The business economist at work: the Economics Department at Fannie Mae - Federal National Mortgage Association

Business Economics, July, 1995 by David W. Berson

The basic functions of the Department can be narrowed to three main areas: forecasting and current conditions analysis, policy analysis, and public dissemination of our views on the economy and housing and mortgage finance markets (or, more simply, public communications). In each of these areas, the Economics Department supports other units at Fannie Mae in their attempt to meet the strategic goals of the company. As is the case at most other firms, the Economics Department is a support function and not a profit center - and as such, is useful to the company only to the extent that it helps those parts of the company that actually earn income. This can be viewed quite broadly, however - and so, for example, supporting the Office of the Chairman (which is comprised of the Chairman, Vice Chairman, and President) is held to be an important role of the Economics Department.

Rather than focusing on my position as Chief Economist at Fannie Mae, it probably is more interesting - and more instructive for other groups of corporate economists - to look instead at the entire Economics Department. The work that we do has increased sharply in the nearly eight years that I have been at Fannie Mae, as it probably has for most of the readers of this article, yet the resources devoted to the Department (especially headcount) have not changed

appreciably over that time - indicating that productivity has risen significantly. This has been the case in large part because of improvements in personal computers and software - we simply would not have been able to do everything that is expected of us today eight years ago with the relatively limited physical tools at our disposal then. While this piece is not meant to be a marketing device for John Qualls' PC Corner column, it is clear to us that improving technology is extremely important the functioning of a successful economics unit today - especially one in which there is increasing demand.

In the remainder of this article, I will cover the specific duties undertaken by the Economics Department, as well as how they are used by other groups in the company - and thus how we provide value-added services to Fannie Mae.

FORECASTING AND CURRENT CONDITIONS

ANALYSIS

The Economics Department undertakes a new forecast of the economy and housing and mortgage finance markets each month. This is keyed off the monthly GDP report from the Commerce Department and typically is completed by the end of the first week of each month. The timing of this forecast is crucial, as the Department is responsible for a monthly report to our Board of Directors due on the second Tuesday of the month - and the revised forecast results are an important part of the report. We use a commercially available macromodel of the U.S. economy as the key tool in developing the monthly macro forecast - although the resulting forecast is our own. One of the benefits of the particular macromodel that we have chosen to use (in addition to its good simulation properties) is the capability to build our own "back end" models that use the macro results as exogenous inputs. None of the commercially available macro-models has the detail in the housing and mortgage sectors that we need, however, so we have developed a large multiequation model of the housing and mortgage markets. Between these two models, we are able to forecast a number of variables of importance to Fannie Mae - including short- and long-term interest rates, home sales, home prices, mortgage originations (divided into purchases and refinancings, adjustable-rates and fixed-rates, and conventional and government-insured), liquidation rates, and serious delinquency rates.

These forecasts are used by senior management in shaping their views on the outlook for the economy, housing and mortgage finance markets, and Fannie Mae business volumes. We play a more direct role in the latter area, as the Economics Department is part of a business volume forecasting team, in which we, our Marketing Division, and our Portfolio Management Division determine jointly what our likely volumes will be (for a three-year period) - based in large part on our forecasts of the economy, housing, and mortgage markets. (In the interests of full disclosure, I should say that while the company uses our interestrate forecasts in projecting business volumes and in its public pronouncements by the members of the Office of the Chairman, it does not take positions based on our forecasts. While we think our forecasts typically are fairly accurate - and that we tend to beat the consensus forecast - the company does not take "bets" on interest-rate movements. Instead, it attempts to hedge any interest-rate risk, earning income on the resulting smaller spread. With sufficient volumes, even small spreads - resulting from significant efforts to hedge interest-rate risk - result in large earnings.)

In addition to forecasting national trends in the economy and housing and mortgage finance markets, the Economics Department is responsible for analyzing regional economic conditions. We average about five analyses per month of specific state and metropolitan area conditions for use by senior management when they visit particular areas (and because we do business in every state and MSA in the country, we eventually get around to analyzing all of them). Perhaps more importantly, however, our regional economic analysis efforts are geared toward forecasting home prices on a regional basis. Fannie Mae and Freddie Mac have jointly developed (under the auspices of the Department of Housing and Urban Development) a repeat-transaction home price index - which uses the approximately 4.5 million homes that either one firm or the other (or both) have purchased or securitized over the past fifteen years. We feel that this repeat-transaction index is a better measure of the underlying value of homes, as compared with other measures of home prices that can be affected by, for example, changing geographic location of sales - and there is a developing literature supporting this contention. (Note that there are commercially available repeat-transaction home price indexes as well.) Changes in the value of a home can have a large impact on the probability of a mortgage going delinquent or defaulting. For example, the probability of a serious delinquency on a home with a current loan-to-value (LTV) ratio of 110 percent is significantly greater than one with a current LTV of only 80 percent. While lenders typically do not originate loans with LTVs over 100 percent, the sharp movements in home prices during the past decade has meant that some LTVs have risen abruptly as the value of homes in some parts of the country have declined. As a result, an accurate forecast of regional home prices is an important determinant of future delinquencies and defaults - and thus of potential future losses. Each quarter, the Economics Department forecasts home prices over a ten-year period using a series of vector autoregression (VAR) models. This is one area in which it is clear that advancing computer technology has enabled us to perform a service to the company that would have been impractical a decade ago.