高利率可能刺破AI泡沫——这里有你该持有的替代资产

真的没有什么比利率上升更能戳破投机泡沫了。

There's really nothing like rising interest rates to burst a speculative bubble.

现在一切都依赖于人工智能泡沫。

Everything is now dependent on the AI Bubble.

你知道，如果我不想持有国债，不想持有公司债券，那么除了这些金融资产之外，你唯一的选择就是实物资产。

You know, if I if I don't wanna own treasuries, if I don't wanna own corporate bonds, your really only other choice outside of these kind of financial assets is real assets.

你好，欢迎来到Wealthion。

Hello, welcome to Wealthion.

我是Maggie Lake。

I'm Maggie Lake.

今天和我一起讨论黄金、债券以及如何为未来做好准备的是Jesse Felder，他是The Felder Report的创始人。

Joining me today to discuss gold, bonds, and how to prepare for what's ahead is Jesse Felder, founder of The Felder Report.

嘿，Jesse。

Hey, Jesse.

很高兴再次邀请你上节目。

Great to have you on again.

很高兴见到你，Maggie。

Great to see you, Maggie.

所以如果我们正处于新一轮的加息周期，我想首先问一下，你相信这会是一系列加息吗？

So if we are on a new rate hiking cycle, I guess, first of all, do you believe it's going to be a series of hikes?

因为对此存在一些分歧，有些人认为美联储是在需求放缓的情况下加息。

Because there's some disagreement about that, and some people think the Fed's hiking into what will be a demand slowdown.

你知道，这种情况已经发生了，他们有点晚了。

You know, that's already happening, and they're kinda late to the game.

所以我们以为会有所有这些加息，但实际上可能不会。

And so we think there's gonna be all these rate hikes, but there's not really gonna be.

你觉得……

Do you...

你听起来像是认为会有加息的那一派。

You sound like you're in the camp that, no, there's gonna be.

是的，我确实认为我们正处于一个新的加息周期。

Yeah, I do think we're this is a new rate hike cycle.

我觉得是因为你听沃什说的，他呼应了前纽约联储主席比尔·达德利上周写的内容，那就是收紧政策的理由很少像今天这么明确。

I think it's because when you listen to Warsh, he echoed something that Bill Dudley, former head of the New York Fed wrote last week, which is The rationale for tightening has rarely been so clear cut as it is today.

美联储只是在任务的一方面没达标，对吧？

The Fed's only missing on one side of the mandate, right?

显然，失业率很低，但自2021年3月以来，他们在通胀方面一直没达标。

Obviously, unemployment is really low, but they've missed on the inflation side really since March 2021.

核心PCE一直高于目标，而且持续这么长时间，现在通胀又可能在短期内再次抬头。

Core PC has been running above target and missing that long, and now inflation is threatening to again in the short term.

所以我觉得现在人们可能没有意识到正在积累的通胀压力。

So I think people are mainly right now, maybe don't appreciate these inflationary dynamics that are building right now.

名义GDP的增速，实际上自1990年以来只有六个季度能跟第二季度一样高。

Nominal GDP is running at its literally, there's only six quarters since 1990 that where nominal GDP has been running as high as it did in the second quarter.

那是在2005年，就在房地产泡沫破裂前，还有1999年，就在互联网泡沫破裂前。

It was in 2005, right before the housing bubble burst, and in 1999, right before the Dotcom bubble burst.

所以经济实际上是在高速运转。

So the economy is literally running.

我们现在经济运行的热度是三十五年来最高的。

We're running the economy hotter than we have at any point in thirty five years.

金融条件处于历史最宽松水平，同时经济热度也是三十年来最高的。

Financial conditions are as loose as they've ever been while we're running the economy hotter than at any point in thirty years.

而且我们现在正面临有记录以来最严重的燃料危机之一，柴油价格创下历史新高。

And we now have one of the greatest fuel crises on record, diesel prices hitting record highs.

进口价格，你知道的，很多是受人工智能相关的东西推动，但进口价格的上涨速度是几十年来最快的。

Import prices, you know, a lot of this driven by the AI stuff, but import prices are rising at the fastest pace in decades.

我们看到的生产者价格指数正在急剧上升。

Producer prices we're seeing are literally ripping higher.

你看任何制造业调查，都在说价格压力又在积聚。

You see any of the kind of manufacturing surveys, they're all talking about pricing pressures building again.

食品价格现在又开始上涨，而且，你知道，如果出现新的厄尔尼诺现象，明年食品价格可能会真的飙升。

Food prices are now starting to turn higher again and threaten, you know, with with a new El Nino, we could have a we could have a real spike in food prices next year.

所以我们说的是可能明年圣诞节前后。

So we're talking about potentially next Christmas.

这真的很重要，对吧，因为你借的所有东西——房贷、信用卡、车贷、还有学生贷款。

And and this matters so much, right, because it's, everything you borrow, your mortgage, your credit cards, your car, your, you know, student loans.

这么多都跟这个挂钩。

So much of this is pegged off this.

我们说的是，大概7%左右吧，可能十年期国债收益率在这个水平才算合适？

We're talking about, what, seven in the seven handle maybe for, like, that's where it feels like appropriate rate for tenure is?

我的意思是，你知道，历史上在加息周期中，十年期收益率至少会上升100个基点。

I mean, you know, and the the tenure historically during a rate hike cycle rises at least a 100 basis points.

这是平均水平。

That's on average.

在历史利率周期中，低端是50个基点，高端是200多个基点。

50 basis points at the low end and 200 plus basis points at the high end through historical rate cycles.

所以这告诉你，这个周期开始时，美联储的十年期利率是5%。

So that tells you this cycle started with the Fed, the ten year rate at 5%.

所以平均来看，可能会涨到6%。

So on average, it would it would probably go to six.

但我觉得你可以说，可能会到六点五甚至七。

But I think you could argue that, you know, you could go to six and a half, seven.

而且，你知道，这只是基于历史上加息周期中的表现。

And, know, that that's just based on what it historically does through a rate hike cycle.

上次我们聊天的时候，你对债券市场的一些信号还挺担心的。

So last time we spoke, you were pretty concerned about some of the signals coming from the bond market.

现在十年期国债收益率一直在5%左右徘徊，还经常突破5%，就在5%上方一点。

Now the ten year has been hovering right around 5%, climbing above it on a pretty consistent basis, just above it.

实际上，全球债券收益率，不只是美国的，都在往上走。

The global bond yields, in fact, not just US, have been pushing higher.

这让你对全球经济形势有什么看法？

What does that tell you about what's happening in the global economy?

有意思的是，关于什么推动了长期收益率上升，大家讨论得很多。

Well, it's funny, there's been so much talk about what's driving, you know, long term yields higher.

你知道，因素有很多。周末的Dealbook发了个东西，列出了推动长期利率上升的十个因素。

You know, there's so many different factors, I think, the weekend deal book, put out a thing, it's like, you know, these are the 10 factors driving long term rates higher.

所以看起来好像所有因素都在推动长期利率走高。

So it pretty much seems like everything is working to push long term interest rates higher.

我觉得，对我来说，用奥卡姆剃刀式的框架来理解很有道理，最简单的解释通常是最准确的。

I think, you know, it makes a lot of sense to me to kind of apply an Occam's razor type of framework, you know, the simplest explanation is usually the most accurate.

而且，在市场上，决定价格的是供需关系。

And, you know, in markets, what drives prices is supply and demand.

所以我觉得我们看到的只是国债供应过剩，同时企业债也供应过剩，都是为了给AI热潮融资。

And so I think what we're just seeing is an oversupply of Treasury debt coming to the market amid an oversupply of corporate debt at the same time to fund the AI boom.

所以供应超过了需求，这就推高了收益率。

So supplies are just outpacing demand, and that's driving yields higher.

但我觉得，考虑到上周的美联储会议和正在积累的通胀压力，我认为投资者现在可能没有充分意识到的是，美联储真的开始了一个新的加息周期。

But I think with the Fed meeting last week and the inflationary pressures that are building, I think what what investors maybe don't appreciate well enough right now is the fact that the Fed is really starting on a new rate hiking cycle.

这个周期可能会比投资者目前预期的持续更久。

It's gonna probably have to go on longer than investors currently appreciate.

我的意思是，我们从今年预期降息，到现在，哇，他们在加息，我认为随着人们开始理解是什么在推动通胀，以及美联储的耐心正在耗尽，市场的这种鹰派转变将会改变。

I mean, we've gone from anticipating rate cuts this year to now, oh wow, they're hiking rates, and I think that's that hawkish shift in markets is going to change as people begin to understand what's driving inflation and, you know, the Fed's willingness to be patient is is running out.

有没有什么……

Is there anything...

因为我们都想知道从投资角度来看该怎么做，对吧？

Because we all want to figure out what we do from an investment point of view, right?

在这种环境下，你有没有看好的东西，或者你会超配哪些资产类别，哪些会表现好？

Is is there anything you're bullish on or what asset classes do you overweight or perform well in that kind of environment?

你描述的是一种高通胀、经济过热的局面。

You're describing a, you know, a run it hot, high inflationary regime.

嗯，我觉得我们可能开始看到资金转向更防御性的板块，比如必需消费品、医疗保健。

Well, I think, you know, we probably are are starting to see a shift towards more defensive type of names, consumer staples, health care.

这些类型的股票实际上已经开始有所起色了。

These types of things have actually started to show some life.

你知道，这对市场来说显然是个问题，因为非必需消费品板块今年已经是市场上表现最差的板块了。

You know, this is obviously a problem for markets because the consumer discretionary sector is already the worst performing sector in the market this year.

这说明消费者已经在挣扎了，对吧？

That's telling you consumers are already struggling, right?

所以如果真的进入加息周期，那对消费者来说压力就更大了。

So if we do get a rate hike cycle, that's going to put even more pressure on consumers.

我觉得从长期来看，如果长期债券收益率要升到5%到6%，那你得开始担心一件事了。上周《华尔街日报》有篇报道说，纽约联储最近几周一直在跟各大银行讨论基差交易，就是那些对冲基金在现金国债和期货市场价格之间做套利，赚取很小的价差。

I think in terms of long, you know, owning long term bonds here, if they are going to move five to six, one of the things you have to start worrying about, and The Wall Street Journal put out a report about this last week, that the New York Fed has been talking to a lot of the major banks over the last couple of weeks about the basis trade, about all these hedge funds that are trying to arbitrage, you know, small price differentials between cash, treasuries and futures market prices.

这个交易杠杆特别高，一旦利率上升，尤其是债券市场波动加剧，这个交易就会承压。

And it's a massively leveraged trade that is, you know, when rates go up and especially bond market volatility rises, that trade comes under pressure.

所以这也是我担心的原因，从5%到6%这个过程可能不会那么顺利，说不定国债市场会出点意外。

So, you know, that's why I worry that, you know, it might not be just a smooth ride from five to six, it could be, we could see an accident in the Treasury market.

与此同时，私募资本那边已经有不少问题在酝酿了，对吧？

And then at the same time, we have the problems that are already brewing in private capital, right?

私募股权基金手里一堆僵尸企业，想甩都甩不掉，但我觉得私募信贷问题更大。惠誉说私募信贷的违约率实际上已经到6.5%了。

We have private equity funds that are dealing with a bunch of zombie companies they can't offload, but I think private credit is an even bigger problem, where, you know, Fitch is saying that literally defaults are already, you know, hitting 6.5% in private credit.

PIMCO提到，如果把实物支付贷款也算进去，私募信贷的违约率已经接近20%，差不多占贷款组合的两成了。

PIMCO talks about if you include payment in kind loans, defaults in private credit are already 20, almost 20% of loan portfolios.

所以利率上升对私募信贷的压力就更大了。

So rising interest rates put even more pressure on private credit.

但我觉得，可能对你的观众来说，最重要的一个方面是，在利率上升的环境下，从历史上看，价值股的表现会比成长股好，对吧？

But I think for maybe to your viewers, the most important kind of aspect is this in an environment of rising interest rates, historically, stocks do better than growth stocks, right?

这是因为当利率上升时，相当于你提高了对未来收益进行折现的折现率。

That's because when when interest rates rise, that's kind of you're raising your discount rate on those earnings that you're discounting from well into the future.

所以，折现率上升对成长股估值的影响远大于对价值股的影响。

So rising discount rates are much more of an impact on the valuation of growth stocks than they do on value stocks.

因此，关注价值股，我认为是一个非常重要的动态。

So focusing on value is, I think, a really important dynamic.

而且，实际上，我们唯一看到在利率上升期间成长股仍然试图引领市场走高的时期，就是1999年到2000年互联网泡沫的末期。

And really, the only time we've seen kind of a period of rising interest rates where growth still tried to kind of lead the market higher was in 1999, 2000 at the end of the Dotcom bubble.

所以我觉得我们现在就处于这样一个阶段，成长股还在试图坚持，希望利率能转向对它们有利的方向。

So I think we're kind of in that kind of a stage where growth is still kind of trying to hang on and hope that rates are going to turn around in their favor.

但如果美联储不得不像我预期的那样去做，那么价值股就会接管市场，这正是2000年发生的事情——当互联网泡沫开始破裂，主要成长股纷纷下跌时，价值股表现非常出色，作为轮动的一部分，推动了许多被低估的传统经济类股票上涨。

But if the Fed is going to have to, you know, do what I think they're going to have to do, then value stocks are going to take over, and that's exactly what happened in 2000 was, you know, when the .com bubble started to burst and the the the major growth stocks rolled over, value stocks did phenomenally well as a part of a rotation that that boosted a lot of undervalued kind of old economy type names.

那在这个高通胀时期呢？

What about, you know, in this high inflationary period?

而且你刚才也提到了这一点。

And you just touched on this too.

我们还面临着大规模供应链中断的背景。

We're also against the backdrop of massive supply chain disruptions.

还有战争，它的前景……

We have war that is...

非常不确定。

Has a very uncertain outlook.

你知道，我们原本以为这事儿能解决，结果现在……

You know, we went from thinking this would be resolved to being...

看起来在升级，虽然情况每天都在变。

Looking like it's escalating, although that changes, you know, day to day.

但是……

But are...

大宗商品在这里面扮演什么角色呢？

How do commodities fit into the picture here?

因为看起来现在应该是大宗商品繁荣或者超级周期的时候。

Because it would seem like this is this is a time where commodities should be in some sort of boom or super cycle.

但我不确定，由于高利率的压力，供需动态会不会在这里起作用。

But I don't know if the supply demand dynamics because of the pinch of high rates factor in here.

你对大宗商品整体怎么看？

How are you thinking about the commodity complex?

是的，你知道，我觉得柴油价格创下历史新高，对吧，这显然会给全球经济带来压力。

Yeah, you know, I think with with diesel prices hitting record highs, right, that's obviously going to put strain on economies around the world.

我确实认为，你知道，我们之前也聊过，过去几年每次我们谈话我都提到，我是大宗商品的坚定看多者。

I do think, you know, and we've talked about this, I think every time we've spoken for the last few years, I'm a huge commodities bull.

我觉得我们正处于一个大型大宗商品超级周期之中。

I think we're in the midst of a major commodity supercycle.

我认为这个周期远未结束，因为这些超级周期是由资本周期驱动的。

I don't think we're anywhere near the end of that yet because these supercycles are driven by the capital cycle.

所以当你开始看到公司说，我们要大规模投资，增加石油、天然气、铜等新供应时，那才是结束的开始。

So when you start to see companies say, we're going to invest massively in bringing on new supplies of oil and gas and copper and these types of things, that will be the beginning of the end.

但我们甚至还没到那个阶段，公司还没有大规模投资新的钻探机会，对吧？

But we haven't even gotten to that point yet where companies are investing massively in new drilling opportunities, right?

他们就是没这么做。

They're just not doing that.

所以这告诉你，市场还没有出现供应端的反应。

And so that tells you that there is no supply response yet to what's going on in the market.

所以我们看到这种情况的地方，显然是在AI交易上，对吧？

So where we're seeing that is obviously in the AI trade, right?

那里有巨大的供应反应，对吧？

Where there's a huge supply response, right?

需求持续太久了，现在每个内存供应商都在全球投资数十亿美元来增加新供应，更不用说中国出于地缘政治安全考虑，政府支持芯片制造业，这也会带来大量供应。相比之下，我更担心那里的周期末期动态，而不是大宗商品领域的任何问题。你知道，我们会看到调整之类的事情发生，但能源板块今年迄今上涨了50%以上，投资者仍然不感兴趣，这对我来说还是有点令人惊讶。

There's so much been so much demand for so long that, yeah, now you have every memory supplier around the world investing billions and billions of dollars in bringing new supply on market, not to mention the fact that China is invested just, you know, literally as a geopolitical kind of security concern as a government to support the chipmaking industry in a way that, you know, is going to bring on massive supplies, you know, I'm much more worried about end of cycle kind of dynamics there in that, than anything in the commodity sector, you know, we're going to see, you know, there are going to be corrections and things that happen, But it's just kind of still amazing to me that the energy sector is up 50% plus so far this year, and investors still have no interest.

他们还是在追逐大型科技股，你知道吗？

It's still chasing the big tech names, you know?

是的，顺便说一下，就在我们录制节目的今天，即使有所有这些新闻和你刚才提到的所有担忧，我们也有点……

Yeah, as we speak, by the way, like on on the day that we are recording this, even with all the news out there and all the concerns that you just mentioned, we have sort of...

你知道，AI交易又回到了领先地位，但我们听到和看到了很多关于它的担忧。

You know, the AI trade is back in leadership, but we're seeing and hearing so many concerns about that.

而且你很早就提到了一些会计问题。

And you very early on, you were talking about some of the accounting issues.

所以这符合你对AI周期末端的判断。

So that's that fits into your end of cycle dynamics around AI.

是的，你知道，没有什么比得上通胀和利率上升，真的没有什么比利率上升更能戳破投机泡沫了。

Yeah, well, well, you know, there's nothing like rising it, we talked about inflation, rising interest rates, there's really nothing like rising interest rates to burst a speculative bubble.

纵观历史，总是利率上升来戳破泡沫。

If you look throughout history, it's always rising interest rates that crop up to burst a bubble.

但我觉得就AI泡沫本身而言，罗杰·麦克纳米最近接受了一个采访，他是Facebook的早期投资者之一，也是过去三十年最成功的科技投资者之一，他基本上说，你有这些AI实验室，对吧？

But I think in terms of where we are in the AI Bubble itself, Roger McNamee recently gave an interview where Roger was one of the early investors in Facebook, one of the most successful tech investors of the last thirty years, and he basically said, You have these AI labs, right?

我们都在谈论那些安全问题，但这其实有点分散了真正发生的事情的注意力。

We're talking about all the security stuff, but it's really kind of a distraction from what's going on.

他们下了很大的赌注在通用人工智能上。

They made a huge bet on artificial general intelligence.

这个赌注没成功，对吧？

The bet hasn't worked, right?

他们没能造出超级智能。

They haven't been able to create super intelligence.

开源的那些人开始真正抢他们的饭碗了。

The open source guys are starting to really eat their lunch.

他们在数据中心和Hugging Face事件上经历了一轮灾难性的公关危机，现在他们快到了市场真的不想再给他们钱的地步。

They've had a disastrous PR cycle with data centers and the hugging face incident, and now they're approaching the point where markets really don't want to give them any more money.

所以，我觉得现在的情况是，我们正处于一种绝望的阶段，如果资金流不来，你知道，上周我们看到甲骨文的贷款出现困境，甲骨文债券价格下跌，还有与简街建设数据中心相关的贷款也出现困境。

And so, you know, I think what's going on is we're kind of in this kind of desperation phase, where if the money flow doesn't come, you know, if start to see distress in Oracle loans last week, prices of Oracle bonds, distress in actually loans related to Jane Street's building of a data center.

软银现在想以垃圾信用评级出售，以便为OpenAI融资。

SoftBank now wants to sell at a junk credit rating in order to fund OpenAI.

所以信贷市场正在收紧，基本上在告诉你，AI交易的钱可能开始枯竭了。

So credit markets are tightening and basically telling you that the money is potentially starting to run out for the AI trade.

如果真的这样，显然会有重大的连锁反应，不仅影响AI，还会影响整个股市和经济，因为经济已经如此依赖AI交易了。

If it does, obviously, there's major ripple effects, not just for AI, but the broad stock market and the economy, which has become so dependent on the AI trade.

我们频道做的所有事情，目标只有一个，就是让你成为一个更敏锐、更明智的投资者，一个能在主流市场反应过来之前，就看懂影响你财富的宏观力量的人。

Everything we do on this channel has one goal, to make you a sharper, better informed investor, someone who understands the macro forces shaping your money before the mainstream catches on.

但信息只是其中一半。

But information is only half of it.

另一半是付诸行动。

The other half is acting on it.

而这正是Wealthion会员的意义所在。

And that's what Wealthion membership is for.

它让你从观看这些对话，转变为把它们应用到自己的投资组合中。

It's how you go from watching these conversations to putting them to work in your own portfolio.

会员资格为你打开三扇门：获得顾问的指导，帮助你把真实资产纳入投资组合，以及为合格投资者提供专业基金的准入。

Membership opens three doors: access to advisors, helping build real assets into your portfolio, and for accredited investors, access to specialized funds.

想了解更多，成为会员的链接就在下方的描述框里。

To learn more, the link to become a member is in the description box below.

嗯，我觉得这正是人们担心的。

Well, I think that's what people worry about.

你是说，从资金方面来看，你开始看到投资者说，好吧。

You're saying from the, you know, from the funding side, you're starting to see investors say, okay.

嘿。

Hey.

等一下。

Hang on a second.

我完全看不到任何形式的投资回报。

I'm not seeing that ROI in any way, shape, or form.

而且高管们自己也在谈论放慢步伐，也许服务……

And you have the executives themselves talking about slowing things down, maybe serving...

也许试图提前应对他们已经看到的资本放缓。

Maybe trying to get ahead of that capital slowdown they're already seeing.

但如果所有这些因素加在一起，你知道，我接触过的一些AI业内人士说，如果是这样的话，你看到的是资产价格的剧烈重估。

But if you have all of that together, you know, there have been some AI folks I talked to that say, like, you're you're looking at a violent repricing of assets if that's the case.

这是主要的……

Is that the major...

跟我说说，这种风险再加上利率上升和通胀环境，会是什么样子？

Talk to me about what that risk looks like paired with rising interest rates and an inflationary environment.

听起来对经济和市场来说是个有毒的组合。

That sounds like a tox brew for the economy and markets.

是的，确实如此。

You know, it it is.

你知道，我们现在在通胀方面确实处于一个艰难的境地。

You know, we're really in a in a difficult place right now in terms of where we are in inflation wise.

这非常滞胀，对吧？

It's, I mean, it's very stagflationary, right?

在消费者已经承压、房地产市场已经紧张的情况下加息，对吧？

You raise interest rates with already a stressed consumer, stressed housing market, right?

而房地产是周期性GDP的主要驱动力之一。

One of the main drivers of kind of cyclical GDP.

现在一切都依赖于AI泡沫，对吧？

And really everything is now dependent on the AI Bubble, right?

经济、股市，还有所有东西，都跟它挂钩了。

The economy and the stock market and everything is kind of levered to it.

所以是的，经济放缓，真的没办法在不引发大规模连锁问题的情况下实现放缓。

So yeah, a slowdown, there's really no way to kind of do a slowdown without causing problems that ripple outward in a major way.

因为你想啊，现在所有半导体公司的盈利预测，百分之百都依赖于未来几年建设数据中心，好让这些半导体都能装进去。

Because when you think about it, literally, right now, all of the forecasts for semiconductor earnings are 100% dependent on building out data centers over the next several years that can handle, where all these semiconductors can be installed.

所以如果我们放缓，我们已经建不起来了，我觉得这就是实际情况，放缓已经在发生了。

So if you know, we slow down, we already can't build it, you know, this is, I think this is really kind of what's going on, is that the slowdown's already happening.

我们根本建不了那么多数据中心来满足盈利预期。

We cannot build the number of data centers needed to meet the earnings estimates.

这根本不可能实现。

It's just not going to happen.

所以很多公司，你知道，很多说法可能是我们只是在找借口，我们是为了安全原因才这么做的。

And so a lot of companies are, you know, a lot of the narrative could be that we're just looking for an excuse, you know, we're doing it for safety reasons.

就像OpenAI，说为了安全原因推迟IPO，但当你听到他们几个月前尝试上市时传出的各种消息，投资银行基本上告诉他们，这事儿成不了，对吧？

It was like OpenAI, saying we're delaying our IPO for safety reasons, when, when you listen to all the kind of noise that came out of their attempt to go public a couple of months ago, the investment banks basically told them, It's not happening, right?

对于一家市值那么大的公司来说，未来几年要亏损三千亿美元，你还想让公众来买单，那根本就不会有那么多需求来支撑一个万亿美元以上的IPO。

The demand for a company that's markets, you're going to lose 300,000,000,000 over the next few years, and you want the public to fund that, there's just not going to be the demand there for a trillion dollar plus IPO.

Anthropic稍微好一点，因为他们声称自己盈利了，但这完全是基于虚假的会计操作，就是说如果把所有开支都排除掉，我们就盈利了。

Anthropics getting a little closer because they are claiming to be profitable, based on literally bogus accounting, you know, saying if you exclude all these expenses, we're profitable.

这不就是WeWork玩过的把戏吗？

Same thing that WeWork tried to do, right?

我记得不太清了，WeWork当时搞了个什么调整后的EBITDA，现在Anthropic也在玩同样的套路。

Know, I can't remember, you know, WeWork's kind of whatever it was, human adjusted EBITDA, but Anthropic's trying to do the same thing.

所以，是的，我们真的快到周期尾声了，因为要支撑下一阶段发展所需的资金量实在太大了。

So, yeah, we're getting really close to the end of the cycle here just because the amount of money needed to to fund the next phase is is just getting too big.

但是白宫那边好像还在踩油门，说我们不会放慢脚步。

Do you, we have a White House that seems to be stepping on the gas though and saying, we're not slowing down.

这是必需的。

This is needed.

这关系到国家安全。

This is national security.

那这会改变经济放缓的可能轨迹吗？

Does that change the potential trajectory of a slowdown?

是不是有联邦政府给AI行业兜底，能缓冲你所说的那种大变动或者清算带来的市场和经济的冲击？

Is there sort of a federal put behind the AI trade that might buffer that sort of market and economic fallout that would come from a from a big change or from a, you know, from a reckoning that you're describing?

我不这么认为，因为想想看，有两个原因。

I don't I don't think so because when you think for, you know, for for two reasons.

第一，有人问过斯科特·贝森特关于最近的安全问题讨论，他直说了，监管这些公司最好的办法就是让他们对黑客事件负责，不给他们免责，这话已经放出来了。

One, Scott Bessent has been asked about, you know, the the recent talk about security issues and whatever, and he said straight up, The best way to regulate these companies is make them liable for their hacking episodes, right, is to not indemnify them, and that's literally, you know, that's come out.

这个AI泡沫最让人着迷的一点就是看到两党都支持，或者说，反AI、反数据中心的运动，对吧？

One of the most fascinating things about this AI Bubble is to see the bipartisan support, or I guess, you know, for the anti AI movement, the anti data center movement, right?

史蒂夫·班农和伯尼·桑德斯手拉手说反数据中心，对吧？

You have Steve Bannon and Bernie Sanders literally holding hands and saying, anti data centers, right?

还有前FTC主席莉娜·汗，她真的想重新启用整个反垄断框架来打击大型科技公司的垄断，还有大卫·萨克斯，他之前是白宫的AI沙皇，后来离开了。

And then you have Lina Khan, former head of the FTC, who really kind of tried to bring back the whole antitrust framework to attack big tech monopolies, and David Sacks, who was a former kind of AI czar at the White House before he left.

所以左派和右派都在说，我们已经有现成的法律可以追究这些公司的责任了。

So you have left and right, both they're saying, We already have the laws on the books to hold these companies accountable.

我们只需要去执行这些法律就行了。

We just need to enforce them.

所以你看，两党都支持这一点，不只是反对建数据中心，还主张起诉这些公司。

So you have this bipartisan support, it's not only anti data center, but it's pro prosecution.

我是说，他们都在讨论要起诉这些公司，因为这种行为本身就是违法的，对吧？

I mean, they're talking about prosecuting these companies for literally, it's illegal, right?

入侵其他公司的系统可是重罪。

Hacking another company is a felony.

所以让公司承担责任，也就是说，哪怕是你的……

So holding the company is liable you know Even if your

哪怕是AI代理干的，你也得负责。

AI agents do it, you're liable.

没错。

Exactly.

就像莉娜·汗说的那样，其实已经有法律条文规定，公司要对推出缺陷产品负责，对吧？

So you, as Lina Khan put it, you know, there are already laws on the books for companies putting out defective products, right?

你把有缺陷的产品推向市场，造成伤害，你就要为此负责。

You put out a defective product out into the public, it causes harm, you're liable for that.

所以我觉得这也是正在发生的一件事，这些公司开始慌了，天哪，我们可能要承担的责任有多大？

So I think this is another thing that's going on, the companies are going, Holy cow, our potential liability?

马特·莱文在彭博社写过这个。

Matt Levine wrote about this in Bloomberg.

他说，你知道，你的潜在市场规模据说是30万亿美元，对吧？

He said, you know, your TAM is $30,000,000,000,000 supposedly, right?

Anthropic想让我们相信他们的总可寻址市场相当于美国的GDP。

Anthropic wants us to believe their total addressable market is equal to US GDP.

确实，在某些情况下，我们的收入可能等于美国的GDP。

Literally, there's some world where our revenues are equal to US GDP.

但他问，你的负面TAM是多少？

But he said, what's what's your negative TAM?

你的负面TAM是多少？

What is your negative TAM?

这就是投资者们开始需要考虑的问题。

That's what investors have to start thinking about.

因为你知道，他有点把这事比作烟草公司，对吧？

Because if you, you know, he kind of likened it to the cigarette companies, right?

如果烟草公司上市的时候，他们说，我们的潜在总市场规模是全世界每个人每天抽两包烟。

If you, when the cigarette companies came public, they had said, you know, our potential, you know, total addressable market is every human being in the world smoking two packs a day.

我们潜在的负面就是所有因为每天抽两包烟得癌症的人提起的诉讼。

Our potential negativity is all the lawsuits that come from people who got cancer smoking two packs a day.

你知道，基本上，AI实验室今天也在告诉我们类似的事情，我们的潜在总市场规模是30万亿美元，但我们的潜在责任是毁灭人类。

You know, they basically, the AI labs are telling us kind of the same thing today that our potential TAM is $30,000,000,000,000 but our potential liability is destroying humankind.

最后是什么

What's the last

一个伟大的监管文件？

thing a great regulatory filing?

但顺便说一句，先毁灭人类是违法的。

But by the way, that's illegal to destroy humankind first.

而且，就算只是试图毁灭人类，法律上要承担什么责任呢？

And what's the legal liability, you know, even attempting to destroy humankind?

所以这有点疯狂。

So it's kind of a madness.

这就是...

This is where...

而在这里，确实很难绕开数学和基本面。

And this is where, you know, it's hard to get away from sort of math and fundamentals.

因为...

Because...

我要说的是，我们收到的反对意见，我们花时间在这上面，是因为每个人都持有...

And I will say the pushback we we get, and we're spending time on this because, everyone holds...

很有可能我们大多数人持有这些股票，或者持有股市的一部分，而这部分现在依赖于此，我认为你和其他人会说，这占了股市的很大一部分。

Very good chance most of us hold these names or holds, you know, a portion of the stock market that is reliant on this now, which I think you and others argue is an awful lot of the stock market.

你知道，这种领导力，我们看到的这些涨幅，都依赖于这个。

You know, this leadership, you know, the gains we've seen depends on this.

所以这是我们所有人都面临的风险。

So this is an exposure that we all have.

所以我觉得这个话题真的值得好好聊聊。

So I think it's really worth having this conversation around this.

但如果真的像这样，你躲不开这个数学问题，那就会让大家开始想，好吧。

But if if this is like, you can't escape this math, this is going to, cause people to say, okay.

那我该怎么对冲这个风险？

Well, how do I hedge this?

比如说，什么……

Like, what...

我……

I I...

我们一直在谈分散投资，但现在可能正在发生一个巨大的尾部风险事件。

We talk about diversification all the time, but this is a massive tail event potentially unfolding now.

那么，杰西，投资领域里还有哪部分跟这个没关系呢？

So, you know, what part of the, investing universe is not attached to this, Jesse?

这个问题问得很好，因为你知道，今天彭博社有篇文章提到，养老金管理者在努力限制他们对AI的敞口方面真的很头疼。

I mean, it's it's a great question because, you know, Bloomberg had a piece out today about how pension managers are really struggling with trying to limit their AI exposure.

他们指出，AI基础设施，现在标普500指数里有40%都是AI基础设施股票。

I think they point out that AI infrastructure, 40%, 40% of the S and P 500 now is just AI infrastructure stocks.

新兴市场中有28%的份额仅仅来自三只芯片股。

28% of emerging markets is just three chip stocks.

三只芯片股就占了所有新兴市场指数的28%。

Three chip stocks make up 28% of all emerging markets indexes.

今年投资级债券发行量的一半都与AI相关，而今年90%的风险投资资金都投向了AI。

Half of investment grade bond issuance this year is tied to AI, and 90% of VC funding this year is all AI.

所以如果你是养老基金，你会想，我到底该怎么分散投资呢？

So if you're a pension fund, you're going, how the heck do I try and diversify?

我觉得这就是为什么我们看到黄金持续受到追捧，因为人们在想，我还能持有其他什么资产？

And I think that's why we're seeing a pretty persistent bid for gold, is people saying, What else can I own?

如果我不想持有国债，也不想持有公司债券，那么除了这些金融资产和风险资产之外，唯一的选择就是实物资产，比如关注大宗商品、贵金属这类东西。

If I don't want to own treasuries, I don't want to own corporate bonds, the really only other choice outside of these kind of financial assets, risk, you know, is real assets, is trying to focus on commodities, precious metals, these types of things.

而且，你知道吗，这真的很吸引人，因为历史告诉我们，在高通胀时期，什么表现好，对吧？

And really, you know, it's fascinating because that is what history tells you does well during periods of high inflation, right?

在高通胀时期，金融资产表现不佳。

During periods of high inflation, financial assets do poorly.

股票和债券表现不佳。

Stocks and bonds do poorly.

它们会很挣扎。

They struggle.

在高通胀时期，实物资产表现非常好，这真的是一个非此即彼的情况。

During periods of high inflation, real assets do phenomenally well, and it really is an eitheror type of thing.

当一个表现好的时候，通常是以另一个为代价的。

When one does well, it's usually at the other's expense.

所以我认为投资者对实物资产的配置太少了。

And so I think that investors are so underexposed to real assets.

我认为现在能源板块只占标普500的3%。

I think the energy sector is 3% of the S and P 500 right now.

这仍然是最低的水平之一。

It's still among the lowest levels.

英伟达的市值比整个能源行业还要高。

Nvidia is worth more than the entire energy sector.

所以，当你从资本周期的角度来看我们现在所处的位置，如果资金要从科技和金融资产中流出，流入这些其他行业。

So when you look at where we are in terms of the capital cycle, if money is going to flow out of technology and financial assets broadly and flow into these other sectors.

那么，能源在标普500中的占比从3%上升到10%，这并不令人意外。

It wouldn't be surprising to see something like energy go from 3% to 10% of the S and P 500 over time.

而黄金的配置，你知道，从零、百分之一、百分之二，增加到机构投资组合平均的5%。

And allocations to gold you know, go from zero, one, two percent, and, you know, the average institutional portfolio to 5%.

你知道，这其实就是推动这些超级周期的真正原因。

You know, that's kind of what really drives these supercycles.

那么，有两个问题。

So two questions.

我们，算是对之前讨论的一个反驳，我们几周前和汤姆·李聊过，你知道，我觉得这是我们都在纠结的问题。

We, sort of as a pushback to what we're talking about before, we we spoke to Tom Lee a couple weeks ago, and, you know, and I think this is what we all grapple with.

现在市场上确实有各种不同的说法。

That that sort of diverse narratives out there.

然后他……

And he...

他的观点大致是说，大多数人不擅长做成长型投资。

His sort of comment was something to the effect that, most people are bad growth investors.

你知道，我们正处在所有这些变革的关口。

You know, we're at the cusp of all of these changes.

好吧，也许个别公司会倒闭，或者一些泡沫会破裂。

And, okay, you know, maybe individually, some companies are gonna fail or bust or some bubbles will burst.

但总的来说，方向是更多的计算，更多的技术渗透到一切。

But generally, the the direction is more compute, more technology in everything.

而且我觉得，当我们回顾过去通胀时期，大宗商品的表现，我们经常讨论这个，对吧？

And I think when we think about this idea in past inflationary periods, commodities, and we talk about this a lot, right?

我们关注的是实物资产、稀缺性，以及如何分散投资到不相关的资产上。

We're all about real assets, scarcity, how to diversify into non correlating.

这就是我们在Wealthion整天讨论的内容。

That's what we spend all our time at Wealthion talking about.

但是……

But the...

你会听到类似的说法，比如在过去通胀时期，大宗商品表现不错，实物资产跑赢了。

You'll hear things like that and say, well, in the past, inflationary periods, commodities did well, and real assets outperformed.

但在一个AI代理已经到处跑、我们通过手机做所有事情的数字世界里，这些同样的指标还适用吗？

But in a world where AI agents are kind of already running around and we're digitally doing everything off our phone, Do those same metrics apply?

还是说我们正在进入一个一切都金融化的数字化世界？

Or are we just moving into a world of digitization where everything is financialized?

你怎么看待这种说法？

How do you think about that, you know, that narrative?

我觉得有两件事。

I I think there's there's two things.

每一次投资热潮中都有这样一件事。

There's one in every investment boom.

这个在某些方面是独特的，但在每一次投资热潮中，你总会走到过度投资那一步，对吧？

And this one is unique in some respects, but in every investment boom, you always get to overinvestment, right?

公司从来不会只投资刚好够用的钱来创造丰厚的回报，然后这些回报就实现了。

Companies never invest just enough to create wonderful returns, and then those returns manifest.

因为贪婪，他们总是过度投资，是的，总是这样。

There's always, because of the greed, yeah, they always overinvest.

我的意思是，这基本上就是人类历史、市场历史，以及一切的历史。

I mean, that's literally the history of humankind, of markets, of everything.

所以问题就是，我们什么时候会到达那个过度投资的点？

And so the question is, when do we get to that point of overinvestment?

所以不仅仅是数据中心供应过剩的问题，还有那些规律……

So it's not just amount of oversupply in data centers, but the laws of...

科技史也告诉我们，互联网泡沫就是一个很好的例子，当你从供应不足转向供应过剩时，同时，因为价格动态变得极其极端，比如内存，这就给开发技术的公司创造了巨大的动力，让他们变得更高效，找到用更少内存的方法，找到更高效利用算力的方法。

The history of technology tells us too, and the .com boom was a great example of this, is that as you kind of cross from undersupplied to oversupplied, at the same time, because those price dynamics, like in memory, for example, become so extreme, it creates a huge incentive for the companies developing the technology to become way more efficient and find ways to use less memory, find ways to utilize the compute way more efficiently.

所以在互联网泡沫期间，我们不仅看到了光纤容量的过剩，而且公司在利用这些容量方面也变得更高效，这就是为什么我们有了暗光纤，而且供应过剩持续了好多年，因为他们不仅建得太多，而且我们在使用上也变得更高效了。

And so during the Dotcom bubble, we not just saw an oversupply of fiber optic capacity, but companies became way more efficient in utilizing that, which is why we had dark fiber literally for years and years and years of oversupply, because not only did they build too much, then we became way more efficient in using it.

所以我觉得今天已经在发生同样的事情了。

And so I think the same thing is happening already today.

我们已经看到公司在内存使用上变得更高效，在计算使用上也变得更高效。

We're seeing companies already become way more efficient in their use of memory, way more efficient in their use of compute.

而且我们甚至看到在个人电脑上本地使用AI模型方面取得了巨大进步，对吧？

And we're even seeing huge strides in the use of AI models locally on your PC, right?

你没法在个人电脑上运行前沿模型，但通过使用正确的工具和正确的代理，你实际上可以用一个小型语言模型在笔记本电脑上获得和大型数据中心几乎相同的结果。

You can't run a frontier model through your PC, but you can literally, by using the correct harness, using the correct agents, these types of things, you can use a small language model and essentially get the same results on a laptop as you can using a monstrous data center.

而且这已经是现在的事了，对吧？

And that's already now, right?

所以这项技术本身也在不断改进。

So that technology is also itself improving.

而且我觉得你会看到，比如我们听到的一个情况是，公司不太愿意使用OpenAI和Anthropic的模型。

And I think you're seeing that, you know, like one of the things that we're hearing about is companies less willing to use open and anthropic models.

一是因为它们太贵了。

One, because they're so expensive.

我们有开放模型。

We have the open models.

开源模型便宜得多。

Open source models are way cheaper.

但而且，他们发现OpenAI和Anthropic在偷他们的知识产权，对吧？

But also, they're finding out that OpenAI and Anthropic are stealing their IP, right?

看着你使用他们的AI，看你拿它做什么，然后说，好吧，我们要把你刚才做的拿过来复制，做出我们自己的产品。

Watching you use their AI and seeing what you're doing with it, and then saying, Okay, we're going to take what you just did and replicate it and create our own product.

所以很多客户说，我们也不想冒那个风险。

And so a lot of these customers are saying, We don't want that risk either.

我们可以省钱，还能避免知识产权被偷，用这些开源模型，而且，你知道，直接下载开源模型放到本地服务器上跑，连数据中心都不用。

We can save money and eliminate IP theft by using these open source models, and also by building, you know, by literally downloading the open source models and using them on local servers, not even have to use a data center.

所以我觉得这些动态是现在大家不一定意识到的，当我们正处在AI的热潮中，大家都在兴奋地建设的时候，就是不仅我们可能会过度建设，我们还会变得更高效，这会自然地带来那种衰退，就像历史上每次投资热潮之后那样。

So I think these are the dynamics that people aren't necessarily appreciating right now when we're kind of in the moment of AI and this excitement of the build and everything, is that not only are we probably going to overbuild, we're probably going to get way more efficient, and it's going to naturally create the bust that you see, you know, after every investment boom in history.

我觉得这个回答很有深度，听你这么说，让我想到，这不意味着……

I think that's a really thoughtful answer, and it makes me think as I listen to that, It doesn't mean...

我觉得人们总是喜欢把人分成不同的阵营，对吧，然后把这个对话隔离开来，我觉得这正是危险的地方，而我们一直在努力搭建桥梁。

I think people tend to wanna put put people in narrative camps, right, and sort of silo this conversation, which is what what I think is so dangerous and we're trying to always bridge.

你可以对技术非常乐观，并且以增长为导向，但同时也要明白，对基础设施的过度投资会发生，而且会带来很多损害和后果，或者只是失败，或者公司不得不缩减规模，或者估值调整，随着技术增长的下一个阶段到来并呈现出不同的面貌。

And you can be sort of super optimistic and growth oriented toward the technology, but still understand that the overinvestment in the infrastructure part of this story will happen, and there will be lots of damage from that and fallout from that and or just like failures or or companies having to reduce size or valuations adjusting as the next phase of that technology growth picks up and looks like something else.

所以当你谈到光纤和所有的……

So when you're talking about the fiber optics and all the...

那些……

That...

为互联网铺设的一切，并不意味着我们没有看到基于此发展起来的应用、服务和其他所有东西。

Everything that was laid out for the Internet, that that did not mean that we didn't see the development of apps and services and all this other stuff built on that.

但这也没有改变一个事实，你知道，有些公司破产了，因为增长故事的那个阶段结束了。

But that also didn't change the fact that, you know, some of these companies went bust because that phase of the growth story ended.

你知道吗？

You know?

而且还有整合、后果和效率提升，这些都改变了那个动态。

And and there was consolidation and fallout and efficiency that changed that dynamic.

所以我觉得，指出这一点并不是说你反对增长。

So it's not like you're anti growth pointing this out, I think.

你知道吗？

You know?

而且不。

And No.

不。

No.

我的意思是，你听起来像是相信科技在改变。

I mean, you sound you sound like you believe technology is changing.

你只是真的在深入研究投资周期的数学。

You're just you're just really digging in the math of the investment cycle.

完全正确。

Absolutely.

我，你知道，我一直是个科技...

I, you know, I've been a technology...

我是科技爱好者。

Fan of technology.

你知道，我八岁那年参加了苹果电脑夏令营，学会了用Logo语言编程。说真的，我真的是个超级科技迷。

You know, when I was eight years old, I went to Apple Computer Camp and learned to program logo, and so, I mean, I'm literally a huge fan of technology.

这可能会暴露我的年龄，不过……

You know, dates me a little bit, but

谢谢你分享这些。

Thank you for sharing that.

但我一直是科技的超级粉丝，生活中也早早尝试过各种新事物。我真的对它感到惊叹，觉得它太棒了，而且我在工作中经常使用 Perplexity。

But I've been a huge fan of technology, and an early adopter of so many different things throughout my whole life, I'm like, you know, I'm amazed by it, and I think it's phenomenal, and I use perplexity a lot in a lot of the work that I do.

我也很欣赏一点，就是它确实会经常出错，毕竟它是有局限性的。

I also appreciate that, you know, it can be wrong a lot of the times and things, and, you know, there are limitations to it.

对，我还想补充一点：你看这一轮科技周期里最出色的那些公司，其实都是诞生于互联网泡沫破裂之后的。

But yeah, I do think, you know, another point that I would make is that when you look at the, I guess, some of the most wonderful companies of this technology cycle, they were born out of the Dotcom bust, right?

如果没有当时互联网相关资产价格暴跌、几乎跌到零，导致上网变得极其便宜，那就不可能有 Google、Facebook，也不会有这些在泡沫期还没出现、却靠泡沫破裂后的产能过剩而崛起的企业。

If you didn't have a collapse in the, you know, pricing of, you know, internet and all this kind of stuff to almost zero to where, you know, using the internet became super cheap, then you don't have Google, you don't have Facebook, you don't have these companies that weren't around during the dot com bubble, but out of the bust, that overcapacity kind of enabled them.

我真的觉得AI会在这里起飞，就是如果我们——想象一下如果真的出现一次大的AI泡沫破裂，算力价格暴跌到几乎为零，那时候人们才能真正利用它来建立可持续的商业模式。

And I really think that's where AI is going to take off, is if we and imagine if you do have a major AI bust, right, and the price of compute plunges to almost nothing, right, that's when people are going to be able to utilize it in a way that they can create sustainable business models.

但现在算力成本这么高，还有人花几万美元去搞AI代理，结果订错酒店、订错日期，还得退款，你知道，就是运行不正常，而且成本太高。

But when computers costs what it does, and you have people spending tens of thousands of dollars on trying to create AI agents that buy them the wrong hotel room on the wrong date and have to refund, you know, it's just not working properly and it's costly.

我们还没到那一步。

You know, we're just not there yet.

但我确实觉得，一旦你……

But I do think, I do think, you know, that once you...

可能最健康的方式，而且我觉得马克·扎克伯格和山姆·奥特曼心里都明白，对技术长期发展最好的事情就是来一次大崩盘，把算力成本压到几乎为零，这样才能为未来真正漂亮的商业模式打下基础。

Probably the healthiest thing, and I think that, you know, down deep, Mark Zuckerberg understands this, Sam Altman understands this, that the best thing for the technology longer term would be a huge bust to get the cost of compute down to next to nothing would set the groundwork, you know, lay the groundwork for some real beautiful business models going forward.

不过，那还得再走几步。

But, you know, that's that's a couple steps down the road.

中间会有一些痛苦……

And some pain...

真正的痛苦。

Real pain in between.

然后，我们接下来要回到大宗商品的供应动态这个话题，因为挺有意思的。

And and, know, we we're we're gonna shift back to talking about supply dynamics with commodities because this is interesting.

你看到这个会说，所有这些AI基础设施和算力的供应太多了，而且太贵了。

You're looking at this and saying there's too much supply of all of this AI sort of infrastructure and compute, and it's too expensive.

所以需求正在远离它，因为它……

And so it's the demand is is moving away from it because it...

经济账算不过来。

The economics don't work.

如果我们看大宗商品这边的供需，当我听到这些，看到连锁反应、估值下跌、股市可能下跌，对经济产生连锁影响，我就会想，这在大宗商品超级周期里是怎么运作的？

If we look at supply demand on the on the commodity side, so when I hear that and I see fallout, valuations going down, maybe equities going down, the knock on effect on the economy, I come back to how does that work with commodities in the supercycle?

因为那意味着，如果我们处于衰退环境和大规模市场调整中，所有东西的需求端可能都会承压，这还没算债券。

Because that means the demand side of everything might be under pressure if we're in a in a recessionary environment and in a major market correction, and that's leaving bonds out of it.

我们回头再聊这个。

We'll we'll circle back to that.

大宗商品超级周期是怎么运作的？

How does how does the commodity supercycle work?

我们会先看到它下跌吗，还是说它会随着需求故事一起进入下行周期，或者你关注的是供应端，寻找进一步的上涨空间？

Are we gonna see a dip in it before it's it's a a part of the cycle that goes down with that demand story, or or does the supply side is what you're focused on in terms of looking for further green?

其实关键在供应端。

It's really the supply side.

你看历史上的大宗商品超级周期，真正推动它的是供应端，因为需求其实没有大家想的那么有弹性，对吧？

When you look at these historical commodity supercycles, it's really the supply side that drives it, because demand is really not as elastic as I think people like to think it is, right?

即使是在金融危机期间，能源需求也没有大幅下降。

Even during the great financial crisis, demand, energy demand didn't drop very, very significantly.

当然，疫情期间确实下降了，对吧？

So it did, obviously, during the pandemic, right?

但那是靠全球封锁，大家不能开车、不能坐飞机，才真正把需求压下来，甚至让价格一度变成负数。

But that took a lockdown, global lockdowns of people, you know, preventing them from driving or flying anywhere to really, really, you know, bring demand down in a way that brought prices to negative for a short period of time.

但需求其实是一个长期稳步增长的趋势，对吧？

But demand is really kind of, a longer term dynamic that is just steadily growing, right?

我们都在用更多能源，新兴市场也在用更多能源，所以能源需求一直在不断增长。

We're all using more energy, and emerging markets using more energy, and so energy demand continues to grow, grow and grow.