The recent announcement of the 2013 Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel has delighted researchers working in empirical asset pricing. The combination of deep economic insight and clever methodological contributions that Eugene Fama, Lars Hansen, and Robert Shiller have brought to this field has revolutionised our understanding of the determinants of asset prices.

A brief note is in order here about why understanding asset price movements is critically important to a broader understanding of the functioning of the economy. Predictions of asset price movements are key inputs into a variety of calculations, including, but not limited to:

  • Whether pension liabilities are appropriately funded;
  • Firms’ capital budgeting and capital allocation decisions;
  • Evaluations of asset managers’ performance;
  • Whether firms should be takeover targets; and
  • Whether governments are solvent or insolvent.

So the foundational work that the laureates have done has a broad range of important real-world implications.

Eugene Fama

Back in the 1960s, economists were putting the finishing touches to the theory of competitive markets, and financial markets were viewed as being as close as possible to the theoretical ideal of such a competitive market. The perception was that financial markets had very little in the way of frictions impeding rapid, profitable trading in the event of dislocations between prices and relevant information – a notion given status and a vocabulary (“an efficient capital market”) by Fama in a series of papers in the early 1970s (see, for example, Fama 1970).

This notion was not merely theoretical – an early empirical tour de force in this literature is Fama et al. (1969). This paper pioneered an approach to empirical asset pricing that persists to this day, blending theoretical concepts, innovative statistical approaches, and the analysis of a large dataset of financial market prices to arrive at clear conclusions. The paper used stock splits – a frequent corporate action of listed firms – to provide evidence that financial markets appeared to satisfy the theoretical concept of efficiency. The methodological contribution of the paper was to provide an approach – used in innumerable subsequent papers – known as the ‘event study’, in which frequently occurring phenomena in capital markets can be analysed before and after an ‘event date’ with appropriate adjustment of price movements to account for variation across stocks in their level of systematic risk. It is worth pointing out here that like many Nobel laureates, Eugene Fama’s subsequent contributions to both methodology and empirics in finance have revolutionised the way we think of the field – and singling out just a few of these does not do justice to the huge and influential body of work that he has produced.

Robert Shiller

A consensus emerged steadily in the empirical asset pricing literature through the 1970s that stock markets did indeed fit the description of the canonical efficient market. Many of these studies utilised high-quality data on individual stocks, which had first been compiled by the Center for Research on Security Prices at the University of Chicago in an immensely useful and prescient exercise that has served subsequent generations of researchers extremely well. The reliance on data using individual stocks, and the method of adjusting their prices against movements in an aggregate index (as the most frequently utilised model of risk-adjustment, the capital asset pricing model, recommended) somewhat obscured a broader issue, which was that the index of all available stocks itself exhibited unusual movements from the perspective of a rational model. This is where the rightly famous paper of Shiller (1981) contributed hugely to the ongoing debate.

Shiller’s contribution, explained after the fact (like most seminal contributions), appears to be simplicity itself. Since stock prices in an efficient market should correctly capitalise all future cash flows derived from owning the stock, with the benefit of hindsight it should be possible to construct an ex-post measure of whether stock prices actually did this. In other words, as an observer of stock markets today, I can easily check, given sufficient data, whether stock prices in the 1970s were justified by the future realised cash flows to the stocks from 1970 onwards. I can do this for every historical date available. If stock prices are high in a rational market, this must mean that future cash flows are expected to be high as well. Conversely, falls in stock prices should be justified by future falls in cash flows. Of course, the key here is the discount rate applied to these cash flows, and Shiller begins with an extremely simple assumption, namely that the discount rate is constant over time. His conclusions can be summarised in two simple pictures, which are reproduced below:

Figure 1. Figures reproduced from Shiller (1981)

In each of these pictures, Shiller takes a well-known stock index (S&P Composite on the left, DJIA on the right), and plots it alongside the ex post rational price, that is, the present value – using a constant discount rate – of the subsequent realised dividends of the stocks in these indexes. The difference is astonishing – clearly stock prices (the highly volatile series in each plot) are way too volatile to be justified by subsequent realised dividends. Something doesn’t compute, says Shiller in the paper, since the only way to justify this is by very substantial variation in discount rates. He states that the required order of magnitude of variation doesn’t show up in nominal interest rates, and hence is unlikely to be in expected real interest rates. He also presciently forecasts a burgeoning line of subsequent research by mentioning (and providing useful reasons for potentially being sceptical of) the possibility that the market may greatly fear an enormous catastrophe.

Shiller finally arrives at the conclusion that the efficient markets hypothesis appears indefensible if required compensation for risk (in the form of the discount rate) varies so greatly over time. He subsequently refined this intuition in combination with his student and frequent co-author John Campbell, in an important series of papers which decomposed unexpected stock returns into ‘discount rate news’ and ‘cash flow news’ using a clever methodology which has subsequently been applied to a range of different assets (Campbell and Shiller 1987, 1988).

Lars Hansen

Lars Hansen arrived at a similar conclusion about the difficulty of explaining stock prices with fundamental information, using a newly developed methodology, different data, and a different modelling approach. Hansen’s starting point, along with Kenneth Singleton, was to empirically analyse the basic consumption-savings decision of a hypothetical capital market investor in the most elegant and plausible setting, and in the simplest possible fashion (Hansen and Singleton 1983). Having invented an important estimation methodology – Generalized Method of Moments, which makes the empirical analysis of such ‘non-linear rational expectations’ models tractable (Hansen 1982, Hansen and Singleton 1982) – he and Singleton applied it to the joint analysis of macroeconomic aggregate consumption data and stock market prices. In an exercise mirroring the conclusions of Shiller, they found that, while real interest rate variation lines up well with consumption growth, risky asset return variation is not well explained by the basic macroeconomic model. The parameters of the model are very noisily estimated when risky assets are included in the estimation, meaning that support for the model is elusive in the data. These insights, as well as Hansen’s many subsequent empirical, methodological, and theoretical contributions, have provided a major impetus to an extremely active area of current work in empirical asset pricing which goes informally by the name of ‘macrofinance’. This broad area attempts to connect asset price variation to macroeconomic variables through improvements in empirical work as well as through the creation of new models.

Conclusions

Together, the laureates have greatly improved our understanding of the underpinnings of modern asset markets. The blend of rigorous statistical methodology, deep connections with economic theory, and a healthy respect for ‘market wisdom’ demonstrated by these economists has provided a huge impetus to one of the most important and active areas of current research in economics. The answers from this research are continuing to help us solve important problems that affect all of our lives in important ways. It is clear, especially at the present time, that better understanding of the vagaries of financial markets can help us make the best possible financial decisions, and teach us how best to formulate appropriate financial regulation; these are essential factors for the proper functioning of the economy as a whole.

References

Campbell, John Y and Robert J Shiller (1987), “Cointegration and Tests of Present Value Models”, Journal of Political Economy 95: 1062–1088.

Campbell, John Y and Robert J Shiller (1988), “The Dividend-Price Ratio and Expectations of Future Dividends and Discount Factors”, Review of Financial Studies 1: 195–228.

Fama, Eugene F, Lawrence Fisher, Michael C Jensen, and Richard Roll (1969), “The Adjustment of Stock Prices to New Information”, International Economic Review 10(1): 1–21.

Fama, Eugene F (1970), “Efficient Capital Markets: A Review of Theory and Empirical Work”, Journal of Finance 25(2): 383–417.

Hansen, Lars P (1982), “Large Sample Properties of Generalized Method of Moments Estimators”, Econometrica 50(4): 1029–1054.

Hansen, Lars P and Kenneth J Singleton (1982), “Generalized Instrumental Variables Estimation of Nonlinear Rational Expectations Models”, Econometrica 50(5): 1269–1286.

Hansen, Lars P and Kenneth J Singleton (1983), “Stochastic Consumption, Risk Aversion, and the Temporal Behavior of Asset Returns”, Journal of Political Economy 91(2): 249–265.

Shiller, Robert J (1981), “Do Stock Prices Move Too Much to be Justified by Subsequent Changes in Dividends?”, The American Economic Review 71(3): 421–436.

2,940 Reads