The Equity Premium Puzzle - Why Stocks Beat Bonds by So Much
Stocks make more money than bonds. Everyone knows that. But here’s the thing. They make way more money than bonds. And when economists tried to explain why the gap is so big, they couldn’t. Not with their standard models. Not even close.
That gap has a name. It is called the equity premium. And Chapter 18 of Burton and Shah’s book is all about why it is so weirdly large that it became one of the most famous puzzles in finance.
The Numbers That Started It All
In 1985, two economists named Rajnish Mehra and Edward Prescott decided to look at 90 years of stock market data, from 1889 to 1978. They calculated that the average real return on stocks (meaning after inflation) was about 7 percent per year. The average real return on safe short-term Treasury bills? About 0.8 percent.
So the equity premium, the extra return you get for holding stocks instead of safe bonds, was about 6.18 percent per year. For 90 years.
Now, nobody was surprised that stocks earned more than bonds. Stocks are riskier. You should get paid more for taking that risk. The surprise was how much more.
Mehra and Prescott plugged the numbers into standard economic models. These models use something called a utility function to estimate how risk-averse people are. Based on previous research, economists believed the risk aversion parameter should be somewhere between 0 and 10. Being generous about it.
With those assumptions, the maximum equity premium their model could produce was 0.35 percent. Not 6 percent. Not 3 percent. 0.35 percent. That is almost nothing compared to the actual 6.18 percent they found in the data.
So either people are absurdly, irrationally risk-averse (someone joked you would need a risk aversion parameter of 55 to make the math work), or something else is going on. That something else is the puzzle.
And it is not just an American thing. Mehra came back in 2003 and confirmed the same puzzle shows up in the UK, Germany, Japan, and France. The gap between stock returns and bond returns is too large everywhere.
Loss Aversion: The Behavioral Explanation
Remember prospect theory from earlier chapters? The idea that losing $100 hurts about twice as much as gaining $100 feels good? That concept is loss aversion, and it might be the best answer we have for this puzzle.
Think about it this way. On most trading days, stocks go up a little. But not all days. Some days stocks go down. And if you are loss-averse, those down days hurt disproportionately. The pain of watching your portfolio drop on a bad Tuesday might outweigh the pleasure of watching it go up on a good Monday, Wednesday, and Thursday combined.
So here is what happens. Imagine two investors. One checks her portfolio every 10 years. She almost never sees a loss. Stocks basically always go up over 10-year periods. She sleeps fine.
The other investor checks his portfolio every five minutes. He sees losses constantly. Small ones, sure, but losses. And each one stings. Over time, the accumulated pain of all those small observed losses becomes unbearable. He moves his money to bonds. Peace of mind restored.
But here’s the thing. Not checking would have made him a better investor. The stock portfolio was going up over time. He just couldn’t handle the emotional ride along the way.
If most people behave this way, if they avoid stocks because of loss aversion rather than rational risk assessment, then fewer people own stocks than should. And when fewer people want to buy stocks, stock prices stay lower than they should be. Lower prices mean higher future returns. And that is your equity premium.
Burton and Shah note that this explanation makes a lot of sense. It is well supported by psychology research. But traditional economists resist it. Why? Because loss aversion breaks the utility functions that most of economic theory is built on. Accepting loss aversion as the answer means admitting that a huge chunk of existing finance theory needs reworking. That is a tough pill for academics to swallow.
The Survivor Bias Argument
There is another explanation, and it comes from looking beyond America.
In 1999, Jorion and Goetzmann pointed out something important. The U.S. stock market has been one of the best performing markets in the world for over a century. But if you are a truly rational global investor, you should own stocks from everywhere, not just America.
And when you look at everywhere, the picture changes. Many countries had stock markets that performed much worse. Some countries had stock markets that literally stopped working during wars or revolutions. If you grew up in the Soviet Union like I did, you know that markets can simply disappear.
Jorion and Goetzmann looked at global data from 1921 to 1996. The median real return across all countries was only 0.8 percent per year. Compare that to the U.S. market’s 4.3 percent over the same period. America was the exception, not the rule.
So maybe the equity premium puzzle is partly a survivor bias problem. We are looking at the winner (the U.S. market) and saying “wow, stocks are amazing.” But we are ignoring all the markets that crashed, closed, or just limped along.
However, even Jorion and Goetzmann admitted this does not fully solve the puzzle. When they built a GDP-weighted global portfolio, it still returned about 4 percent real. Smaller than the U.S. number, but still a big premium over bonds. The puzzle shrinks but does not disappear.
Other Explanations
A few more ideas have been thrown around.
The future will be different. Some argue that the high equity premium is a historical artifact. Future stock returns will be lower because of resource limitations, slower growth, or structural changes in the economy. Essentially a modern version of Malthus saying we will hit the limits of growth. Maybe. Nobody knows.
Status quo bias. This one is interesting. Studies of retirement accounts show that people rarely change their initial asset allocation. If you start your 401(k) and put everything in safe bonds because you are 22 and don’t know better, you will probably still have everything in safe bonds at 55. You just never get around to changing it.
Agnew, Balduzzi, and Sunden looked at 7,000 retirement accounts over four years and found most people went all-in. Either 100 percent stocks or 100 percent safe assets. And once they picked, they almost never switched. The “path of least resistance” kept them stuck.
But this explanation has problems too. You are only looking at one retirement account. The person might have other investments, a spouse with different allocations, inherited assets. One account does not tell the full story.
Are Stocks Always Best for the Long Run?
Jeremy Siegel wrote a famous book called “Stocks for the Long Run” arguing that over periods of 17 years or more, stocks have never lost money after inflation in U.S. history. So if you are a long-term investor, stocks are actually the safest choice for preserving purchasing power.
And the math backs this up to some degree. If stocks return 12 percent with 16 percent variance and bonds return 4 percent with no variance, the probability that stocks underperform over one year is about 31 percent. Over 5 years, it drops to 13 percent. Over 20 years, just 1.3 percent. Over 40 years, nearly zero.
But economist Zvi Bodie made a smart counterargument. He asked: how much would it cost to buy insurance against stocks underperforming bonds over long periods? The answer is that the insurance cost approaches the entire value of your investment. Why? Because while the probability of underperformance gets tiny, the potential size of the loss gets enormous. A low probability event with catastrophic consequences is still expensive to insure against.
So “stocks always win in the long run” is true most of the time. But it is not a law of nature. It is a historical observation that could break.
Still a Puzzle
So where do we stand? Loss aversion is probably the most compelling explanation. It accounts for why people avoid stocks beyond what rational risk aversion would predict. But accepting it means traditional finance theory needs serious revision.
Survivor bias explains part of it. The U.S. market has been a winner, and looking only at winners inflates the numbers. But even globally, the premium persists.
And maybe future returns will be lower. Maybe status quo bias plays a role too.
But as Burton and Shah conclude, the equity premium puzzle remains a puzzle. We know stocks have beaten bonds by a huge margin for over a century. We know the standard models cannot explain why the margin is so large. We have good guesses. But no definitive answer.
And that is honestly kind of refreshing. Not everything in finance has a clean solution. Sometimes the market just does something that nobody can fully explain, and the honest response is to admit it.
Previous: Calendar Effects in the Stock Market