Divining stock price forecasts

Written by Madhav Raghavan | Updated: Oct 16 2013, 09:35am hrs
It is hard to imagine that people can be awarded prizes for saying opposite things, harder still to picture them sharing the prize for it. Yet that is exactly what has happened in economics this year.

The 2013 Nobel Prize for economics has been jointly awarded to three peopleEugene F Fama, Lars Peter Hansen (both of University of Chicago), and Robert J Shiller (of Yale University)for their empirical analysis of asset prices. Fama and Shiller are known for their financial market theory, while Hansen made his name developing econometric tools that have been widely used in studying asset prices, among other things. This split is in keeping with the recent trend of jointly honouring theoretical and empirical contributions to the field.

In the aftermath of the financial crisis, the study of asset prices is particularly important. Everybody agrees that the crisis was exacerbated by the complexity of the assets being traded. Even now, nobody knows how to accurately price a collateralised debt obligation (CDO) or a mortgage-backed security (MBS). Even though they generate relatively high returns, the markets for these products are a fraction of the size they once were, because nobody trusts their valuations any more. Forecasting their prices is a fools game.

Yet simpler assets like stocks and bonds continue to be traded widely, and in large volumes. Some of this is due to confidence: investors have more faith in how stocks and bonds should be priced, and in forecasts. This faith comes from extensive researchin a large measure based on the works of this years laureates.

Eugene Fama is mainly known as the proponent of the Efficient Market Hypothesis (EMH), which asserts that since markets are efficient, there is no investment strategy that is guaranteed to make money via short-term trading.

In particular, the EMH says that since anybody can become an investor, financial markets are fully (and fiercely) competitive. In such a market, everybody brings their information to be priced in, and so todays opening price of a stock reflects all information up to today. And so, todays price changes must reflect only todays news, which is inherently unpredictable. Stock prices therefore follow a random walk, which cannot be forecasted in the short term.

Instead, if price movements were predictable, a rational investor would be able to make a killing with this information. This seldom happens, except for cases involving insider trading, where some traders are made privy to news before the rest of the market. This work has been so influential that even now the phrases investors are rational and markets are efficient are benchmark assumptions for analysis of financial markets.

Robert Shiller, on the other hand, initially rose to prominence based on his work that disputedeven contradictedthe EMH. Markets are not particularly efficient, he argued. Since the amount one pays for stock is essentially related to the future returns expected from holding the stock, it follows that there should be a relation between the stock price and its discounted future dividends. If markets were efficient, the relation should be more or less constant, since important news would affect the share price only by affecting the prospects of future company profits. But Shiller demonstrated how stock prices are far more volatile than dividend payments, and couldnt therefore be based on a solely rational view of the future.

However, when he considered the longer term, Shiller showed that the relation between stock prices and dividend indeed follows a trend. Whenever the ratio between the two deviates too far, this is usually followed by a reversion to the trend. In this way he showed that, despite short-term volatility, there is indeed some predictability in longer-term movements of stock prices. The same analysis works for bonds too, and this reasoning allows fund managers to make confident longer-term forecasts.

Robert Shiller has a great track record in spotting bubbles. In his book Irrational Exuberance, published in 2000, he argued that dotcom stock prices were unjustifiably high; indeed, within a year, prices collapsed. He is also known for the Case-Shiller index of American house prices, which was used to argue, even as early as 2005, that US houses were highly overpriced. And we all know what happened two years later.

Of course, all these theoretical results and claims mean nothing unless extensively verified by systematic empirical analysis. Unfortunately, financial markets, and the data that they generate, are notoriously hard to analyse. We know too little, and there are too many moving parts. Lars Peter Hansens share of the Prize lies in his econometric contributions, which have helped significantly with financial market analysis. In fact, a lot of econometrics relies heavily on tools that he developed.

In particular, Hansen developed a technique called the Generalised Method of Moments (GMM). This allows statisticians to estimate parameters in their models, based on the least possible assumptions about the data they analyse. This is very useful when researchers have little idea about what is going on. As a happy corollary, many of the existing econometric techniques were shown to be special cases of this more general method, depending on what is known about the data, and so his work has unified a lot of econometrics.

In practice, their works have led to many innovations that benefit the small-scale investor. For example, the growing popularity of index funds that passively track a basket of stocks is based on the realisation that there is no way in the short term to beat the market. Their work has also spawned whole fields of research, such as behavioural finance. It has also changed the benchmark models of modern financial analysis.

In any case, the three laureates would be the first to admit that there is a lot more to do before we really know how asset prices work. For the sceptics, the turmoil of the last five years should be proof enough. Forecast prices at your own risk.

Madhav Raghavan is a doctoral student of economics at the Indian Statistical Institute, New Delhi, and specialises in matching theory