Tuesday, March 11, 2014

Asset Prices in a Lifecycle Economy

I have just completed a new paper on asset prices, "Asset Prices in a Lifecycle Economy".  The paper is available here from the NBER or here from my website.  This is a good time to comment on asset price volatility and the apparently contradictory findings of two of the 2013 Nobel Laureates because my paper sheds light on this issue. 

In 2013, Gene FamaLars Hansen and Robert Shiller, shared the Nobel Prize for their empirical analysis of asset prices.1 

Fama won the Nobel Prize for showing that financial markets are efficient. He meant, that it is not possible to make money by trading financial assets because markets already incorporate all available information. 

Shiller won the Nobel Prize for showing that financial markets are inefficient. He meant that the ratio of the price of a stock to the dividends it earns, returns to a long run average value; hence, an investor can profit by holding undervalued stocks for very long periods. 

These apparently contradictory results are consistent  because Fama and Shiller are referring to different concepts of efficiency.

When Fama says that financial markets are efficient, he means informational efficiency. There is a second concept that economists call Pareto efficiency. This means that there is no possible intervention by government that can improve the welfare of one person without making someone else worse off. The fact that markets are informationally efficient does not necessarily mean that they are Pareto efficient and that fact helps to explain why financial markets appear to do such crazy things over short periods of time.

Long-run predictability is not the only feature of asset prices that is hard to understand. Economists also have a hard time explaining why asset prices are so volatile (the excess volatility puzzle) and why the return to equity has historically been six percentage points higher than the return to holding government debt, (the equity premium puzzle).

My new working paper explains both these asset pricing puzzles.  I don't need to assume that there are financial frictions, sticky prices or irrational behavior.  The only thing that is different about my work from standard macroeconomic models is birth and death. Rather than assume, as do many models, that there is a single representative agent in the world,  I assume instead that people are born, they work, they retire and they die. 
Figure 1 Source "Asset prices in a Lifecycle Economy", (c) Roger E. A. Farmer
Figure 1 compares simulated data from my model (left panel) with US data (right panel). The blue line in both cases is the safe rate of return and the green line is the stock market return. 

In the simulated data, like the actual data, the stock market is risky with a higher average return than a safe short asset.  What explains my results?

Financial markets are like insurance markets. If you own a house you will insure the house against fire. And since not all houses burn down at the same time, the premiums we all pay for fire insurance can be used to compensate the unlucky few who suffer from a loss in any given year. But to benefit from insurance, you must purchase the insurance before your house burns down.

Oreopolous and coauthors have shown that people who start their working life in a recession are worse off for their entire lives than people born in a boom. Given the opportunity, we would all choose to purchase insurance over the state of the world into which we are born, for the same reason that we buy fire insurance on our house. I call our inability to purchase this insurance, the absence of prenatal financial markets. It is the inability of the young to trade in prenatal financial markets that explains why the financial markets are not Pareto Optimal.

In my model, huge inefficient asset price fluctuations occur because the unborn are not around to profit from them. Risky assets trade at a premium because retirees have no other source of income and cannot afford to gamble away their savings.  If our children's children's children could trade in the markets, they would short stocks that are overvalued, and buy those that are undervalued. But for those of us with finite horizons, life is too short to make those trades. As Keynes quipped; Markets can remain irrational for longer than you or I can remain solvent. 
1. Lars Hansen won for his work on estimation and I will not have a lot here to say about his contribution. It has already become a part of the curriculum for every Ph.D. student in economics.  


  1. Eventually, I'll try to read your paper. But initially my thought is it's not just that different cohorts have different risk-aversions at their current life stages. Even within a cohort, it doesn't look hard to show irrational, or at least extremely uninformend and/or unexpert, behavior. Even within, say, the young or middle-aged cohort, I don't think it would be hard empirically to find a large percentage of investors not jumping all over, with the liquid money they did have, the realtively low P-E ratios in early 2009 – and instead doing the exact opposite, fleeing.

    1. I agree. But simple models don't always capture every element of reality.

  2. "But for those of us with finite horizons, life is too short to make those trades. As Keynes quipped; Markets can remain irrational for longer than you or I can remain solvent."

    So, are you agreeing, basically, with Siegel's premise of stocks for the long run, that if you can take a long-run view you can get better risk-adjusted returns (or much better)? This would imply that the young should hold a relatively higher percentage of stocks (which some like Bodie seem to disagree with). And the same would hold true for money the old want to leave as a bequest (as well as for other long-lived funds, like university endowments).

    1. Richard
      Yes, stocks for the long run is clearly a winning investment strategy for the young

  3. Stock prices fluctuate because market makers and insiders play games to sell high and then buy low again and there is nothing else into it. You can try to fit to a model as many auxiliary hypotheses as you like but the truth of the matter is that facts are over hypotheses and the stock market is a game of wealth redistribution from speculators to equity issuers and insiders. Trying to look at stock prices from an economic perspective is, in my humble opinion of course, fallacious thinking. Many will get prizes while the facts are constantly denied. Therefore, since you handle the math extremely well, why not building a model in which there are four agents: the equity issuer, the well-informed insider, the uninformed speculator and the trader/gambler. You will find out that your simulation will also fit reality.

    Why the market has been going up the last year? Because companies bought back their stock primarily. At some point, those insiders and issuers with dump the pumped stock on the face of the speculators and create another huge swing, like the past two, in 2000 and 2007.

    Microfounded Economics or Macrotheory? http://t.co/J3uvOg6cv8

  4. Link to David Glasner should be http://t.co/J3uvOg6cv8


Note: Only a member of this blog may post a comment.