IL54: Why Smart Investors Keep Making Crazy Decisions ft. Alex Edmans

Alex Edmans · London Business School

National mood swings from World Cup exits move the UK stock market about 0.5% the next day.

Football results are an exogenous mood shock that should not affect profits, dividends or inflation, except in a few sectors like replica merchandise. Edmans found that when a country is knocked out of the World Cup, its stock market falls significantly the next day, about half a percent after controlling for world market moves. Applied to the UK market, that is over £10bn wiped off in a day. He concludes that professionals are not immune to emotion, so investors should guard against these extraneous factors. He cites the England-Argentina defeat, where the UK market opened lower despite good growth news, partly attributed to the negative sentiment.

“when a country gets knocked out of the World Cup, the market falls significantly the next day. To put some numbers on it, it's about half a percent even when you control for the broader world market movement.”at 4:05

Alex Edmans · London Business School

Playlist happiness tracks national mood and is significantly correlated with stock prices.

Mood is hard to isolate from fundamentals, so Edmans measured its effects instead of its drivers. Behavioral congruence means people play music that matches how they feel. Using a valence score for song happiness (not lyrics) from Echo Nest, he aggregated the country's top streamed songs into a national mood gauge. That gauge was significantly correlated with stock prices, and negatively correlated with government bond prices. He explains the bond link as optimism pulling money from safe assets like gilts into equities.

“we used that algorithm to look at how happy or how sad a country's playlist was as a reflector of national mood. And what we showed was that was also significantly correlated with stock prices.”at 8:45

Alex Edmans · London Business School

Momentum works across assets and countries because investors underreact to news.

Momentum buys stocks that did well over six months and shorts those that did badly. The mechanism is underreaction: good news is only moderately priced in, because confirmation bias leads people to downweight news that challenges their view. Edmans says momentum has been found over any reasonable period, not just in the US but around the world, and in bonds, currencies and virtually any asset class. That breadth points to a common psychological driver rather than a local quirk.

“people have found that momentum typically outperforms. And interestingly, this is not just in The US, it's around the world. It's not just in equities, but it's in bonds and currencies and virtually any asset class”at 31:22

Alex Edmans · London Business School

High VIX reliably signals momentum underperformance, so it indicates when to exit momentum.

Momentum loses badly and can crash. Edmans cites research by two professors who also worked as investors at AQR and Goldman Sachs. They found that when the VIX volatility index is high, momentum typically underperforms. That gives a practical signal for when to get out of momentum. He concedes it is not a perfect predictor, since momentum can still crash unexpectedly. He links this crash risk to why momentum persists: managers fear the crash, so they under-exploit it.

“when VIX, which is the volatility index is high, momentum is typically going to be underperforming. So this tells you when to get out of momentum. Now, it is not a perfect predictor”at 32:11

Alex Edmans · London Business School

AI looks like a bubble, but he won't sell AI stocks and risk missing the boom.

Edmans says he knows there is an AI bubble, yet he won't sell his AI stocks. If the bubble keeps inflating, he doesn't want to be sitting out of AI and miss the big boom. He ties this to the risk in a bubble: if the market then turns south, as it did for electric vehicles in 2022-2023, the whole market could fall. His point is that an active manager who sees a bubble still holds positions to avoid career risk from underperforming, and that conviction is what clients pay for.

“So even if there's an AI bubble right now and I know there's a bubble, I don't want to sell my AI stocks... if the bubble continues to inflate, I don't want to be the Muppet [sitting] out of AI and missing out on this big boom.”at 23:24

Alex Edmans · London Business School

AI hyperscaler valuations look like a big-market delusion; not every hyperscaler will win, so he cut AI exposure.

The 'big market delusion' is when excitement about a new technology lifts every company in the industry so far that the summed valuation is implausible, because not every company can win. Edmans says AI will be the future, but not every AI hyperscaler will win, and some may over-invest or invest in the wrong things. Applying the framework to hyperscalers, he argues investors may be pricing them as if all will succeed. He cut his AI and tech position six to nine months ago, partly to lock in gains that funds a mortgage refinancing about a year away. He is unsure whether the sale will prove right long term. His concern is a bubble that turns south as EVs did in 2022–2023.

“Not every AI hyper scaler will win. It may well be that some of them ended up over investing or investing in the wrong stuff.”at 56:54

Alex Edmans · London Business School

Higher active share predicts outperformance; benchmark-hugging funds just overcharge clients.

Active share measures how far a fund deviates from its benchmark. Academic research nearly twenty years ago found it was generally correlated with outperformance. Deviating reflects managers using their specialist knowledge, such as buying firms with strong culture or intangibles. Funds that hug the benchmark are not using any expertise and are simply charging active fees for index-like returns. He warns that the measure can be gamed, since trading just to raise active share adds no value, so the real test is whether active positions match the claimed expertise.

“this active share was generally correlated with outperformance. So funds which were having conviction to deviate from the benchmark... the ones who are inactive, who are hugging the benchmark, are the ones who are not using any expertise at all, they're just overcharging you.”at 24:00

Alex Edmans · London Business School

Stocks' equity premium is too rich; it should be about 3-4%, so stocks offer good value.

Stocks have beaten T-bills over any reasonable period. Over the ten years to December 2025, T-bills averaged about 2% a year and the S&P 500 about 15%, a gap of about 13 points, though other periods show about 7%. Some argue that gap is compensation for risk. Edmans says that when reasonable risk-aversion parameters are put into the numbers, the equity premium should be about 3-4%, not 7%. The conclusion is that the stock market over any reasonable period offers good value.

“the equity premium should be something like three or 4%, not 7%. So the stock market over any reasonable period does offer good value.”at 27:52

Alex Edmans · London Business School

The early-2020s EV industry was overpriced: combined valuations of EV makers exceeded the traditional car market.

As an example of the big market delusion, Edmans adds up EV makers such as NIO, Tesla and Xpeng and says their combined valuation, about five years ago when EVs were still nascent, was even higher than the traditional car market. The market priced the industry as if every company would win, which is impossible. He attributes this to fear of missing out, in that investors did not want to hold only NIO and Tesla in case BYD took most of the market. The EV market later turned south in 2022–2023.

“If you were to take the value of the electric vehicle industry, adding up NIO and Tesla and Xpeng and so on, you got a combined valuation which was even higher than the traditional car market.”at 56:03