Economic Forecasting (updated 27 Feb 99)

Almost every financial services firm has an extensive economic forecasting effort. It is usually part of a so-called 'top-down' investment process, which starts with an outlook for the economy and monetary conditions, continues to the strongest industries, follows with detailed company study for stock selection and may include an overlay of technical analysis to provide a timing dimension. Some would add analysis of social and political conditions even before economic studies (see POLITICS AND INVESTING).

Economic forecasts derive from models - usually of the aggregate national or global economy, but sometimes of parts of those economies: particular industrial sectors, regions of the world or even single products or firms. Basic approaches to forecasting simply extrapolate the past; more sophisticated models attempt to understand the sources of past changes and build them into their forecasts. The latter requires knowledge of economic history and economic principles, though, even then, forecasting is by no means an exact science. But while the accuracy of economists' predictions is frequently a target of jokes, forecasting remains a popular pursuit.

Forecasts for the macroeconomy are published regularly by academic institutions, thinktanks, governments, central banks and international organizations like the OECD and the IMF. In these places, modeling can, to a certain extent, be conducted free of the constraint of producing quick and usable data on a daily basis. But in the investment world, forecasts are required to be done 'early and often'. A relatively short-term outlook is normally the limit of their aspirations - what will happen to interest rates within the next month? - with decision-makers demanding rapid output that they hope will be directly relevant to their immediate problems.

Much of the output of financial market models is naturally closely guarded in the hope that it may bring advantage to its owners and their clients. But, at the same time, investment economists like to maintain a public profile for marketing purposes, and are often called on by the media to give their opinion on the latest macroeconomic developments. Their interpretations of economic data may give some clues as to how the financial markets will react, though more often than not, they are explaining why the markets have already reacted as they did. Invariably, too, there are disagreements about what various indicators mean, depending on different beliefs about the economy, and whether the firm is taking an optimistic or pessimistic view of the markets.

Each month, the Economist polls a group of financial forecasters and calculates the average of their predictions for real GDP growth, consumer price inflation and current account balances in a variety of countries. More specialized services like Consensus Economics survey over three hundred economists each month and offer details on average private sector predictions.

Economic forecasting guru: Peter Bernstein

Despite the pressures for 'early and often' forecasts, a number of Wall Street and City economists do as good a job as any forecasters, among them Abby Joseph Cohen, Steve Roach and Ed Hyman. Most such investment economists are good students of market conditions - careful keepers of useful data, and on occasion creative in extracting some kind of signal out of the noise. Ed Yardeni, for example, the chief economist at Deutsche Bank Securities, turns his website into a cyber chart room. If you want to access data and view charts, Yardeni's site is an essential stop. He also makes his commentary available in a section for clients that is password protected, but a substantial amount of the content is openly accessible.

One economic commentator stands amid the few that many of us would class as the best: Peter Bernstein. He grew up heading his father's investment firm, Bernstein MacCauley in New York. He was the first editor of the Journal of Portfolio Management, founded by Gil Kaplan, and has received many awards, among them the highest honor granted by the investment management industry's professional body, the Association of Investment Management & Research.

Bernstein is able to walk on both streets - with practitioners and academics. He writes a newsletter, Economics and Portfolio Strategy, to test and disseminate his analyses. And writing is one of his main strengths: his two books on the history of risk and on how 'capital ideas' came to Wall Street have been regulars on the business bestseller lists during the 1990s.

Like a good academic, Bernstein marshals all the arguments, especially those that are counter to his own position. His mid-February 1998 letter, for example, examined the case for exuberant stock prices in the United States, giving particular emphasis to the markets' reliance upon an all-knowing Federal Reserve for economic management. Bernstein concluded that 'stocks are a risky investment and should be managed accordingly'. Since that analysis was approximately the same as his November 1997 conclusion, he was ahead of the wave and for the right reasons. Bernstein is also faster than most to admit where he has been wrong and to try to examine what led him astray - or, as he jokes, 'what led the market astray when it failed to act the way I thought it would'.

Counterpoint

Financial analysts are professional forecasters. But why study the economy, a traditional lagging indicator, if you want to forecast investment measures? The investment record of this process is only rarely better than random - and when you take account of the expenses of achieving these results, they come out a little bit less than chance. Why do it at all with that unconvincing record of success?

Economic forecasts are supposed to be meaningful. But if you believe that asset prices reflect a forecast of future outcomes, it would seem quite difficult to use a technique that reaches back into the past to get an idea of the future. But that is what economic forecasting does. It is teased for forecasting three recessions for every one that actually happens. No wonder it is called the dismal science.

Financial Times economics columnist Sir Samuel Brittan makes a pointed reflection on the practice of forecasting: 'The golden rule for economic forecasters is: forecast what has already happened and stay at the cautious end. Forecasts tell us more about the present and the recent past than about the future.'

Poor methods, bad models and inaccurate data are all blamed for the recurrence of serious forecast errors. But according to Oxford economics professor David Hendry, these are not the primary cause of systematic mistakes. Rather, unanticipated large changes within the forecast period are the culprit. The primary fault in economic forecasting is not rapidly adjusting the forecasts once they go wrong.

Hendry uses an analogy from rocket science: a rocket to the moon is forecast to reach there at a precise time and location, and usually does so. But if it is hit by a meteor and knocked off course - or destroyed - the forecast is systematically badly wrong. That outcome need not suggest poor engineering or bad forecasting models - and certainly does not suggest that Newtonian gravitation theory is incorrect.

Guru response

Peter Bernstein comments: 'For better or worse, economic forecasting is an essential ingredient in investing because earnings and interest rates are both conditional on economic conditions. So you have to do it or use it in some fashion. Furthermore, although a forecast of next quarter's GDP or even next year's earnings per share may be wrong, the kind of forecasting I do - and really that Abby Cohen does too - is to try to define the basic environment - inflationary or not, fast growth or not, competitive or not, and so on. That kind of thing is most helpful and has paid the biggest dividends over the years, not just in this cycle.'

At the front of this book is further commentary from Peter Bernstein: a grand sweeping history of the markets reprinted from the 1 December 1998 issue of his newsletter.

Where next?

Forecasting is a key task in financial institutions because of the profound effects economic developments can have on potential profits. And while leading economic indicators might provide a hint as to what the economic future holds, they do not anticipate what the additional effects of powerful economic agents like government policy and the financial markets themselves might be. To try to get ahead of the competition, companies will aim to model more accurately, and with more consideration of possible discontinuities in the markets.

One way to make forecasts more useful - though not necessarily better - might be to follow the principle of 'truth-in-labeling' used on food packages and elsewhere. We could describe the kind of forecast we are making more accurately. For example, if we are using backtesting, we should say that that is exactly what we are doing and which of two varieties.

One form of backtesting is 'momentum': the forecast is derived from a view that the past momentum will continue in roughly the same direction - often straight line - as it has in the past. The other form is 'regression to the mean': we think things will not go back or up or down, but return to average conditions. This is like a series of coin flips that goes ninety-nine times in one direction, and we think the next event is related to the preceding one.

Alternatively, we can say that our forecast comes from our own insight or novelty, and label it that way so we know that it is essentially out of our head and our own creativity - or lack of creativity, which we will know in time. Sometimes different techniques like high-frequency forecasting come from this. Or it can come from news and our response to new news. This is not necessarily insider information but news that is not necessarily generally recognized by others - a form of forecasting derived from information.

Finally, the most common form of forecasting is waffle: we do benchmark investing or stick to the middle because we do not know what else to do. That is perfectly all right, but we should label it as such. Let us say that is what we are doing, so people can understand what they are getting when they listen to us. Most of the time, a waffle is the right thing to do, but at all times, we can make our forecasts better by correctly labeling them.

Read on

In print

Peter Bernstein, Against the Gods: The Remarkable Story of Risk Peter Bernstein, Capital Ideas: The Improbable Origins of Modern Wall Street Peter Bernstein's newsletter Economics and Portfolio Strategy Reports from Consensus Economics

Online

econwpa.wustl.edu/EconFAQ/ - Bill Goffe's 'resources for economists on the internet', one of the best entry points on the internet for economic information
www.economics.ox.ac.uk/hendry/Frontpage.htm - David Hendry's research project on the econometrics of macroeconomic forecasting
www.economist.com - website of The Economist
www.ft.com - website of the Financial Times
www.yardeni.com - Ed Yardeni's website