How the Markets Lost their Predictive Power — ft. Aswath Damodaran | Prof G Markets

Gold's enduring demand stems from properties that made it work as a store of value across cultures and millennia—compact, transportable, hard to destroy, malleable—and its pricing is partly emotional (people will hold gold even down 50% because they like its feel); whether Bitcoin has this cross-cultural, cross-time emotional connection is unproven, since its demand currently comes almost entirely from having done well over one decade.

“gold has endured... It was compact enough that you could transport it. It was very difficult to destroy. You could make it into different things.”

The 'big market delusion' is a recurring pattern where a buzzword-driven big market (PCs in the 1980s, internet in the 1990s, online advertising and social media in the 2010s, now AI) attracts overconfident entrepreneurs funded by overconfident VCs, each overestimating their own success, so the summed value of all players vastly exceeds what the market can deliver—leading inevitably to a correction, yet this overreach-and-correction cycle is how change happens and is ultimately healthy.

“It's part of what I call the big market delusion. I've written about it... It's because I've seen this movie before.”

Portfolio managers and investors have fundamentally different missions: a portfolio manager judged year-to-year on returns would be irresponsible not to hold Nvidia/AI because omitting it creates relative disadvantage, whereas a long-term investor's only responsibility is to preserve and grow their own wealth, which may mean staying out of overpriced spaces—neither approach is better, just different.

“if you're going to be judged on an endowment fund or a pension fund on the returns you make each year, not having Nvidia will put you at a disadvantage”

The search for 'multi-baggers' (stocks that go up tenfold) is one of the most dangerous things in investing because making the big hit your entire focus skews your whole investment process toward big bets; it is healthier to let multi-baggers happen 'accidentally'—Damodaran's Nvidia and 1999 Apple wins came from buying undervalued companies expecting only ~5% over-market returns, not from hunting the next Microsoft or Amazon.

“the search for multi-baggers, which is the stock which will go up tenfold, is one of the most dangerous things in investing”

Bubbles are healthy and necessary because overreach is how human beings advance—people overreach, correct, feel pain, and clean up, but the world changes as a consequence (the dot-com bust still transformed how we live), so Damodaran is 'an optimist in bubbles' rather than wishing for a world run by risk-averse actuaries.

“Would you want to live in a world run by actuaries? I'm okay with bubbles. I'm okay with overreaching.”

Investing and trading are fundamentally different: investing assesses TAMs, revenues, and cash flows to ask if a price is reasonable for the business; pricing/trading is about demand and supply, mood, and momentum—and roughly 95-99% of money currently in AI is pricing and trading, where investors don't care about the TAM because they treat it as a 'pass the cushion' game where flipping to the next buyer earns money even if wrong.

“Investing is about assessing TAMs and estimating revenues and cash flows... Pricing is about demand and supply, more than momentum.”

AI is changing the scam game and 'upping the ante': scams now use deepfaked likenesses (Damodaran's own video promoting Palantir and Nvidia to recruit a fake investment club promising 60-80% returns), and scams function 'like a tango' requiring both scammer and scammed—so scams won't disappear until the scammed recognize their own role; the target audience is people who don't do homework and want to invest in a name.

“AI is changing the scam game. It's upping the ante.”

Markets are driven by the economy and earnings, not talk; what would actually shake markets is real evidence of economic slowdown (turning negative/recessionary) translating into earnings damage—likely showing first in third-quarter earnings—because the link between economic data and earnings has weakened, making the adjustment gradual rather than sudden.

“Ultimately, markets are not driven by talk. They're driven by the economy and earnings.”

Big tech companies are 'money machines' that gain a relative advantage as uncertainty rises, because their large cash reserves let them act quickly and spend tens of billions without needing external capital, outcompeting rivals precisely when conditions worsen.

“as you up the uncertainty, you actually give these companies an advantage over the competition”

Meme stocks have broken the normal link between price and value—where value usually acts as a gravitational force pulling price back—and instead become 'unmoored,' driven purely by demand, supply, mood, and momentum; once a stock goes there, its underlying business no longer matters.

“they've broken away from the gravitational push of pull of value. Usually we have price and value value operates as some kind of force pushing price back. With meme stock that link is broken.”

Markets have shifted from being a proactive 'prediction machine' to a reactive 'wait-and-see' stance because economist and market-guru predictions have lost their accuracy in an interconnected, chaotic global economy where tiny inputs produce large, unpredictable outcomes—so the market rationally declines to predict until data confirms.

“have you seen the quality of predictions from economists and market gurus for the last decade?”

The independence of the Fed is partly an illusion that has been 'leaking out' for 20-30 years; Jerome Powell could not do what Paul Volcker did in 1981, when no senator or House member publicly demanded he stop—reflecting that US political consensus and institutional robustness have eroded since the late 1970s.

“The notion that the Fed was independent to 2025 is an illusion... Jerome Powell could not have done what Paul Volcker did in 1981”

A CEO's real job is to be the company's storyteller who frames and sells a compelling narrative—not to fix supply chains—which is why Jeff Bezos drove Amazon's narrative for decades and why Google suffers from lacking a storyteller (Sundar Pichai being a good operator but not a narrative-setter), and why $30B of compensation may be justified to keep Elon Musk as full-time Tesla storyteller.

“That's a CEO's real job. It's not to fix supply chains. It's to be the storyteller for the company.”

A good collectible needs three things—scarcity (Bitcoin has its 21M cap), enduring demand (gold has lasted thousands of years), and acting as a hedge that holds value when financial assets collapse—but Bitcoin fails the hedge test: in Q1 2020 the S&P fell 33% while gold held and Bitcoin fell 56%; over the following six months stocks rose ~50% while Bitcoin doubled, showing Bitcoin behaves like a very risky stock, not a collectible.

“good collectibles operate as hedges against financial assets... when stocks and bonds go into collapse, a good collectible holds its value”

Nvidia is the 'lazy investor's answer' to AI exposure because it is one of the few companies actually making money on AI (mostly architecture); but like Cisco in the dot-com era—which was not the biggest winner (Amazon was)—the architecture companies will recede into the background while product and service companies move to the foreground, so Nvidia is unlikely to be the ultimate AI winner.

“When all was said and done, Cisco was not the biggest winner from the dot-com era. It was amazon.com. I'm waiting to see who the amazon.com of the AI boom will be. But I don't think it's going to be Nvidia.”

Equity markets have already been pricing in US institutional weakening for 7-15 years, which is why there is no massive market reaction in 2025; it is the experts who are stuck in a pre-2008 worldview and being dragged 'discontinuously' to recognize what markets already absorbed gradually.

“equity markets have been treating the US more like the rest of the world for the last 7, 8, 10, even 15 years. And that's why you're not seeing the massive reaction in markets”

The US historically received a 'buffer' allowing it to run trade and budget deficits other countries couldn't, because it was the largest economy with the largest financial market—but 2025 has removed the illusion of that slack, and the US will increasingly be treated in markets like the rest of the world.

“if any other country had run the trade and budget deficits that the US had, they'd be treated as a basket case... We've been doing it for 40 years”

Strong corporate governance should not be seen merely as a check on founders but as a mechanism that brings founders back to reality when they get distracted; OpenAI would be a better company if Sam Altman were more open to people who disagree, which is what a board with different viewpoints provides.

“people often think of strong corporate governance is putting a check on founders. I don't think of it as that way. I think it brings founders to reality when they get distracted”

The best AI companies (OpenAI, Anthropic, SpaceX, Stripe, Databricks) are staying private and have no trouble raising capital, but many are 'corporate governance nightmares'—which may be a reason they stay private since it lets them do things public companies couldn't—and this lack of accountability, even to investors, risks them not growing in healthy ways and eventually biting them.

“Many of them are corporate governance nightmares. Which might be one reason they're staying private”

Cash is held by companies as a shock absorber for safety and stabilization to get through bad times—never to earn high returns—so moving corporate cash from liquid, near-riskless treasuries into Bitcoin is a category error, because Bitcoin is too volatile to serve cash's stabilizing purpose even given its strong historical returns.

“cash was never supposed to have that purpose of delivering high returns. It provides safety. It's a shock absorber.”

Bitcoin can only be priced and traded, not valued or invested in, because its price depends on getting ahead of the demand-supply game; getting companies, mutual funds, and endowments to buy Bitcoin (especially post-2024 election expectations of more buyers) is 'insidious' because with fixed supply, increased demand mechanically pushes price up—but this trading thesis can fall apart fast since a 30% drop dissipates the demand.

“There's an insidious effect of getting companies to invest in Bitcoin... If they invest in Bitcoin... Increase the demand side of Bitcoin, and since the supply is fixed, the price goes up.”

Google/Alphabet trades below the S&P 500 average PE multiple despite strong AI, YouTube, and Waymo growth because a decade of its 'bets' (Nest, Google Glass, the six other businesses) being cash drains with little bottom-line return has made markets skeptical and lowered expectations; if Alphabet delivers on those bets it could beat lowered expectations, making it Damodaran's preferred Mag 7 entry point.

“if I were an investor and I've never held any of the Mag 7 stocks and I want to enter, Alphabet would be my choice”

For Nvidia to justify a $4.4 trillion valuation, AI architecture spending would need to reach $2-3 trillion, which in turn requires AI products and services to be worth $12-15 trillion to make the money back—a market size Damodaran does not believe exists, making Nvidia's valuation hard to justify on economics.

“if you spend $3 trillion on AI architecture, AI products and services have to be $12 trillion, $15 trillion in value to make your money back. I think AI is great, but I don't see the market as that big”

Since 2008, country risk has shifted from a binary (safe developed markets vs. risky emerging markets) to a continuum because globalization means every country's risk spills over into others, and the US has moved along this spectrum—no longer truly safe but not as risky as Brazil—evidenced by losing its triple-A rating across S&P (2011), Fitch (2023), and Moody's (May 2025).

“the US has lost its triple A rating. First with S&P in 2011, then with Fitch in 2023, and finally with Moody's in May of 2025”

A catastrophic economic event would not crash markets in a single day but unfold over weeks in alternating good and bad news, because markets are highly adaptable and can switch from reactive to predictive 'overnight'—unlike experts, markets have no egos and don't cling to methods that stopped working.

“I don't think it'll happen in a day. It'll happen over a series of weeks”

Big VC investors in large private companies are passive and do not demand accountability—they put money in and hope to profit from exits in 2-5 years—so they are not asking the questions needed to make these companies truly great.

“VC investors, especially in these big companies, are passive investors. They take their money, they put it in, and they hope to make their money from exits.”

Bitcoin fails as a currency because it is inefficiently designed (requiring thousands of miners to guess a nine-digit number just to buy a coffee), and despite 15 years of effort, actual transactional use to buy and sell things remains minuscule—so advocates have shifted to calling it 'the new collectible' / digital gold.

“if it takes a thousand Ukrainian miners guessing a nine-digit number for me to buy my Starbucks cappuccino, it doesn't work for me”

CFOs are terrible traders who lack a sense of market tops and bottoms and tend to trade at the wrong times, so letting them move corporate cash from T-bills into Bitcoin (or equivalently Beanie Babies or Picassos) is a bad idea; even a Bitcoin believer should prefer the company return cash so they can buy Bitcoin themselves rather than have the CFO do it.

“CFOs are terrible traders. They don't have a sense of what the bottom is, what the top is. They often trade at the wrong times.”