Source: The Wall Street Journal
There are market-based solutions that would improve the forecasts of GDP, that do not involve model building or a dependency on a particular school of economics. Event-based financial instruments where the amount of the payout is directly and computationally dependent on observable economic indicators, observable prices or observable outcomes, such as election outcomes in the Iowa Electronic Markets, inflation forecasts of US Treasury TIPS and other prediction-based actively traded instruments are more accurate than model-based forecasts.
A US Treasury GDP bond of different maturities, where payouts are computationally dependent on actual reported GDP of different time periods, would allow for the computation of future expected (forward) GDP growth rates. These forward growth rates, if they behave like other prediction market-based forecasts, will have greater accuracy than any model based forecast, including a Federal Reserve model.
One has to wonder why the Federal Reserve has not promoted the idea of a market-based, GDP dependent, financial instrument to improve the accuracy of its GDP forecast and to allow for computation of expected GDP growth rates for many different future time periods. Is it a fear of immediate accountability and feedback? The ability of investors to read real-time, market-based forecasts of the future of the US economy from a GDP-based bond would provide a level of oversight, feedback and accountability to the Federal Reserve's actions and judgments to improve future GDP growth that the Federal Reserve is not used to having.