R. David Ranson calls his economic forecasting method a ``pure discovery, a complete accident.'' Whatever it is, Mr. Ranson and his partners at H. C. Wainwright & Co. Economics, a Boston consulting firm, have used this method to forecast the economy with extraordinary accuracy since he came upon it in 1979.
They correctly forecast the last recession. They predicted the subsequent recovery, including the rapid growth in 1984 and the slow growth in '85.
At this moment, they figure the economy will grow on average at a 5 percent real rate this year and next. In other words, they see ``a return to supernormal growth.''
That growth rate is far above the ``consensus'' or average forecast of most economists of about 3 percent.
Mr. Wainwright's reputation as a forecaster has prompted slightly more than 100 institutional investors, managing perhaps $100 billion altogether, to subscribe to the service. Most pay through ``brokerage'' -- they instruct brokers carrying out stock transactions for their institutional portfolios to pass on a portion of their fees to Wainwright.
What Ranson ``discovered'' in 1979 is that by examining interest rate futures, especially for three-month Treasury bills, he could use an econometric model to predict the economy 18 months or so ahead. Its accuracy proved out very soon. But he had no economic theory as to why the model succeeded. He worked that out with his colleagues later.
A key part of the theory is what Ranson terms ``incentive economics.'' It holds that economic behavior is governed by the anticipated returns to effort and sacrifice. If ``work'' offers a favorable financial return, an individual will give up some leisure to work. The balance between work and leisure will be affected by changes or anticipated changes in taxes, inflation, etc.
An increase in interest rate futures may imply bad news in the future: for instance, higher inflation. So activity will step up now and drop off in the future.
Because financial markets are ``efficient'' -- that is, they reflect known information fully and without bias -- these markets in Treasury-bill futures reflect the collective wisdom of those active in them.
``It is safer to be an empiricist than a Keynesian, a monetarist, or anything else,'' said Mr. Ranson, speaking of prominent schools of economic thought. ``The economic evidence of these theories is always ambiguous. But there is sufficient evidence to confirm any theory if the adherent is sufficiently flexible.''
Monetarists, for example, hold that after a lag, the economy will follow shifts in the growth rate of money. Ranson concedes money is important to the economy. ``The trouble is that money is inscrutable.''
He maintains that money is hard to measure, because many financial instruments, in addition to checking accounts and currency (the usual definition of money), can be used as money when demand is sufficient. People can use Eurodollar deposits or trade credits, or else issue their own debts, such as junk bonds, and all these forms of credit can have an effect on economic activity.
Ranson's own model uses interest rates rather than money. Individuals cannot control interest rates as they can generate credit, he says. So he regards interest rates as a more suitable base for forecasting.
Wainwright does not attempt to make quarterly forecasts, as many economists do. But its annual forecast is updated each month and published each quarter. If interest rates change, the forecast will change. No personal judgment is added to the forecast of the model.
Ranson says institutional customers realize that people's expectations about the economy change and that this influences the economy. In mid-1984, for instance, the model forecast 4 percent growth in 1985. By the end of the year it was forecasting 2.5 percent, close to the actual 2.2 percent for that year.
As for this year, the Wainwright model predicts strong sales in housing, motor vehicles, and other consumer durables, nondurables, and services. But it holds that industrial and commercial construction will decline.
As for stocks, Ranson sees only a 50 percent probability they'll outperform Treasury bills the next 12 months. Last June his model had that as 85 percent likely.