The Compound and Friends - Big Market Delusions

Episode Date: September 15, 2023

On episode 109 of The Compound and Friends, Michael Batnick and Downtown Josh Brown are joined by Rob Arnott in front of a live audience at Future Proof. They discuss: lessons from the tech bubble, smart beta, big tech valuations, Tesla and Nvidia, how AI might impact markets and the economy, and much more! Thanks to KraneShares for sponsoring this episode. KraneShares and Rockefeller Asset Management are launching the Ocean Engagement ETF. This ETF invests in public companies with significant positive impact on oceans and ocean resources. Learn more at: https://kraneshares.com/ksea/ Check out the latest in financial blogger fashion at The Compound shop: https://www.idontshop.com Investing involves the risk of loss. This podcast is for informational purposes only and should not be or regarded as personalized investment advice or relied upon for investment decisions. Michael Batnick and Josh Brown are employees of Ritholtz Wealth Management and may maintain positions in the securities discussed in this video. All opinions expressed by them are solely their own opinion and do not reflect the opinion of Ritholtz Wealth Management. Wealthcast Media, an affiliate of Ritholtz Wealth Management, receives payment from various entities for advertisements in affiliated podcasts, blogs and emails. Inclusion of such advertisements does not constitute or imply endorsement, sponsorship or recommendation thereof, or any affiliation therewith, by the Content Creator or by Ritholtz Wealth Management or any of its employees. For additional advertisement disclaimers see here https://ritholtzwealth.com/advertising-disclaimers. Investments in securities involve the risk of loss. Any mention of a particular security and related performance data is not a recommendation to buy or sell that security. The information provided on this website (including any information that may be accessed through this website) is not directed at any investor or category of investors and is provided solely as general information. Obviously nothing on this channel should be considered as personalized financial advice or a solicitation to buy or sell any securities. See our disclosures here: https://ritholtzwealth.com/podcast-youtube-disclosures/ Learn more about your ad choices. Visit megaphone.fm/adchoices

Welcome to The Compound and Friends. All opinions expressed by Josh Brown, Michael Batnick, and their castmates are solely their own opinions and do not reflect the opinion of Redholz Wealth Management. Today's live episode of The Compound and Friends is brought to you by Crane Shares and Rockefeller Asset Management, who are launching the ocean engagement etf that's ticker casey ksea which will invest in public companies with significant positive impact on oceans and ocean resources reflecting opportunities within the blue economy a subset of the ocean economy focused on solutions that are sustainable and have ocean positive benefits for For more information, go to Craneshare.com slash Casey. All right. Hi, everybody. Welcome to... There we go. I like that energy. There we go. Adam. I've been looking forward

to this all year. So great to see everyone. Fans of The Compound and friends, attendees at Future Proof, welcome to the show. This is a live recording. You'll be able to play this back on any podcast app that you choose. You'll also be able to rewatch it on YouTube or share with friends. You guys as the audience are actually part of the show. That's what separates this from a normal podcast. So feel free to get real loud, get real excited. Let's get rowdy. All right. And we are really lucky to have our guest

today is Rob Arnott. Rob is the founder and chairman of Research Affiliates, an asset manager focused on multi-asset, active equity, and alternative indexation strategies. Rob is a core portfolio manager of the PIMCO All Asset, All Authority, and PIMCO RAE Funds. Welcome, Rob, to the show. Thank you so much. And Rob is actually native to the area. You're in Orange County, yes? It took 16 minutes to drive here. It was really brutal. Not bad. Not bad. Not bad. Rob is, in my view, Rob is a legend in the industry. I've learned a lot about investing from him over the years. I remember attending some of Rob's events in New York City,

due diligence events, et cetera. Rob talks to financial advisors, family offices, the whole spectrum of asset management and wealth management. Thank you so much for being here. Michael, how did we want to start off today's show? All right. Last night in a bar in, I think it was Milwaukee, the owner said, being proximate to Green bay and the packers the owner said to his patrons if the jets lose drinks are on the house and so the jets are governed by murphy's law unfortunately

for my friends who are jets fans anything that can go wrong will go wrong and so when aaron rogers popped his achilles tendon The patrons decided to go after it. And we love our Midwesterners here. You guys like to drink. But the Jets won. And so the bill came due, like I said, Murphy's Law with Jets and Jets fans and everything like that.

Value investing in smart beta is not quite governed by Murphy's Law, but it's been a rough couple of years. I wouldn't say a rough couple of years. I'd say a rough 15. So as you all remember, we're at a financial conference here in 2013, 14, 15, right really almost up until the pandemic. Smart beta, factor investing, fundamental indexing, as Rob and his firm like to call it, that was a topic in and of itself. And conferences have gone away.

We're back. But so has this obsession with smart beta. Rob presciently wrote a post or paper, I should say. Excuse me. A post is way beneath you. Rob wrote an essay in 2016, How Can Smart Beta Go Horribly Wrong? And Rob, I'm going to read your intro and then I'll hand it over to you. Rob wrote an essay in 2016, how can smart beta go horribly wrong? And Rob, I'm going to read your intro and then I'll hand it over to you. You said, because active equity management has largely failed to deliver on investors' expectations, investors have acquired

a notable appetite for any ideas that seem likely to boost returns. In this environment, impressive past results for so-called smart beta strategies, even if only on paper, are attracting enormous inflows. Investors often choose these strategies as they previously chose their active managers based on recent performance. If the strong performance comes from structural alpha, terrific. If the performance is due to the strategy becoming more and more expensive relative to the market, watch out. That was 2016, Rob. Good call, Rob. What happened and where are we today?

Well, I wrote a paper early this year entitled, That Was Then, This Is Now, in which I pointed out that the vast majority of factors involved in factor investing are cheap now. Now, what do I mean by cheap? Let's use a simplistic example. Back in 2016, how can smart beta go horribly wrong? Suppose I'd written a paper in which I said if a stock doubles in price and its fundamentals haven't moved, its past returns will be brilliant and if there's any mean reversion, out it's going to disappoint what would be the reaction to that kind of paper this guy's a moron everybody

knows that but saying exactly that same thing about factors and strategies was suddenly wildly controversial. I don't get it. So what we found back in 2016 was that the great majority of factors were trading rich. What do I mean by that? Quality is always at a premium. But it can be at a small premium or a large premium. You look at its valuation, how cheap it is relative to the market, and compare that with its own history.

In the case of the quality factor, the norm is about an 80% premium. If it's trading at a 40% premium, that's a bargain. It means you have a possibility of a tailwind from quality getting more expensive. If it's trading at 150% premium, watch out. Well, back then it was 150% premium. So what we were looking at in 2016 was money pouring by the billions every single month into multi-factor strategies. And we're players in the multi-factor arena, so it's a little weird to talk, to speak ill about your own strategy suite. But our message was, you know, these mechanistic factors are getting popular enough that they're trading rich relative to

their own historical norms. If there's any mean reversion, watch out. By 2020, summer of 2020, everything was at massive extremes. Most factors were extremely cheap, but nobody would touch them with a 10-foot pole. Some were trading extremely rich, but most were extremely cheap. And almost all of them were either cheapest or richest decile of history in terms of rollout evaluation. or richest decile of history in terms of rollout evaluation. Today, you're looking at about 80% of the major factors that are out there are trading cheap relative to history. So now, when you don't have much money flowing into multi-factor, now's the time to embrace it. It's actually pretty cool. Why do you think it got so crowded? I felt as though factor investing had a boring name, and then smart beta was a little bit

sexier. It was easy. It was a catchphrase almost at all the conferences. Did that contribute to those factors trading rich, just people piling into the funds that were overweighting small cap, overweighting quality, overweighting value? Is that the story? over weighting quality, over weighting value. Is that the whole story? That's part of it. That's part of it. Now, smart beta, the term smart beta was coined by Towers Watson back in 2007 or 2008. They coined it based on fundamental index, the idea we introduced in 2004.

And why did they coin it? They realized that fundamental indexing has a reliable source of alpha that has absolutely nothing to do with fundamentals. Interesting. Fundamental index weights, chooses companies based on how big their business is, not how popular they are, not how expensive they are. How big is their business? How big is their economic footprint? And then weights them by that same measure. So you take growth stocks and you reweight them down to their economic footprint. You take value stocks, you reweight them up to their economic footprint.

You wind up with a stark value tilt. So early critics said it's just a clever repackaging of value investing. Well, they got one thing right, and that is its value. It has a value tilt. It always has, always will. But does the alpha come from the value tilt? No, it comes from having a stable anchor, the size of the company's business that moves much less than the price does, and uses that anchor to contra trade against the market's most that moves much less than the price does and uses that anchor to contra-trade against the market's most extravagant bets. So if the price soars and the fundamentals don't, Rafi will say, oh, thanks for that

high price, let me trim it. Now fundamental index doesn't say the price is wrong. The price might be right. The company is probably a terrific company with good products, good management, but it's in the price. Markets are moved based on narratives and those narratives almost always are in large measure true. That's the good news. The bad news is that the price already entirely reflects the narrative. So you don't make money based on narratives at all.

You make money based on narratives being wrong and identifying where they're wrong. So fundamental index doesn't identify which narratives are right or wrong. It just contrates against the constantly changing narrative and earns a rebalancing alpha. So the slide that's on the screen in front of us, we're looking at Raffi divided by the Russell 1000 value. And even though value stocks have had a difficult time weighting it this way. So I'm making this up, but let's say that Amazon, the market cap of Amazon is, or was, I don't know, three, four times ExxonMobil. They might have a similar economic footprint, but in the- Smaller.

There we go. ExxonMobil's bigger. But so you would take Amazon from whatever, six down to, again, I'm making this up, but it's directionally right, 2%. And so that's how you get the scenario where you're not necessarily saying, take all these growth stocks down to zero, the biggest names. Let's get them more in line with actually how big they are. So what's cool about this exhibit is the red line shows from 2007 to date what's happened to value. What a miserable time to be a value investor. And how on earth do we still have $130 billion when we've been in this kind of market for value investing? The red line shows what happened to value, down peak to trough 37%.

That's if you just waited on the value factor? That's if you just owned the Russell 1000 value and compared your wealth with somebody who bought the Russell 1000. Rafi, the ups and downs look just like the red line. Goes up and down just like the red line. During the global financial crisis, value was getting hurt. We were getting hurt during the snapback in the middle of 2009. Value rebounded slightly. We rebounded huge. And the reason for that is that we rebalanced,

contra-traded against the market. Market pushed value to extraordinary lows, so we went into the deepest value tilt we had ever had and got the snap back. Then value goes sideways for seven years, we go sideways to up a little. Then value crashes, we crash by less. Then value snaps back, we snap back by more. Then value tests its lows, we don't, we recede a little bit. Then value snaps back again, we snap back more. And net-net, value is still down almost 20% and we're up a thousand basis points. The green line shows how we do relative to the Russell 1000 value index. Now that relative performance is remarkable. That relative performance is 5,000 basis points added. And this isn't a back test. This is live indexes. 5,000 basis points

with no drawdown even as large as 2% and with no unsuccessful two-year spans. It's added value not every year, but every rolling two-year span, it's added value. 5,000 basis points with 2% volatility and with no drawdown as large as 2%. That's why we wrote the paper Rafi rocks. Wait, hang on. What do you mean no drawdown of 2%? Do you mean relative to something else? Relative to value. Okay. So listen, the proof is in the pudding, and I have an enormous amount of respect for you,

but I do have to push back against this one thing that you wrote. Rafi wins. I'm going to leave. Rafi wins by breaking the link. I'm going to leave too. By breaking the link between the price of a stock and its weight in the portfolio so that overvalued stocks are not a stock and its weight in the portfolio,

so that overvalued stocks are not automatically, and that's a key word, automatically overweighted relative to their unknowable future fair value and vice versa, undervalued stocks underweighted versus their future fair value. The part that I would push back against is why are the largest stocks in your estimation automatically, I know they're automatically overweighted, but why are the biggest stocks overvalued? I mean, Apple's a great example. That's not what I'm saying. So say more. What I'm saying is any stock that is overvalued,

that's trading above its fair value, and let's use as a proof statement what defines an overpriced stock, it's that it underperforms in the future. So any stock that's going to underperform in the future is accorded too much weight because you're linking the weight to the price. Now, indexers have heard that criticism ever since the dawn of indexing. And the pushback is you haven't said anything useful. That is a truism, but until you can tell me which stocks are overpriced and which are

underpriced, haven't said anything useful. Turns out they're wrong. And they're wrong for a very simple reason. If you break the link with price, then if a stock is overvalued, it might be overweight or it might be underweight. You've broken the link. You're not systematically overweighting the overvalued and underweighting the undervalued. You've broken the link. You're not systematically over-weighting the over-valued and under-weighting the under-valued.

You're breaking the link. And by breaking the link, roughly half of your portfolio will be over-valued, half under-valued. The errors cancel. Jack Traynor wrote a wonderful article in 2005, Why Valuation in Different Indexes Add Value. And what he pointed out is,

if price movement is 90% due to the fair value of a company going up or down, new news, new information, new products, new disasters, whatever. If they go up or down 90% due to genuine changes in fair value and only 10%, 10% due to the market getting it wrong. When the market gets it wrong, it's going to try to fix it. Eventually it will. And so that part of the price movement is mean reverting. If it goes too far up, it's a downward pressure. mean reverting. If it goes too far up, it's a downward pressure. So if you break the link with price, you're going to contra-trade against changes in fair value. Well, that's fine. The

fair value has changed, but you're also going to contra-trade against errors. And it's the contra-trading against errors that adds the value. So when Towers Watson coined the term smart beta, they looked around for other ideas that shared that attribute, not weighting companies according to price or market cap. And they found a bunch of them. Equal weighting, minimum variance, and so forth. They coined the expression smart beta. It's a clever nickname for a strategy, and everybody embraced it, including a lot of cap-weighted strategies, including a lot of trend-following strategies like momentum. And so all of a sudden, smart beta didn't mean smart beta.

It didn't mean a rebalancing alpha. It just meant something mechanistic, formulaic, whether it worked or not. So all of a sudden, smart beta wasn't smart beta. It was smart, stupid, and neutral beta all rolled under the same umbrella. Well, the market likes when you take something and turn it into almost like just an expression that people toss off. And it almost doesn't matter what it means at a certain point, because people are using that language to just express what type of investor they are. You don't hear a lot of people identifying themselves as smart beta now.

Right. And let's put up this chart, Mike, the multi-factor strategies. So this is… Go ahead. You got it? Yep. Okay. Walk us through what we're looking at here, Mike.

No, let's let Rob do it. This is his chart. Okay. All right. So this chart shows the relative cheapness of various factors on the horizontal axis, and this is in the form of a Z-score. For those of you who aren't quantitatively minded, Z-score just means how many standard deviations away from normal are you.

And so plus one means you're in the top 14, 16% of the historical distribution. Plus two means you're in the top 2.5%. And so you see a whole array of factors back then that were trading rich. You had a handful that were trading cheap. And what you found was that over the subsequent seven years, the ones that were trading rich generally did rather badly, and the ones that were trading cheap did okay. They didn't hurt you. So there was a weak linkage here. But low beta, which was attracting billions of dollars a month at the time, we got flack from one of our distribution partners

because they had a low beta ETF that was sopping up lots of money. And basically- Blink once if it was SPLV. I don't remember. It was something. But anyway, they were pissed off because we were basically saying, watch out, low beta is going to hurt you. And six months later, the market was up and low beta was down. So it didn't soften the upside, it downright hurt you. I actually got a call

from the CEO of that unknown firm at the end of the year saying, all right, I'm over it, you were right. But bottom line is when they get expensive, it's hard for them to add value. The relationship gets really strong when you have a time like summer of 2020, when the relationship was a correlation of 90% because you have two, three, four, five sigma outliers. Where are we now? You got the opposite relationship where most of the factors are trading cheap. Is this the best time to be looking again at value, small quality in a long time that you can remember? Wait, Josh, before we get there, I just want to rewind. So that low beta, the low volatility, I think a lot of that, and I'd be curious to hear your opinion before we get to where we are today. I think a lot of that was driven by people's desire for some sort of income because you couldn't get it with bonds.

And so, okay, Campbell Soup, Clorox, these are bond proxies. These are bond substitutes. Sure, I'll take some equity risk, but I'll clip my 2.5% coupon. And those companies, those strategies attracted a ton of attention. Obviously, interest rates are super important to invest, whether it's value or growth or anything in between. Do you think that's a fair explanation as to why those strategies were so popular and therefore underperformed? I think that's part of the reason they were so popular. The other part of the reason is that they had gotten, they'd attracted money, so their trailing performance was brilliant.

And the biggest mistake investors make is performance chasing. Past performance has nothing to do with future performance. He's the biggest chaser you've ever been on stage with. Whatever just worked last month, he's super bullish. So this graph is kind of interesting. The leftmost bar, that red band, goes from roughly 80 to 100. Excuse me. I can't see it. It's so far away. Anyway, it's all above one. That's because it's quality.

People pay a premium for quality. The white circle tells you where the pricing was as of late 2022. And that's relative to its own history. Relative to its own history. And its own history is spanned by the red band. So anything above that band is top decile. Really expensive. Anything below that band is bottom decile.

Really cheap. And this says quality in the US was pretty darn fully priced. Momentum was getting cheap. And this says quality in the U.S. was pretty darn fully priced. Momentum was getting cheap. Low volatility was getting cheap. Value was off the charts cheap. And as you look across, the bright red is U.S. large companies. The dark red is U.S. small. Blue is international. The dark red is US small. Blue is international. Beige is emerging markets. Value, you can see the rightmost band in each of those sections. Off the charts almost.

Off the charts in the US, off the charts in emerging markets, and OK pricing elsewhere. So this gives you a broad picture. Roughly 80% of the factors were trading cheap. Roughly half of the factors were in their cheapest quintile ever. RAOUL PAL, Do we need a catalyst for these value names to work? We've been having this conversation for a long time about how cheap they are. Do we need a catalyst or can investors' preferences just change? PEDRO DACOSTA, Yeah, could they stay cheap for 10 years absent any kind of, I don't want

to call it an intervention, but absent some sort of an event? How much longer could this go on for? What, like the Fed funds rate going from zero to five and a quarter? That can make a difference. The notion of a catalyst is a really fun parlor game. Why do I describe it as a parlor game? Because any catalyst that's big enough to move markets, by definition, is a surprise to the market. And so identifying a surprise to the market ahead of time is very cool and very difficult. You do need a catalyst. You don't have to know what the

catalyst is. But if I wanted to speculate on catalysts for value to win in the coming three to five years, one, off the charts cheap. Two, interest rates much higher than they were, hence discounting for future profit growth of growth companies means that their advantage relative to value diminishes. Three, inflation. Inflation, people have been popping champagne corks. We wrote a recent paper, basically hold the champagne, because what you're looking at in terms of inflation is that inflation has been tumbling this year to date due to a base effect.

Last year was two years in one. First half of the year, 6% inflation. Second half of the year, half a percent. So you were replacing, first half of the year, you were replacing a percent a month. Month after month after month. So we hypothesized, what if inflation matches the trailing three-year average, which was 47 basis points a month? You're replacing 1% with 0.47. That means it falls half a percent a month. Just year over year, just on comps? Yeah. Second half of the year, you're replacing a tenth of a percent with 47 basis points.

So it gets harder. Yeah. Second half of the year, you're replacing a tenth of a percent with 47 basis points. It gets harder. It gets much harder. And in fact, between now and year end, we have 0.2% total inflation that we're replacing. So is inflation going to be ratcheting up between now and year end? High odds. Not certain, but high odds. And all it has to do, if it rises, if you get 20 to 40 basis points a month, which is conservative, you'd get to four and a half to five and a half inflation at the end of the year. So my betting is four and a half to five and a half. I was talking to a Fed watcher who said that he expected year end inflation to be three and a half. I said

to get there you have to have almost have deflation between now and year-end are you sure about that? And he said well what's your number? And I said four and a half to five and a half. I said what's the market reaction gonna be if it's five and a half? And he said catastrophic. So why should going to be if it's 5.5? And he said catastrophic. So why should it be catastrophic if it's just a base effect and everyone should do the arithmetic and expect it? My own view is it won't be catastrophic, but it will break some of this current return of the growth bubble.

So I do think there's pretty darn good odds that value is coming back in a big way. I want to go there now, actually, because you have been on a one-man crusade to warn us all about the dangers of NVIDIA. Taking too much NVIDIA before bedtime, snorting NVIDIA. We are all to stay far away from NVIDIA. All right. I get the premise and you have done amazing work looking at prior tech bubbles and you have made the Cisco analog, which is a situation where a company actually does have earnings growth for 15 years and a stock price that basically never goes anywhere. So I understand the premise, but walk us through why NVIDIA has become so emblematic of this phenomenon to you in 2023. Now, firstly, the title of the paper was the

AI NVIDIA Singularity. And if you do AI Singularity or NVIDIA Singularity, it'll take you right to it. But the point of that paper wasn't that there's anything wrong with NVIDIA. It's a great company. They've got visionary product. They're just priced for beyond perfection. And so the title goes on as Breakthrough, Bubble, or Both. And cut to the chase, we say both. Now, that doesn't mean NVIDIA itself is going to fare badly, but you have so much of the market surrounding AI,

so much of the narrative surrounding AI, suggesting this is going to be huge, and it will be. This is going to... The technology itself. The technology itself. This is going to displace millions of jobs. And it will. It's not that anyone here is going to... Nobody here is going to lose your job to AI. If you lose your job, you're going to lose it to somebody who knows how to use AI better than you do. So get on board. Learn about AI. knows how to use AI better than you do. So get on board, learn about AI. But the issue isn't that the AI revolution is a fraud, the issue is the simple pricing.

Nvidia getting to 42 times sales, that's an extraordinary multiple. Now, our definition of bubble is really simple and can be used in real time. It's you would have to use implausible growth assumptions to justify the current price. And part B, a cross check on the first part of the definition, the marginal buyer doesn't care about valuation models. Is that true of NVIDIA? I think so. I think so. Today. So now not all bubbles pop. The exception that proves the rule is perhaps Amazon in 2000. I would have said that's a bubble. And it's performed brilliantly. But for the first 10 years, the decade of the aughts, it underperformed the S&P. It's only the

last dozen years that it caught up with the S&P and then soared past it. So bubbles don't inevitably burst, but they have very high odds of bursting. When I said implausible growth to justify current pricing, I didn't say impossible growth. I debated Cathie Wood. How'd that go? That was fun. And she's smart.

She's smart. She's been on this show before. Yeah, she's smart. She's very capable but i said i described that definition and i said so what are your growth assumptions that justify a target of 3 000 for tesla and she said well it's going to grow 89 a year in the next five years and then it'll be price peripasu with uh today's fang stocks I was asked by my host to play nice, so I didn't do what went through my mind, which was to say- What went through your mind?

That's 25-fold growth in five years. Amazon's grown 14-fold in 10 years. You're saying the one company will grow twice as much as Amazon in half as many years. So you're telling me there's a chance. will grow twice as much as Amazon in half as many years. So you're telling me there's a chance. Not impossible. It doesn't have to be impossible. But bubbles happen.

By the way, can we get this slide up while Rob's talking? Which one? The top 10 tech names of 1999. Yep. So I think this really illustrates what he's saying. Yeah, you look at the top 10 tech names of 1999. One beat the S&P, Microsoft. Between then and the end of 2022. Right. And it beat the S&P by 2% a year. So what are the other names on here just because we're audio? Microsoft. Cisco, Intel, IBM, AOL. Remember that? Oracle, Dell, Sun Microsystems,

Qualcomm, Hewlett-Packard. Three of them are gone, don't exist. Right. And nine of the 10 underperformed. But the point here is everything we were saying in 1999 that the internet could be came true and then some, but that did not necessarily translate to owning the top 10 companies that were at the forefront of that breakthrough. Is that? Brad Cornell coined the expression big market delusion. When you've got a big market on the horizon, a new market, an exciting market, investors will want to identify who's involved in that market, be it the internet, be it electric vehicles. Of course, place your

bets. Be it AI. Right. And when people place those bets, the tacit assumption is these current dominant players, they're all going to succeed. They're all going to win. But wait a minute, they compete against one another. They can't all win. And oh, by the way, there's going to be some competitors on the scene in 10 years that we don't even know about that might not even exist today. This is the question I wanted to ask you, though. Do things change? Because I could have said in 2017, Apple, Alphabet, Microsoft, Amazon, these companies now are all starting to compete with each other. They actually had people sitting on each other's boards, and that came to an end. I could have made the same case that, okay, everyone's now excited about cloud computing.

It's not possible that all these companies will win. And there's a big market delusion. Everyone is bidding up these stocks far in advance of what the reality is. Here we are seven, eight years later, every one of those stocks is now multi-trillion. They're competing with each other, but somehow record gross margins maybe take

out Amazon, if we're just talking about AWS. So have the FAANGs broken a lot of what we thought about investing? Have they broken- Fundamental investing. Have they broken fundamental investing because of how well run they are? And are they better companies than this list in front of us? And their margins have not been reverted a bit. That's correct. So firstly, remember my definition.

Implausible growth is needed. Back in 2017, you didn't need implausible growth assumptions to justify Apple or Microsoft's price. You needed aggressive assumptions, and those aggressive assumptions were, at least until recently, exceeded. So that's cool. Bubble territory would be, we wrote a paper, Electric Vehicles, The Big Market Delusion. And people focused on Tesla because it's the big player. Is that 2021?

Yeah. People focused on Tesla because that's the big player. 2021? Yeah. Okay. People focused on Tesla because that's the big player. But what was interesting was that there were nine electric vehicle specialists that only make electric vehicles. We looked at those nine. Tesla was 24 times sales. Tesla was the second cheapest on the list. Second cheapest. Right.

Not a list that Tesla normally makes. Right. Right. And so people said, well, you identified Tesla. And by the way, it's only underperformed a little bit since then. Yeah. But what about the other nine? The other eight, I should say. So nine out of nine underperformed. And it was in linear, almost exact linear proportion to their starting price to sales. There was one that had sales that were less than a million dollars that was priced at 10,000 times sales. Was that the one where they pushed the truck down a hill? Which one was that? Was that Nikola?

I think that was. that was. Okay. So Rob, do you think that our investors over indexing for the tech bubble, because it was so big and it caused so much damage and it's so vivid in our memories. There are a lot of differences between the type of companies, the Amazons, the Googles today versus the Sun Microsystems, all that sort of stuff. Is it too easy and too convenient and too neat to make that sort of parallel? Or is there wisdom in learning from what happened in the past? There's always wisdom in learning from what happened in the past, always. But there's also always wisdom

in keeping your hubris under control. The NVIDIA AI singularity. Do I think NVIDIA is a bubble and is likely to underperform? Yeah. Am I certain of that? No, I'm not. And do I have higher confidence that other AI companies will underperform, ones that are trading at even higher multiples?

Yeah. So I look at the AI singularity as path-breaking. I don't know how many people here saw the wonderful piece in Wall Street Journal, I think it was in the Friday paper, in which AI went up against MBAs in developing product ideas for a target demographic of college students that could be marketed for under 50 bucks. And MBAs at a university collectively came up with 200. Chat GPT-4 came up with 100. And then was trained further.

They realized they'd been unfair. They trained it further by giving it examples of successful past ideas. And they ran another 100 examples. So 400 product ideas. They did polls with students saying, would you buy this definitely or definitely not or somewhere in between? And so all of the products had a score. 40% of the MBA's ideas scored well enough that they might be an interesting product launch. 49% of the chat GPT.

But where it got interesting was in the top decile, the 10% most saleable products, 35 out of 40 were chat GPT. So that's terrifying because that's not clerical work. No. That's not a word calculator like an LLM. Why terrifying? Brilliant, wonderful. Okay, so say more. I have an opinion, but I'd rather hear yours.

Why is that not terrifying? I look on any technological innovation from the perspective of how can we use this? Whether you're talking about the- We could use this to terrify people. Yeah, yeah. The media will. Their job is to promote terror and anger. We love all our media guests.

Thank you so much for coming. That was Rob. Now, I would also say that you don't strike me as the kind who promotes terror. No. Be that as it may, every technological innovation from the loom to the telephone and telegraph to the auto, the airplane, the computer, the internet, they've all destroyed millions of jobs.

They've all created millions of jobs. They've all massively improved productivity. Look at the lifestyle that we lead today and compare it with the lifestyle of the affluent of 200 years ago. It's a horrible comparison, and today is really quite spectacular. Then ask yourself the question, how many of the jobs that computers eliminated are missed today? How many of the jobs that the internet eliminated are missed today? How many of the jobs that the internet destroyed are missed today? These technological revolutions destroy millions of jobs but they're jobs that aren't missed when they're gone

because people find other things to do with their time, often with the technology. The technology is empowering. Can an AI create a fundamental form of investing that none of us have conceived of yet that has market-beating returns so long as other people don't catch on to what it's doing? Short answer is absolutely. Slightly longer answer is I'm sure it already exists. So the quant funds that are playing with AI have probably stumbled upon some things that are

systematic. I would think more the super jumbo hedge funds, Citadel's on the other side of what, 10% of all trades that happen? Yes. I think they have a system, Rob. They have a system. They have a system. And I'm highly confident that they don't have human analysts deciding, oh, I'm going to do this trade. That's right. No. They have algorithms that are built on a foundation of AI that learns. AI is massively data hungry.

If you have thousands of samples of data, AI is useless. If you have millions, AI starts to become a little useful. If you have billions or trillions, now you're talking. AI can be very useful if you have vast data. This is one reason that I think that AI won't necessarily play a major role in long horizon investing, but will play a big role in short horizon trading and already does. Rob, the stuff that you talk about, the reason why it's so especially relevant today, breaking the link between price and weight. And I'm generally a wisdom of the crowd sort of guy. So I will defer to market cap. Now, obviously, at extremes, they exist. And maybe we're there today, maybe we're not. But undeniably, the biggest stocks are getting bigger as a percentage of

the overall weight of the index. The top seven are now 27% of the weight. They contribute. They've driven a lot of the returns year to date. Are we reaching, are we nearing a tipping point? Because these companies have economically outperformed even high expectations. They just continue to find a way to pull a new lever, create a new category. Where are we today? You buy when you're at peak fear. You worry when you're at peak enthusiasm and peak optimism. Or peak certainty, even. Or peak certainty. Yeah. Good choice of words.

This is an environment right now in which you have huge enthusiasm, huge optimism, very little fear. And I view that as a wonderful time to rebalance into the segments of the market where there is fear. time to rebalance into the segments of the market where there is fear. Would you say with the 10-year trading where it is, is this one of your biggest surprises of 2023 that the NASDAQ 100 is up 40 some odd percent on the year? Yeah. I did not have that on my bingo card. Well, I did, but I didn't share it with anyone. And that's, I bought NASDAQ. It was on one of your bingo cards. Well, so if somebody had set up the year for you this way, you sat down with someone on Jan 1, and they said, these two things are going to happen. The Fed is going to go to 5.5% Fed funds rate very rapidly, multiple 75 basis point rate hikes

from nowhere. We're going to get there. And 45% NASDAQ rally led by Tesla, Apple, Google, Amazon, those two things would seem highly incongruous just on their surface because the narrative about why those stocks were working for so long was money's cheap, Tina, no alternative, got to do something. So that bet would have gone very awry to say these stocks will be victims of higher rates. This is why hubris is so dangerous in our business. Because you're right.

I would have said those two realities don't generally coexist very happily and don't exist for very long. So I would draw some comfort from the latter. But, you know, chat GPT really did upend a lot of people's perceptions of AI. AI has been around forever. I was using neural networks in the late 80s. But it's become interactive in a fashion that people who aren't remotely tech savvy can use AI in fascinating ways. My son, who's in a totally different business from me, said last November,

Dad, you've got to try this chat GPT thing. And he sent over a picture, and I said, what is it? He said, I'm not sure it was chat GPT. It Um, and he sent over a picture and I said, what is it? He said, I, uh, I'm not sure it was chat GPT. It might've been Dolly. Anyway, he, he tasked it to, um, uh, draw a picture of a, an ayahuasca temple based on a, that's really specific. Wait, why don't I just tune that out? Based on a dodecahedral structure in a forest. And then he tasked it to draw a floor plan. And it did. And it did all of that in five seconds. So I went on and I, with Dolly, I tasked it to paint the Brooklyn Bridge in the style of Andy Warhol. My wife is a serious enthusiast and collector of contemporary art and a big fan of Warhol.

And I took the pictures that came out, and I sent them over to her, and I said, South Bay's is having an auction. Southbees is having an auction the Warhol archives have just released some pieces that are new that were in the archives and this looks really interesting do you want me to bid on one of these

and she looked at it and she said the second one is fantastic I want that and I said well it was just painted by a computer. It took five seconds. Yay. Psych. It's a buy-in video. Wow. So you won husband of the year shortly after. Oh, I win that daily. I mean, she's the only one I know who kicks me.

Fair enough. So given that you see the potential for this technology, is there a universe where we look at a company like Meta, which is about to release Lama to the public in an extraordinary way, almost akin to what we saw with ChatGPT last November? have 2 billion users of Facebook's platforms doing whatever they want with very sophisticated AI within the next couple of years. Is that the universe in which meta being up 150% year to date actually makes sense or are we still subject to this big delusion and there's a trap door somewhere that none of us are thinking about? I would say yes and yes. When I say you'd have to make implausible assumptions, I don't mean impossible. Those assumptions may be correct.

But there's also the issue of upstarts, new companies. There was a confidential memo that was circulated from within Google that somehow leaked, where one of the senior executives said, we have no moat. NVIDIA has no moat. Anyone can come into this space. And that's true. NVIDIA has no moat? Can you think of another company that can manufacture GPUs at scale today? David Morgan No moat doesn't mean that it happens tomorrow. David Morgan Okay, eventually.

David Morgan And also, think in terms of job displacement. My wife and I were on a three-week trip to Europe in May, and we played the thought experiment as we met people along the way, asking, will this person have a job in five years? And we didn't ask the people involved. Were you allowed club med? We didn't say, are you going to have a job in five years? But, you know, driver from the airport, maybe.

Yeah. The greeter at a hotel, the bellman, the waiter at a restaurant, the bus boys, the chefs, their jobs aren't at risk. About 80, 90% of the jobs out there aren't at risk. Financial planner. Financial planner, not at risk from the human element of financial planning. Breathe out. And here's where it gets interesting. Can AI do a better job of managing money than the average financial advisor?

Perhaps I would even verge on saying in reasonably short order in the next few years, probably yes. But can they relate to the clients? I think the most important job a financial advisor does is to educate their clients and steer them away from horrific blunders. I've often said if a financial advisor has zero alpha for their clients, they're doing them a huge service. Agreed. And they're doing them a huge service because they're dissuading the client from having a minus 500 basis point a year negative alpha from performance

chasing and blunders and fads and you name it. I think that's a really nice place to leave it. Rob Arnott, ladies and gentlemen. Bye.