
Stock Market Bubble Expert Warns "Things Could Get Bloody Fast" | Jonathan Treussard
What this covers
Adam Taggart interviews Jonathan Troussard, a former partner and head of product at Research Affiliates who now runs his own firm and wrote academic work on bubbles at UCLA. The conversation begins with a market that has risen roughly 25% in four and a half months and asks whether that is a new era or another round of exuberance. Troussard places the market late in a bubble, and much of the hour covers how it got there and what could end it. He points to liquidity that has expanded since the October 2022 low, even though quantitative tightening was nominally in force. He argues it was effectively put on hold when regional banks failed, a claim marked as contested. The market has also survived Silicon Valley Bank, recession forecasts and commercial real estate stress, but he treats each escape as a postponement.
The middle of the discussion turns to Nvidia as the emblem of the AI trade. Troussard works through a price-to-sales ratio near 30x. He questions whether Nvidia can stay a monopolist and argues that training data loses value once every major player has it. He also says the company has high real economy beta, so AI spending would be cut when payrolls come under pressure. His account of the mood is existential FOMO), a fear of missing a world-changing shift. Taggart adds the Tesla drawdown and a comparison with dotcom-era corporate spending. Nvidia's 50% fall during the 2022 crypto winter serves as an example of narrative risk.
The macro portion is where the claims become more contentious. Troussard says inflation is stuck in an elevated regime, that elevated inflation also means more volatile monthly prints, and that uncertainty about prices works like a wage cut. Taggart argues the lag effect of rate hikes has been delayed rather than repealed. Troussard also covers mortgages as a short call option, why a return to 3% mortgage rates would be a bad sign, and how low implied volatility leaves options sold in buy-write and structured note products underpriced. He also criticizes asset managers for selling whatever investors want.
Troussard argues that US stock markets are in a late-stage bubble (4/5 severity) driven by liquidity conditions and multiple expansion rather than earnings fundamentals, with valuations at 97th percentile historically leaving investors vulnerable to sharp corrections if economic conditions deteriorate or Fed policy tightens further.
- US large-cap valuations at 97th percentile of Cape data since 1880 provide minimal margin of safety
- Liquidity has risen since October 2022 bottom supporting both economic growth and asset prices, but this may reverse as Fed emergency programs end
- Market has dodged multiple bullets (SVB, regional bank failures, recession predictions) but lag effects from rate hikes remain unresolved
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
High inflation regimes are also high inflation volatility regimes, meaning month-to-month inflation prints will be more volatile once inflation is elevated, making it harder to discern true disinflation trend
“a high inflation regime is also a high inflation volatility regime that's just the nature of it so once inflation is high then you know the month-to-month prints are just going to bop around just a lot more”
It is irrational to expect Nvidia will remain a monopolist in AI chip markets long-term, yet the current valuation pricing assumes this, making sustained profitability at current price-to-sales levels unlikely
“I think it's irrational to think that Nvidia will forever be the monopolist in this market that's number one uh and clearly part of the excitement is it's the only game in town today but is it really going to be the only game in town in the next 10 years that seems fairly unlikely”
US large-cap stocks are currently trading at the 97th percentile of all Cape ratio data going back to 1880, representing extremely stretched valuations relative to historical norms
“the US market and I'm talking about us large caps uh are currently trading at the 97th percentile of all Cape data going back to 1880”
Tesla's decline from $400/share in late 2021 to $178 in present demonstrates that even companies with real competitive advantages and massive addressable markets can experience severe valuation corrections when expectations prove unrealistic
“Tesla hit a a high of over 400 uh bucks a share right at the end of 2021 it's now trading at about 178 so it's it's been more than cut in half over the past two plus years”
AI regulation is currently minimal in the US, representing best-case pricing scenario, and it is unrealistic to assume this permissive environment will persist indefinitely
“there isn't a lot of Regulation certainly not in the US uh that's been put in place and so this is kind of like best case scenario pricing where you know most anything is on the menu um it's hard to imagine that we're going to be living in a no AI regulation world forever”
Nvidia had a 50% correction when the crypto/NFT narrative shifted to 'crypto winter' in 2022, demonstrating the stock is vulnerable to narrative shifts even with strong underlying demand
“Nvidia had a major major correction the last time around when people went from uh you know crypto and nfts for all to crypto winter very quickly in 2022 uh I don't remember exactly the numbers but some version of of down 30% or down 40% seems yeah I'm pulling it up it's it's hard to see right but if you go to like if you just go to that Peak around the number 2022 what is that three 330 call it 330 went down to yeah right so so that's that's down 50”
Data is only valuable if it is gated (exclusive) to the data owner; once all major AI companies have access to the same training data, the marginal value of that data rapidly decreases because competitive advantage is eliminated
“the only way in which data truly are valuable is to your point if they're gated uh and if everybody cuts the check and everybody learns on it then the marginal value of that of that training ground is is very quickly backed down to kind of you know the Foothills because um it's it's now part of the bloodstream”
Quantitative tightening (QT) was effectively put on hold in March 2023 when regional banks failed, despite official communications suggesting continued tightening, leaving markets with an impression of tightening when liquidity has actually been expanding
“operation whatever tightening at large whether it's quantitative or real uh was put on hold back in March of 2023 when those Banks went under and so it's just we've been kind of floating out there with with the impression that we're tightening but not really really”
Historical data on market valuations demonstrates empirically that when valuations are high, prospective returns are lower, a relationship that does not depend on theory but is an established empirical finding
“we know uh empirically and again as you noted I'm a PhD Economist and I've and I've you know thought long and hard about both the the theory of Finance as well as with the data inform us um this does not require Theory this is a a a an empirical result which is when uh valuations are high uh prospective returns are lower period”
Fresh government-backed mortgages at 6% are functionally equivalent to being long a 10-year Treasury plus selling a call option, because if rates decline, homeowners will refinance, capping upside for mortgage investors
“like a fanny mortgage um a fresh one you know one that was just issued recently at some version of 6% right uh is effectively functionally identical to being along a US Treasury roughly a 10-year treasury uh and selling a call on it uh because what you know is going to happen is if rates come down right people will refinance their mortgage that's the prepayment risk”
Nvidia has a price-to-sales ratio around 30x, meaning investors would need to wait approximately 30 years without any capital expenditure or operational activity for the company to generate enough free cash to return their initial investment
“you know price to revenue there's you know the number is some version of 30 uh which and probably actually higher now which means that what you would have to do for you know as an investor to get your money back out of this thing uh would be to stop all activity uh all capex and and so on and so forth for the next 30 Years”
Asset managers have perverse incentives to sell whatever products investors demand regardless of whether those products are in investor interests, as evidenced by product development at major Wall Street firms
“when I was head of product at research Affiliates we went by our mission which was to do things for the benefit of investors so we often had requests for new products and new strategies and we would sit around and say but do we believe this isn't the benefit of investors and if we said no we would just say like I don't really care that there is you know you know money in the banana stand so um but that's the way a lot of product groups around Wall Street in particular in Asset Management operate if there is money to be made I don't really think long and hard so to speak about what it means for you know whether or not this is an advisable product”
The lag effect from rate hikes (where increased cost of capital takes time to impact economic activity) has not been repealed but merely delayed through liquidity-supportive Fed policies, and will eventually arrive as these support programs end
“I like to use the term lag effect you know we've had this this we've dramatically increased the cost of capital and yes that's been cushioned by a lot of net liquidity factors um including things that the FED has been in charge of like the reverse repo program and the the btfp um both of which by the way are are kind of ending at this point right running out running out”
A decline back to 3% mortgage rates (which happened before) would be a negative signal for the broader economy, making this a scenario investors should not necessarily want to occur despite the prepayment upside
“I don't know that we're going back to a 3% mortgage world and by the way I'm not sure we want to go back to that world because that's that means a lot of bad things have happened”
In corporate environments today, just as in the 1990s dotcom era, there is pressure to invest heavily in the new thing (AI) even without clear ROI because nobody wants to be the company that 'missed' the transformation; this drives corporate capex allocation toward speculative AI investments.
“there's probably a lot of that with AI right now which is hey you know seems like that's where it's going we better we better invest big in AI now even if we don't fully understand it because we just don't want to be the dinosaurs in the story”
The market has been walking a fine line, successfully avoiding multiple potential catastrophes (recession expectations from early 2023, regional bank bankruptcies including SVB, ongoing commercial real estate problems) but each avoided crisis doesn't eliminate underlying risks—it merely postpones them.
“the reality of it is the US economy is strong and that has caused I think you know some of the runway we've seen in markets it's stronger than we all anticipated certainly about you know 15 months ago when we went into 2023 with you know near 100% probabilities of recession and so in a lot of ways we've dodged a lot of bullets and by the way some bullets we hadn't seen come along uh you know before they hit us things like uh the bankruptcies of of of regional Banks and and and Silicon Valley Bank”
The current market froth is driven partly by existential fear (EOMO - existential FOMO) where participants fear they will be left behind if they don't position themselves for a world-order-changing technological shift, rather than purely from greed or simple FOMO
“forget fomo maybe you and I are coming up with a new acronym eomo which is you know existential fomo uh and I think in a world in which you're looking at again think about everything we've lived through in the last you know Call It Whatever Whenever you start the clock but let's just call it covid okay and you're saying this world is changing radically and in in ways that I don't love and don't care for and you know I've been put through the ringer uh and now there seems to be this new thing that may be taking uh you know the economy and and the world in which I operate in a completely different uh Direction and and and I think it's reasonable to again I I'm trying to be educated by what I learned from looking at the big bubbles in history some times the social order changes and when it does it people want to make sure they're on the right side of the Ledger”
Liquidity has been recovering since October 2022 (when markets bottomed), and rising liquidity has driven both economic growth and financial asset price appreciation; however, there is major uncertainty about how long this liquidity expansion can continue given forward monetary policy dynamics.
“on a net basis liquidity has been r in all through then that it really bottomed in October of 2022 which is right when the markets bottomed and as liquidity has continued to increase uh so has uh economic growth and so has uh so has uh financial asset prices”
Current implied volatility on US stocks is very low (around 30th percentile historically), making options that investors are selling through buy-write and structured note strategies underpriced relative to the downside protection being purchased
“implied volatility on US Stocks is really low right now you know we can talk about percentile some version of into 30 percentile going back in history but the point is these options you're selling you're selling for cheap on a thing that leaves you with a bunch of downside and not a lot of upside”
The US economy has been stronger than anticipated, particularly relative to recession predictions that had near 100% probability at the start of 2023, which explains some of the market runway since October 2022
“it's stronger than we all anticipated certainly about you know 15 months ago when we went into 2023 with you know near 100% probabilities of recession”
Commercial real estate sector contains significant unrecognized losses on the books of financial institutions, and despite Fed backstop of the banking system, some of these losses will eventually spiral and trigger economic disruption
“I would be shocked if this you know we had heard all we're going to hear uh on the commercial real estate set of things just think there's a lot of losses on a lot of books uh and uh and yes the FED has basically just back stob the entire banking system for the last year um but at what point does you know one of those thing you know spiral again”
Current inflation data suggests we remain stuck in an elevated inflation regime despite Fed's 'mission accomplished' rhetoric, with real-time price pressures (gasoline prices) continuing to rise
“I think it's hard to to look at the recent inflation data and not think that we're kind of stuck here um we all know the things that are going to wash out eventually if we're lucky you know the Oar the uh you know the rental equivalent so on and so forth um but I don't know about you I drive past a gas station every day and it's pretty clear that you know oil is is going up and and gasoline going into you know the tanks of your neighbor's cars is going up”
Nvidia has significant 'real economy Beta' meaning its revenues and profitability are highly correlated with overall economic cycles, and when the economy slows and corporates face payroll pressures, AI capital expenditure will be deprioritized
“I believe that Nvidia has a huge amount of what I would call real economy Beta And what I mean by that is not their correlation to the stock market but they're correlation to the real economy and if and when because we know these things will happen when the economy really slows down and takes a hit and you know corporates are wondering you know how to make payroll or certainly where to cut my suspicion is you know us corporates will will worry about those things before they worry about investing in AI applications”
Inflation uncertainty functions as a wage cut even when inflation hasn't yet occurred, because consumers worried about future purchasing power make conservative spending decisions, creating real economic headwinds despite nominal wage growth
“perspective inflation is the fact that I don't really care what my wages are today I'm looking at my paycheck and I'm worried that I don't know what it's going to buy me in a year uh and so in a real way uh no pun intended um inflation uncertainty is like a a wage cut right now now even though it hasn't happened”
The narrative around Tesla shifted from pure EV superiority to questioning actual vehicle efficiency and environmental benefits, and the market is now reconsidering whether EVs or hybrids represent the optimal solution, suggesting similar narrative reversals could occur with AI hype
“the bloom is kind of coming off the rose of the EV Market you know I'm not here to to to denigrate it but um there's definitely been a lot of questions that have been asked about um you know do these cars perform as well as we thought uh are they as efficient as we thought are they as green as we thought and there's a lot of arguments to suggest that they're not”
Troussard's non-financial investment recommendation is to regularly assess oneself on multiple dimensions of well-being (beyond wealth), following Martin Schwarzenegger's multi-dimensional life assessment method, to ensure balanced life development
“I uh someone mentioned to me Martin Schwarz multi-dimensional whatever life assessment method um you know comedian uh and I think you can Google the Satur night life comedian Martin yeah yeah yeah and basically the St and and I've been doing it for myself and I've I've found it very helpful”
NVIDIA shareholders are capturing more than 100% of value creation from the AI boom so far, whereas in prior technology revolutions (1995-2000) value was distributed across many players, indicating extreme concentration of gains
“what we're seeing is NVIDIA shareholders capturing more than 100% of all gains from so far first approximation”
Some of Nvidia's premium pricing may reflect a risk premium for the possibility that US export restrictions prevent the company from accessing the Chinese market, reducing addressable market and justifying lower valuations
“I think some of it is a premium related to the idea that these chips shouldn't be finding their way into China but there is no way they're not uh the data go of you know Import and Export going into China suggests otherwise”
Reddit's IPO was driven by AI companies like Google licensing Reddit's archive to train models, but this is a one-time revenue event that cannot be replicated indefinitely, yet markets are applying high multiples as if revenue growth from licensing will persist
“the reason why you know Reddit became red hot and wanted to IPO to capitalize on it was um I can't remember which AI player um but maybe it was Google I think approached them and said hey we want to basically license your your archive right we want to we want to we want to train our AI on your your you know gazillions of posts from regular people right it's ton of content that that we're going to learn from right and okay just taking that at face value I get it but that's a one-time gain to Reddit when you talk about how real are these Revenue this Revenue growth right that's a one-time hit”
AI is genuinely transformative and will change reality, but is currently being sold with Fantastical expectations that exceed near-term practical capabilities; the hype has reached levels where everyone pitches 'AI solutions' that can theoretically 'do anything' rather than solving specific business problems.
“someone was pitching him an AI solution for his company and he said well great what does it do uh and the pitch was well I can do anything you want it to do and you're like how about you build a product that actually helps me in my job and then you come back”
Bubbles are wealth redistribution mechanisms where some participants gain and others lose, making the key question not whether a bubble will form but rather who will be holding assets when it deflates
“bubbles are a wealth redistribution mechanism there are people that lose money and people that make money and that's the bottom line it's just the reality of it um and so I think part of it is like who's gonna by the end of this thing who's going to end up with more chips and who's going to end up with fewer”
Buy-write strategies (selling covered calls on stock positions) have become popular through ETF formats specifically because they create high distributions that appear attractive, but this obscures the fundamental trade: giving up upside in exchange for downside protection at unattractive prices
“buy right strategies hugely popular uh in you know kind of wealth management at large uh has been made more popular by the fact that a lot of it has been going into ETF format recently and and sucking up a lot of AUM uh what's a buy write strategy you buy the stocks and you write you sell call options on top of it to you know Spike the the distributions”
High valuations universally predict lower prospective returns, but Troussard acknowledges humility about timing: he cannot say when the top will be hit or when a substantial correction will occur—only that the environment is not conducive to deploying fresh capital into U.S. equities at current prices.
“I'm not here to you know uh bead the drum that that's it we've seen the top I have no way of knowing that um but I'm not excited about putting more money in US stocks at these prices”
Nvidia's stock price increased from ~$150 at start of 2023 to ~$940 currently, driven primarily by hyperbolic revenue growth, but the sustainability of revenue growth at historical rates is questionable given competitive market dynamics
“the revenues I mean the revenues really have gone hyperbolic there's no question the point is uh you know are the revenues you know sustainable and real and not real in a accounting fraud sense of the term as as in you know reliable as a go forward indicat um and and that's the challenge”
People are experiencing fatigue from the past 4+ years of uncertainty and volatility (COVID, inflation, banking crisis); this intellectual and emotional fatigue makes it hard to process information clearly and is causing confusion in markets.
“if you're tired that's okay it's been a minute uh and as a result of that uh I think a lot of people are trying to process information in real time while you know Running on Fumes intellectually”
Risk has been realized for many US households over the past four years (since COVID), causing them to adopt prospective rather than retrospective thinking about economic conditions
“risk risk has been realized for a lot of people over the last four years and as a result of that there there thinking about it on a prospective basis”
Reddit data quality issues ('garbage in, garbage out') mean that training AI models on Reddit's user-generated content may produce models that learn from low-quality or problematic human behavior, reducing the value of the training data
“there's the concern I have again maybe knowing Reddit a little bit better than the average person of kind of garbage in garbage out right I wonder like what is your AI Gonna Learn like I prefer not to consume some of this stuff uh and uh whether that's in a direct fashion or indirect fashion through some kind of you know machine learning you know on on it”
Investors considering structured notes should ask: 'Am I excited about earning an extra 4% on top of treasury yield in exchange for selling market insurance at low volatility prices?' The answer should be 'no'—especially in environments where markets are stretched and tail risks are elevated.
“are you excited about the idea of earning an extra 4% on top of a really healthy treasury yield for the benefit of selling Market Insurance uh and again at fairly low prices today because implied volatility is low”
We are in a 'late-stage market bubble' where either a substantial price correction or prolonged stagnation is necessary for valuations to revert to reasonable levels
“late stage Market yeah late stage market bubble meaning we're getting close to the point where the gains we've seen are not going to be able to continue and really one of two things are going to have to happen either a a pretty substantial correction to get prices down into better value or a prolonged stagnation”
Troussard does not believe this is the beginning of a market run ('third inning') but rather that much of the move has already occurred and valuations have largely priced in future gains
“you have to look at things from the perspective of the structure you have um and the structure we have which are things like historical data and and market valuations and you know all of that um aren't suggestive that this is uh the beginning uh they're suggestive that you know a lot has already happened and a lot is already in the price”
Troussard believed soft landing was credible at beginning of 2023 but current evidence suggests the scenario has shifted toward 'no landing' (no recession but no slowdown) which is an unusual and unsustainable economic configuration
“I thought uh at the beginning of this year that you know the the predominant argument which was soft Landing you know had a lot of credibility and I think it's very clear that the cards at a been turned over since you know are taking us away from some version of of soft Landing to you know kind of this no Landing World which is a really bizarre world”
If the Fed chooses to maintain its current messaging about inflation being solved and resists tightening further, it could surprise markets that have priced in current funding rates, causing both market and economic shocks
“if if somehow someone gets a little bit more of a of a a spine uh over the next few months and say we're no we're not quite there yet then it could the market could be in for for a real surprise and therefore the economy could be in for a real surprise because you know I think we've all made peace with this funding rate for a period of time”
On Troussard's bubble meter (1-5 scale), current market conditions rate 4 out of 5 in bubble severity, with the caveat that the Japanese 1980s stock bubble reached 5 or higher, leaving room for further appreciation despite extreme valuations
“on a meter uh you know one through five or something I think it's a solid four I think it's hard I think it's hard to say everything points to Bubble I think there some things don't uh but the predominance of the evidence suggests that this Market isn't priced in a way that leads to really great outcomes for the next 10 years”
Bond interest rate volatility is very high while stock volatility is very low; this creates an environment where selling stock options (buy-write) is bad while buying bond exposure (mortgages) is relatively good, creating an asymmetric opportunity set.
“stock market volatility implied volatility is very low right now so it's not a great time to be selling options all over town um bond interest rate volatility is very high”