在更像是进入尾声了。

I think we're more in the endgame with respect to the Iran war.

我觉得这件事会在 midterms 之前解决。

I think it will be resolved before midterms.

我觉得,你知道,你可能会看到能源价格、shipping rates,还有所有这些东西都出现很大的回落。

I think, you know, you could see a big reversion, obviously, in energy prices, shipping rates, all this stuff.
F1
F113:17

对,它确实能很快恢复过来。

Yeah, it can come back quite quickly.

我的意思是,有些人说,我们可能很快就会到某种临界点,到那时储备都已经用光了。

I mean, some people say that we're going to be at some kind of tipping point soon where reserves have all been used up.

但希望吧,对,希望在那之前就能解决。

But hopefully, yeah, it gets resolved before then.

不过我觉得,你在关税政策上应该是更直言不讳的,我想。

But I think you were more outspoken about the tariff policy, I think.
M2
M213:30

当我认为总统做的是对的事情时,我一直是支持他的。

I've been supportive of the president when I think he's doing the right thing.

而当我感觉到——我是说,当我觉得那种关税做法,尤其是最开始实施的方式,风险非常非常高的时候,我的态度就会更谨慎,甚至有些担忧。

And I've been, I would say, cautious to concerned when I felt that— I thought the tariff approach was very, very high risk the way it was sort of initially implemented.

我同意总统的看法,就是你知道,长期以来,我们让其他国家——用个引号的话说——利用了美国,通过不公平的贸易关系占了美国的便宜,而关税是解决这个问题的一个好办法。

I agreed with the president that, you know, over time we've allowed other countries to, quote unquote, take advantage of the United States with unfair trade relationships and that tariffs are a good way to resolve them.

但是那种几乎一夜之间就推出大规模关税的做法——是我先提出了这个九十天暂停的想法,然后总统某种程度上采纳了。

But the introduction of massive tariffs in a kind of overnight fashion— I initiated this 90-day pause idea, and the president kind of took it up.
F1
F114:02

有意思。

Interesting.

我是说,如果风向又转回去,Democrats 赢得下一次选举的话,你觉得会发生什么?

I mean, do you have thoughts about what you think is going to happen if it were to swing back and the Democrats would win the next election?

你觉得他们会保留那种更偏 America First 的政策,想继续把关税维持下去吗?

Do you think they would keep that kind of more America First policy and want to keep tariffs in place?
M2
M214:15

嗯,Supreme Court 对关税这件事其实已经表达过他们的看法了,所以这就有点难。

Well, the Supreme Court has kind of had its view on tariffs, so it's hard.

总统如果没有 Congress 的支持,就很难把这个当成一种工具来用。

The president to use that as a tool without the support of Congress.

你知道,还是那句话,我们的制度跟你们的有点不一样。

You know, again, our system's a little bit different from yours.

总统在中期选举之后也还是总统。

President's still going to be president after midterm elections.
F1
F114:29

哦,对,对,不好意思,我是指二零二八年。

Oh, yeah, no, sorry, I meant in 2028.
M2
M214:32

那在这个时代里,感觉就像是特别遥远的事。

That just seems like in this era, very far away.
M1
M114:36

好。

Okay.

Bill,我们转到市场的话题吧。

Bill, let's switch to markets.

更广泛地说,还有 AI,我想听听你对这个的看法。

AI more broadly, and I want to get your take on this.

感觉好像每隔几个月,围绕 AI 的叙事就会摇摆一下,会冒出一些担忧。

So it feels like every few months we have a kind of swing in the narrative with a bit of worry about AI.

我知道在美国,AI 在消费者当中其实很不受欢迎。

I know that in the US AI is very unpopular with consumers.

你觉得 AI 现在在美国是不是在某种程度上有形象问题,这可能部分是像 Anthropic 这样的公司造成的,因为他们出来说模型太强了,在推出之前必须得,你知道,被压一压、管一管?

Do you think that AI has an image problem in the US at the moment, to some extent potentially caused by companies like Anthropic who are coming out and saying that the models are too good, they have to be, you know, held back before they go out?

你觉得这会怎样帮助或者阻碍 AI 的扩散,不只是停留在那些 hyperscalers 层面,而是进一步下沉到终端消费者和企业?

And how do you think that's going to help or hinder the proliferation of AI through more than just the hyperscalers, but down to end consumers and enterprises?
M2
M215:19

其实,我不同意你刚才说的。

Actually, I disagree with what you said.

我觉得消费者对 AI 的反应其实是非常正面的,至少从他们真的会去使用这个产品这点来看,是这样,对吧?

I think consumers have had an incredibly favorable reaction to AI in terms of they actually use the product, right?

AI 的采用,你知道,最早 OpenAI 推出时,就用户数量在相对较短时间内的增长来说,是历史上任何产品里采用速度最快的。

The adoption of AI, you know, the original OpenAI rollout is the fastest adoption of any product in history in terms of number of users in a relatively short period of time.

而且很多人都在把 AI 融入他们的日常生活。

And a lot of people are integrating AI into their daily life.

无论是用它来检查自己发现的健康问题,还是用它来帮助自己工作。

Whether they're using it for health checking issues they've identified or they're using it to help them in their jobs.

我觉得,没错,人们普遍也确实会担心失业的问题。

I think, yes, there's also a fear among people generally about job loss.

而且我会说,这种情况比任何其他技术发展都更明显,至少在我有生之年里肯定是这样。

And I would say more so than any other technology development, certainly in my lifetime.

所以我觉得这里面有一种挺有意思的平衡。

So I think there's sort of this sort of interesting balance.

它确实会冲击某些类型的工作,但我越来越有信心的是,从整体经济上看,它会大幅带动就业增长。

And it will be disruptive to certain kinds of jobs, but I'm increasingly confident it's going to be a big boom economically to job growth generally.

只要人们能适应,我觉得它对就业增长会是非常非常正面的。

And as long as people adapt, I think it would be very, very positive for job growth.
M1
M116:15

你怎么看 AI 泡沫还是没泡沫这个争论?

Where do you sit on the AI bubble, no bubble debate?

这件事似乎经常会被提起,至少在英国的新闻里是这样。

This is something that seems to come up quite often, at least in the news flow in the UK.

美国那边我就不太清楚了。

I don't know about in the US.

你对现在在 AI 基础设施上的投入规模满意吗?

Are you happy with the amount of spending that's going on AI infrastructure?

对于你投资组合里的那些公司,它们是这个领域里最大的投资者之一,你有没有看到你希望看到的回报?

Are you seeing the returns you want to see from the companies in your portfolio that are some of the biggest investors in this space.

就像,帮我们理解一下,你在脑子里是怎么框定这个问题的。

Kind of like, help us think about how you kind of frame that in your mind.
M2
M216:42

所以我看的方式很不一样。像 Microsoft、Google,或者你随便说一个,比如 Meta,这些公司并不是在拿自己的资产负债表去冒险,硬闯进这段大规模资本开支的时期。

So I think very differently, you know, Microsoft, Google, you know, pick your Meta, for example, these are companies that are not risking their balance sheets to, you know, enter into this massive period of capital expenditure.

它们这么做,是在回应巨大的需求。

And they're doing so in response to enormous demand.

它们这么做,也是因为谁先到达 superintelligence,显然谁就在这个世界上拥有巨大的竞争优势。

And they're doing so in response to the first person to superintelligence obviously has an enormous competitive advantage in the world.

你不会想输掉这场竞赛。

You don't want to lose that race.

正是这场竞赛在推动竞争,也推动了这种大规模投入。

That race is what's driving competition and this sort of massive spending.

不过我会说,现在要准确判断这些投入最终会带来什么回报,还为时过早。

I would say it's too early to know precisely what the returns will be from that spending.

如果你想看实时的数据点,你可以看看 SpaceX 的招股说明书,或者 SpaceX 最近的一些公告。

If you want real-time data points, you look at the SpaceX prospectus or recent announcements from SpaceX.

Elon 以制造和硬件能力著称,他建了一个 data center。

Elon, well known for his manufacturing and hardware capability, built out a data center.

然后他把它按短期方式租出去了,显然这是他自己的选择,而且回本周期非常非常短。

And then he's rented it on a short-term basis, apparently his choice, where the payback period is very, very short.

所以,对 AI 驱动下对算力的巨大需求,以及这个工具本身的力量,我并不怀疑。

So I don't question the enormous amounts of demand for compute driven by AI and the power of the tool.

我觉得真正会亏掉很多钱的地方在于,不是每一家沾点 AI 边的私人公司,最后都能变成下一个 Anthropic。

Where I think there's going to be lots of money lost is not every private company that's AI adjacent is going to turn into the next Anthropic.

而且,你要是看这些前沿 AI 模型的话,OpenAI 原本是领跑的前沿玩家,而现在 Anthropic 看起来似乎已经暂时跑在前面了。

And, you know, if you think about the sort of frontier AI models, OpenAI was the frontier and, you know, Anthropic seems to have pulled ahead here.

Gemini 也在里面,这就是一场竞赛。

And Gemini, it's a race.

而我会说,绝大多数人其实并不需要最强的模型,来完成他们绝大多数要做的事情。

And I would say the vast majority of people won't need the best model in order to implement, you know, the vast majority of things that they do.

比起建设基础设施的人,我更担心的是 LLM 这个商业模式。

I worry more about the LLM business model than I do about the people building the infrastructure.
F1
F118:21

对,我觉得我们现在看到的是,围绕定价已经开始出现一些紧张关系了,比如说运行这些东西到底要花多少钱,拿 Uber 使用 LLM 来举例。

Yeah, I mean, I think that what we're seeing at the moment is there are tensions now around the pricing, about how much it costs to run, you know, for example, for Uber to use LLMs.

而且我读到过不少报道,说现在 Silicon Valley 的很多创业公司都在用中国的模型,因为便宜得多。

And I've read that quite a lot of Silicon Valley startups are now using China's models, which are considerably cheaper.

所以我想,问题就在这里:如果你并不需要最尖端的东西,那投资能拿到怎样的回报,自然也就成了一个问题。

So I suppose there is that question about how, if you don't need the very cutting edge, and therefore question about the returns that you can get on that investment.
M2
M218:47

这有点像,某种程度上,如果你拿它去类比互联网早期的话,当时有一家叫 Webvan 的公司,我会说它有点像今天的 Amazon,就是你可以在相对短的时间里收到各种东西,食物也能在相对短的时间里送到。

You know, it's a bit like, in some, you know, if you make the comparison to the early days of internet, there was a company called Webvan, which was, you know, I would say like Amazon today, where you get everything delivered in a relatively short period, food delivered in a relatively short period of time.

但它给消费者的补贴非常巨大,因为他们并没有一个真正跑得通的经济模型。

But the consumer subsidy was massive because they didn't have an economic model that worked.

而这套模式也只能在他们还能持续融到资金的前提下运转。

And that could only run as long as they could raise capital.

你知道,OpenAI 早期之所以能被迅速采用,很大程度上是因为那种随便用、管够的定价模式。

You know, the early OpenAI adoption was driven by the all-you-can-eat pricing.
F1
F119:14

嗯。

Hmm.
M2
M219:14

现在我们正在转向按 token 计费,尤其是面对企业客户这块。

And we're moving to a per-token-based pricing, particularly for, you know, the enterprise customer.

随着这种变化发生,再加上这些工具变得非常强大,你会看到,各个部门很快就把预算用光了。

And as that's happened, and the tools become really powerful, you know, you have seen, you know, departments use up their budgets.

如果本来就没有预算限制,那就会在很短时间里烧掉大量的钱。

And if there wasn't a budget, burn through massive amounts of money in a short period of time.

所以我觉得,企业客户接下来对价格会敏感得多。

So I think there's gonna be a lot more price sensitivity on the part of corporate customers.

还有就像你说的,正如我之前讲的,如果你只是想找一家最好的餐厅吃饭,你没必要去跑 Mistral 或者 Anthropic 的模型。

And to your point, as I said before, you don't need to be running the Mistral or Anthropic model when you're looking for the best restaurant to eat in.
F1
F119:44

没错。

Exactly.
M2
M219:46

而且到目前为止,实际上也还没出现什么真正的差异化,你基本上就是点那个最强的模型来干活。

And at this point, there's been really no differentiation, you know, you just click on the most powerful model to do stuff.

我觉得免费模型在很多使用场景里已经够用了。

And I think the free models are, in many use cases, good enough.
F1
F119:56

对,我明白你的意思了,就是到目前为止,他们其实还不算特别依赖外部融资。

Yeah, I see what you say, that until now they haven't been sort of very indebted.

这些公司一直在产生大量现金流,然后把这些现金流投入到他们正在做的这些投资里。

These companies, they've been generating a lot of cash flow, which they've put towards this investment that they're doing.

但现在确实感觉这个情况正在变化。

But it does feel that that now is changing.

比如说,Alphabet 就宣布了一笔大规模股权融资。

You know, you've got Alphabet, for example, announcing a big equity raise.

我们现在也能看到,大概有百分之十到百分之十二的 investment grade 和 high yield,现在某种程度上都和 tech AI-

We can see now about ten to twelve percent in investment grade and high yield is now sort of tech AI-

相关。

related.

那对于这些公司一轮接一轮继续做股权融资或者债务融资,你的容忍度会有多高?

So how tolerant are you going to be for continued rounds of equity or debt raising by these companies?
M2
M220:29

如果你看 Alphabet 这个例子,或者其他那些大型公司,你知道,这些公司在历史上基本都是没有负债的企业,资产负债表上现金很多。

If you look at the Alphabet example or the other sort of big companies, you know, these have been historically debt-free enterprises, a lot of cash on the balance sheet.

而且如果真要说的话,过去这些年它们一直还在回购股票,把钱还给股东,因为它们对那些多出来的资本并没有足够好的用途。

And if anything, they've been buying back stock over time, returning to shareholders because they didn't have an adequate use for that excess capital.

而这一点在过去六个月里当然已经反过来了。

And that's in the last 6 months certainly reverted.

就是很多这样的回购计划都停了,而且现在资本其实有非常好的用途。

Where a lot of those buyback programs have stopped and they're actually, you know, very good use for capital.

你知道,Amazon 还有其他公司,现在都在做资本投资,花个一千五百亿美元甚至更多。

You know, Amazon and others are spending whatever, $150 billion or more on capital investment.

但那是资本投入。

But it's capital.

我对这个问题的看法和另一种情况很不一样。比如说,如果你有一家企业,我随便举个例子,一个零售商,为了保持竞争力,不得不把所有门店都升级,因此它现在必须比过去多花很多 CapEx。

I think very differently about, you know, if you have a business, I don't know, a retailer that needs to upgrade all of its stores in order to be competitive, and therefore they have to spend a huge amount on CapEx versus previously.

我会把那种看成是 maintenance CapEx。

I think of that as maintenance CapEx.

那其实会削弱公司的价值。

And that's a diminution in the value of the company.

而这里的支出不是 maintenance CapEx。

This is not maintenance CapEx spending.

这是一种投资。比如说,扩建 Azure cloud 来满足需求,购买服务器、房地产,或者任何为了提供算力所必需的东西,而市场愿意为这些算力支付一个价格,使他们能够获得远远高于、显著高于其资本成本的回报。

This is investment into, for example, building the Azure cloud out to meet the demand, you know, buying servers, real estate, whatever is needed in order to be able to deliver compute that the market's prepared to pay a price that enables them to earn a return well in excess of, vastly in excess of their cost of capital.

Warren Buffett 一直都说,世界上最好的生意,就是那种资本回报率很高,而且你还能把这些资本以很高的回报率继续投入到新项目里的生意。

Warren Buffett always said the best business in the world is one that generates high returns on capital, and you can reinvest that capital at high rates on new projects.

这显然就是这种情况的例子。

These are clearly examples of that.
M1
M121:50

你觉得,市场叙事要转向这种想法——也就是实际上有地方可以投放你的资本是一件好事——需要什么条件?

What do you think it's going to take for the market narrative to shift towards this idea that actually having somewhere to put your capital is a positive?

Meta 大概就是个很典型的例子,每次它公布新的 CapEx 水平,市场往往都会惩罚它。

And Meta is probably a perfect example of this, where it tends to get punished in the market every time it comes out with a new CapEx rate.

如果他们下个季度出来说要削减 CapEx,市场大概率反而会奖励它。

If they were to come out next quarter and cut CapEx, it'd probably be rewarded from the market.

但按照你的这套逻辑,这其实是一种错误的看法。

But from your kind of logic, that's the wrong way to think about it.

你觉得拐点会在什么时候出现,也就是市场什么时候才会跟上这种思路?

When do you think is going to be the inflection point where the markets catch up with that way of thinking?
M2
M222:18

当收入和盈利增长能体现出这些资本支出所带来的高回报时,大家就会对此感觉相当不错。

When revenues and earnings growth reflects the high returns that are being earned on the capital spend, people are going to feel pretty good about it.

但如果突然之间,全世界每家公司为了增长都在大量借钱、大量发股,那这个叙事就不可能长期站得住脚。

And if all of a sudden every company in the world is borrowing a ton of money and issuing a ton of equity in order to grow, that's not going to be a good narrative on a sustained basis.

不过我不认为我们会走到那一步。

But I don't think we're going to get there.

当然,我也可能错。

I could be wrong.

我倒是挺惊讶 Alphabet 发了八百亿美元的股票。

I was surprised that Alphabet issued $80 billion of stock.

他们其实并不真的需要这么做,但我觉得他们当时的思路是——你看,我们之前是 Alphabet 的大股东。

They didn't really need to, but I think the way they thought about it— look, we were a big shareholder of Alphabet.

那对我们来说是一笔非常成功的投资。

It was a very successful investment for us.

我记得我们在三到四年的时间里,回报大概接近投资额的三点五到四倍。

I think we made something approaching 3.5, 4 times our investment over a 3, 4-year period of time.

但我们在这个季度退出了,我们这么做并不是因为我们不觉得 Alphabet 是一家很棒的公司,而是因为它的估值到了三十一、三十二倍 earnings,我们觉得那笔资本可以被配置到更好的地方。

But we exited in this quarter, and we did so not because we don't think Alphabet's a great business, but it got to a valuation, 31, 32 times earnings, where we felt that capital could be better deployed.
F1
F123:04

对。

Yeah.
M2
M223:04

而且我们实际上还买入了 Microsoft 的一个很大的持仓。

And we actually bought a big stake in Microsoft.

我觉得 Alphabet 看了看自己的股价和估值,然后他们说,你看,现在对我们来说是个不错的发股时点。

I think Alphabet looked at their share price and their valuation and they said, look, this is a decent place for us to be issuing.
F1
F123:11

对,尤其是如果你面前正有一大波新资金要涌入市场,那你还不如早点进去,宜早不宜迟。

Yeah, particularly if you're facing a wall of new money coming to the market, you might as well get in there quicker, sooner rather than later.
M2
M223:20

从竞争角度看,当你的所有竞争对手都得去融资、而且接下来还会融,比如 Anthropic 和其他公司要上市,那为什么不通过多发一点股票,把公开市场上的一些泡沫拿走,这样也能让你的竞争对手更难受一点。

Competitively, while all your competitors are having to raise, are going to, you know, you have Anthropic and others coming public, why not take away some of the froth from the public markets by issuing more stock so you can make it a little harder on your competition.

市场会支持这种做法的。

There will be support for that.

我的意思是,看到像 Google 这样的公司下调他们的 token 价格,我真的很震惊,这会让 OpenAI 和 Anthropic 的日子变得非常难过。

I mean, it's shocking to me to see, you know, the Googles of the world reduce their token prices, which is going to make life really hard for OpenAI and Anthropic.
F1
F123:43

对。

Yeah.
M2
M223:43

他们之所以能这么做,是因为如果你看看竞争格局,Anthropic,尤其是 OpenAI,直到非常最近——Anthropic 显然也是——一直都在烧掉巨额现金。

They can do so because, you know, if you think about the competitive landscape, you've got Anthropic and in particular OpenAI up until very recently—for Anthropic apparently—but they've been burning huge amounts of cash.

这是一种高风险的商业模式。

That's a risky business model.

因为你依赖资本市场持续给你提供资本。

You're reliant on the capital markets giving you capital.

这也就是为什么这些公司现在要上市。

This is why these companies are now going public.

如果我是 Alphabet,而且我想这么做的话,凭借他们资产负债表上的优势,他们完全可以大幅下调 token 定价,真正打乱市场,让竞争对手的处境变得困难得多。

If I were Alphabet, and I wanted to, you know, with their balance sheet advantage, you know, they could reduce token pricing significantly and really disrupt, make it much more challenging for the competition.
F1
F124:11

我会说,这对 OpenAI 来说可能比对 Anthropic 还更具挑战,因为 Anthropic 看起来更明显地在往 enterprise 端走,想在那边扩张,我觉得这也帮助他们某种程度上超过了 OpenAI 现在的位置。

I would say it's probably even more challenging for OpenAI than Anthropic, who seem to be much more going towards the enterprise side and trying to grow there, which I think has helped them sort of overtake where OpenAI has got to.

而且 Gemini 对 Google 来说其实也非常成功,对吧,至少在捍卫它们市场地位这件事上是这样。

And Gemini has actually been very successful for Google, hasn't it, in terms of defending their market position.

说到 Microsoft,今年年初我们看到那种软件股的抛售。

On Microsoft, we had that sort of software sell-down at the beginning of this year.

你怎么看?

How do you feel?

你觉得大家对软件的担忧是不是有点过头了?

Do you think that's been overdone in terms of people's fears about software?

还是说,你觉得 AI 实际上会帮助软件更广泛地普及,所以你会更愿意用 Microsoft,而不是自己做一个 Excel 之类的版本?

Or do you think AI will actually help the dissemination of software, and you will want to use Microsoft rather than making your own version of Excel or whatever?
M2
M224:47

对,我觉得这得一家公司一家公司地分析。

Yeah, I think it's a company-by-company analysis.

我觉得有些软件公司,过去靠向每个用户收很高的费用,赚到了某种意外之财;而在一个企业自己非常有动力去替换掉那种每年一万、两万、三万美元、甚至更贵的软件产品,转而用内部自己开发的东西的世界里,这些公司会面临问题。

I think some software companies that have made sort of windfall profits charging a lot to each user, in a world where companies themselves are highly incentivized to replace an expensive $10,000, $20,000, $30,000 per annum or more software product with something they can develop internally.

你看,典型的 Microsoft 365 企业用户,一个席位大概就付二十美元。

You know, the typical Microsoft 365 enterprise user, they're paying $20 a seat.

这买到的是一整套产品;如果你分别通过 Zoom、Gmail 这些渠道去买,等等吧,你得付五十美元甚至更多。

That's for a basket of products that if you were to buy them via Zoom, via Gmail, you know, etc., you'd have to pay $50 or more.

他们的用户数,我觉得,已经接近五亿了。

They've got, I think, something approaching half a billion users.

而且我觉得 Microsoft 在 Copilot 这件事上一直非常聪明。

And I think Microsoft's been, you know, very smart about Copilot.

我觉得 Microsoft 会有点像 Uber,变成大家在工作场景里接入 AI 的那个平台。

And I think Microsoft becomes a bit like Uber, the platform by which people access AI in the workplace.

而现在,一个 CIO 最不想做的事,就是自己去做一套 Word、Excel、Zoom,或者其他这些产品的版本。

And the last thing a, you know, CIO wants to do now is create his own version of Word or Excel or Zoom or any of these other products.

这还没算上你再去考虑网络安全层面的影响。

That's before you think through the cybersecurity implications.

Microsoft 这个平台以非常低的人均成本就带来了很多价值,然后在这个基础上,他们再去收取额外的增量费用。

A lot gets brought to the table with the Microsoft platform at a very small per-user cost, and then that becomes the base from which they charge incremental dollars.

你知道,我觉得 Microsoft 未来对企业非常重要的一个角色是:你会希望有一种技术,能把你指向那个为了完成目标所需要的最低成本模型。

You know, one of the really important roles Microsoft, I think, is going to play for the enterprise is you want technology that can point you to the lowest-cost model that you need to accomplish the objective.

比如说,做个电子表格,你没必要非得用最前沿的模型。

You don't have to be at, you know, the cutting edge building out a spreadsheet.

所以 Microsoft 会成为企业接入 AI 的入口,而且他们是以一种让 CIO 感到相当放心的方式来做这件事。

And so Microsoft becomes the gateway to AI in the enterprise, and they do it in a manner that, you know, the CIO feels quite comfortable with.
M1
M126:23

Bill,你是不是属于这样一种看法:语言模型整体上都会逐步商品化,而这对你的投资组合公司总体来说是利好的,因为像你说的那样,你可以全面提供不同模型的不同版本?

Bill, do you sit in this camp of language model commoditization across the board, and that being broadly positive for your portfolio companies because you can, as you say, offer different versions of different models across the board?

当然会有一些差异化,但也会出现更便宜的选择。

There will be some differentiation, but there's going to be cheaper options.

而关键就在于这种把用户导入进来的思路。

And it's that idea of funneling the user.

你知道,现在你在不同的时候大致还能自己挑想用哪个模型。

You know, at the moment you can kind of pick which model you want to use at different times.

但总体来说,我觉得大多数人其实并不想这么做。

But, you know, broadly speaking, I think that most people don't want to do that.

他们基本上只是想直接用自动切换功能,让系统自动切到完成他们任务所需成本最低的那个。

They kind of just want to be able to use the auto-switch and have it switch through to the lowest cost to do their task.

对于那些你认为会成为 AI 胜出的软件公司来说,这整体上算是净利好吗?

Is that a net benefit for the software companies that you think are going to be AI winners?
M2
M226:58

我觉得这取决于商业模式。

I think it depends on the business model.

还是那句话,如果你是那种成本很高、比较小众的软件产品,我觉得你的风险会大得多;相比之下,如果你是一家拥有接近五亿用户、成本很低、而且已经深度嵌入企业体系的公司,那情况就不一样。

Again, if you're a high-cost, kind of nichey-type software product, I think you're at much greater risk than if you're a company approaching half a billion users at a very low cost that's built into the enterprise.

我的意思是,还是得一家公司一家公司地看。

I mean, again, it's a company-by-company analysis.

但我确实觉得,像 Microsoft 这样的公司会是赢家。

But I do think the Microsofts of the world are going to be winners.
F1
F127:19

好,太好了。

Okay, great.

那我们稍微换个话题,你刚才提到了一些 Berkshire Hathaway。

So if we shift a little, you talked a bit about Berkshire Hathaway.

我们能不能先从 Howard Hughes 最基础的情况讲起?

Can we really start with the basics on Howard Hughes?

因为这算是一个战略上的调整,我们的散户投资者可能还没有完全理解。

Because this is a bit of a strategic shift that our retail investors might not fully understand yet.

所以你能不能讲讲买入 Howard Hughes 股份背后的动机,然后再聊聊 Vantage,也跟我们多讲一点这件事、你们的方法,以及你们希望在这里打造什么?

And so can you talk about the motivations behind buying the stake in Howard Hughes and then touch on Vantage and just tell us more about that and the approach and what you're hoping to build here?
M2
M227:45

当然。

Sure.

Howard Hughes,如果你看我们的投资组合,它其实不算是那种匹配的标的,因为我们持有的绝大多数公司,都是大盘股、超大盘股,非常高质量、能够长期持续、而且资本回报率非常高的公司。

So Howard Hughes, if you look at our portfolio, it's not a fit in the sense that the vast majority of the companies we own are large-cap, mega-cap, very high-quality, durable companies that earn very high returns on capital.

那才是我们的商业模式,而不是去持有那种相对净资产价值有很大折价交易的小盘房地产公司。

That's the business model, as opposed to small-cap real estate companies trading at big discounts to net asset value.

我们之所以持有这笔股份,更多是历史原因,也就是从我们买入 Howard Hughes 的股份到现在,我想已经差不多十七年了。

The reason for the existence of this stake is one of history, which is it's now been almost, I guess, 17 years since we took a stake in Howard Hughes.

Howard Hughes,当时是因为金融危机,失去了获得房地产债务融资的能力。

Howard Hughes, as a result of the financial crisis, lost the ability to access real estate debt financing.

不,不是 Howard Hughes,是一家叫 General Growth 的公司。

It's not Howard Hughes, a company called General Growth.

抱歉。

Excuse me.

这是我们当时买入了百分之二十五股权的一家公司。在金融危机期间,因为资本市场某种程度上的停摆,这家公司的股价跌了百分之九十九点五。

This was a company we bought 25% of the equity in, a business that was— stock was down 99.5% during the financial crisis due to the kind of shutdown in the capital markets.

我们主导了一个破产重组方案,并买下了这家公司百分之二十五的股份。

We led a bankruptcy restructuring plan. We bought 25% of the company.

后来它成了我们有史以来最成功的一笔股票投资。

It turned into our most successful equity investment ever.

但这个破产重组方案里有一部分内容,是把 General Growth 里面那些市场不喜欢的资产剥离出去。

But part of the bankruptcy restructuring plan was getting rid of assets inside of General Growth the market didn't like.

我们把所有这些资产都放进了现在这家叫 Howard Hughes 的公司里。

We put all those assets in this company now called Howard Hughes.

之后十五年里,我们一直在慢慢梳理这些资产,也经历了几届管理团队。

And for 15 years, we kind of sorted through those assets, went through a couple of management teams.

但归根结底,这里面有一项我们非常喜欢、可以用几十年维度来看待的核心业务,就是拥有这些基本上可以说是小型城市的资产。只是这种业务,尤其在 Wall Street 越来越短线化的环境里,并不会特别吸引人,而且它本身也相当复杂,需要花时间去做尽职调查。

But ultimately, a core business that we really liked on a multi-decade basis, a business of owning basically these small cities, not one that particularly in a shorter and shorter-term Wall Street environment is going to be appealing to people and a business that is quite complicated, that takes time to do due diligence.

所以 Pershing Square 这家管理公司,也就是有些人说的 GP,又对这家公司追加了一笔投资。

So the management company, the GP, as some people call it, at Pershing Square made an incremental investment in the company.

所以我们通过 Pershing Square Holdings 持有 Howard Hughes 大约百分之三十的股份,再通过我们的其他基金持有这家公司另外大约百分之二。

So we own a stake through Pershing Square Holdings of about 30% of Howard Hughes and through our other funds, another 2% of the company.

我们又通过管理公司买下了这家公司百分之十五的股份,这样我们合计就持有了这家公司百分之四十七。

We bought 15% of the company with the management company, giving us 47% of the company.

而且我们正在推动这项业务进行转型。

And we're taking the business through a transformation.

所以我们要把它从一家纯粹持有这类小型城市资产的房地产公司,转型成一家多元化控股公司,而这个转型会由对一家叫 Vantage 的公司的投资来带动。

So we're going from a pure-play real estate owner of these sort of small cities to a diversified holding company led by an investment in a company called Vantage.

Vantage 是一家专业保险和再保险公司。

Vantage is a specialty insurer and reinsurer.

其实我们上周刚完成这笔交易,而整个商业计划,就是借鉴一下 Buffett 先生做过的事情。

We closed on this transaction actually last week, and the business plan is to take a page from what Mr. Buffett has done.

Buffett 先生一开始拥有的是一家处境非常艰难的纺织企业,当时纺织品生产正越来越多地转移到亚洲。

Mr. Buffett started with a very challenged company in the textile industry as textiles were increasingly being manufactured in Asia.

后来随着时间推移,他清算了那个核心业务,然后投资了保险、银行,还有糖果公司。

And over time, liquidated that core business, invested in insurance, in a bank, candy company.

这个故事大家都知道,但绝大部分价值,其实是通过保险公司内部资本的复利积累创造出来的。Buffett 把保险公司的资产——当时这些资产几乎全都投在债券里——以及保险业务产生的全部 float,拿去投资 Treasury;再把多出来的盈余,也就是股东权益,投到普通股里。

And everyone knows the story, but the vast majority of the value was created by the compounding of capital inside of an insurer where Buffett took the assets of the insurance company, which were almost entirely invested in bonds, and took 100% of the float generated by insurance, invested in Treasuries, and took the surplus or the equity, invested in common stocks.

而且他很会选股票。

And he was a good stock picker.

这就推动了 Berkshire 的价值增长。

That drove the value of Berkshire.

我们现在做的事情非常相似。

We're doing something quite similar.

所以,借助 Vantage,我们会把这家保险公司的 float 拿来投到短期 Treasuries 里,基本上等于是不承担什么风险。

So, with Vantage, we're taking the float of the insurer, investing in short-term Treasuries where we take effectively no risk.

然后 surplus,我们会按照 Pershing Square 的策略投到 common stocks 里。

And the surplus, we're investing in common stocks according to the Pershing Square strategy.

如果我们做得好,管理团队在保险业务上承担审慎风险这件事也做得好,那我们就在资产负债表的负债端赚到利润。

If we do a good job, the management team does a good job taking on prudent risks in insurance, we'll make a profit on the liability side of the balance sheet.

同时,我们也会按我们预期的方式去投资,在资产端获得很高的回报。

And we'll earn a high return on the assets, investing it the way we expect.

这样一来,希望就能在很长一段时间里,以一个很高的速度让这家公司的 equity 持续复利增长。

That will compound the equity of this company at hopefully a high rate over a long period of time.

所以这就是一门生意。

So it's a business.

你知道,Buffett 做得特别重要的一件事,就是他在保险业务上拿到了很高的回报,而且随着时间推移,他几乎没有增发股票,或者只增发了很少的股票。

You know, one of the things that Buffett did so important is he earned high returns in his insurance operation and he issued no stock or very little stock over time.

所以,这家公司的每股价值才能在六十年的时间里,以百分之二十的速度持续复利增长。

And so the per-share value of the company was able to compound over 60 years at 20%.

这可以说是史上最伟大的投资故事之一。

And that's, you know, one of the great investment stories of all time.
F1
F131:21

你为什么觉得,其他保险公司连自己都不会想着在 equity 投资上更大胆一点呢?

Why don't you think other insurance companies even themselves would think to sort of be more adventurous with their equity?
M2
M231:28

因为最优秀的投资人不会选择去保险公司工作。

Because the best investors don't choose to take jobs at insurance companies.

如果你是个特别有天赋的投资人,你会去 hedge fund 工作,或者你会去更传统的 long-only asset manager 工作。

If you're a super talented investor, you're going to go work for a hedge fund, or you're going to go work for a more traditional long-only asset manager.

你不会选择去一家保险公司上班,光是它给出的薪酬水平就不具吸引力。

You're not going to sign up to go work for an insurance company, just the pay scales that are offered.

在 Harvard Business School,基本没人真的会讨论毕业后去保险公司工作。

At Harvard Business School, no one's really talking about going to work for an insurance company.

这不是一件会被认为很 sexy 的事,我觉得这么形容最贴切。

It's something that's not thought of as sexy, I guess is the best way to describe it.

Buffett 带来的关键在于,他自己就是一位伟大的投资人,同时又拥有一家保险公司,所以他可以免费管理这个投资组合,而不是把它交给 hedge fund 去做。

What Buffett brought to the table is that he was a great investor that owned an insurance company and he could manage that portfolio for free as opposed to putting it out to a hedge fund.

我们也是在免费管理 Vantage 的这个投资组合。

We're managing that Vantage portfolio for free.

Vantage 将能够零成本享受到 Pershing Square 的投资能力。

Vantage will have the advantage of Pershing Square's investment acumen at no cost.

而这正是它具备竞争优势的原因。

And that's what gives it a competitive advantage.

所以我们也不需要专门为 Vantage 招人来管理这个投资组合。

And therefore we don't need to recruit people to Vantage to manage the portfolio.

我们是以第三方的方式在做这件事。

We're doing it on a third-party basis.

我们持有这家公司非常大比例的股份,所以让 Vantage 取得巨大的成功,完全符合我们的经济利益。

We own so much of the company, it's in our economic interest to make Vantage highly successful.
M1
M132:24

太棒了,谢谢你,Bill。

Awesome, thanks, Bill.

我们的时间到了,非常感谢你今天来做客。

We are out of time, so thank you very much for joining us today.

刚才的内容真的很有启发。

That was really insightful.

谢谢你,Bill。

Thank you, Bill.
F1
F132:32

谢谢你,真的太精彩了。

Thank you, that's been wonderful.

谢谢你们给我这个机会。

Thank you for the opportunity.
M2
M232:36

祝你今天愉快。

Have a great day.
F1
F132:36

谢谢。

Thank you.

这周就到这里。

That's all for this week.

本期录制于二零二六年六月十一日,所有信息在录制当时都是准确的。

This session was recorded on June 11th, 2026, and all information was correct at the time of recording.

下周,Helen Morrissey 和 Clare Stinton 会回来。

Next week, Helen Morrissey and Clare Stinton will be back.
M1
M132:51

这档播客里的内容都不构成个人建议,如果你不确定什么适合你,应该去寻求专业建议。

Nothing in this podcast is personal advice, and you should seek advice if you're unsure what's right for you.
F1
F132:55

投资的价值会上涨也会下跌,所以你拿回来的钱可能会少于你的投入,而且过往表现并不能作为未来表现的参考。

Investments rise and fall in value, so you could get back less than you invest, and past performance is not a guide to the future.
M1
M133:03

这也不是在建议你买入、卖出,或者继续持有我们讨论过的任何投资或公司。

This isn't a recommendation to buy, sell, or hold any of the investments or companies we've discussed.

顺便说明一下,我个人持有 Microsoft 和 Meta 的股票。

And for the record, I have personal holdings in Microsoft and Meta.
F1
F133:13

Pershing Square Investment Trust 可以使用 derivatives,也可以使用 gearing。

The Pershing Square Investment Trust can use derivatives and can use gearing.

而这确实会增加持有这项投资的潜在风险。

And this does increase the potential risk of the holding.

这里提到的个股都不是投资建议,而且也要记住,分散投资永远是你投资组合的朋友,尤其是在市场波动的时候。

The individual stocks mentioned aren't recommendations, and it's worth remembering that diversification is always your portfolio's friend, particularly when the markets are volatile.

最后,非常感谢我们的制作人 Elizabeth Hodgson。

Finally, thank you very much to our producer, Elizabeth Hodgson.
M1
M133:34

也非常感谢所有观看和收听的朋友。

And thank you very much to everyone watching and listening.
F1
F133:37

刚才你听到的是 Bill Ackman、Matt Britzman,还有我,Anna MacDonald。

That was Bill Ackman, Matt Britzman, and me, Anna MacDonald.

再见。

Goodbye.
M1
M133:41

再见。

Goodbye.