Economic modelers spend their lives forecasting the past with ever-greater accuracy. No, this is not a typo or a misstatement. You have to describe the past statistically to calibrate any model, and you have to keep reducing the model's historic error band to have any confidence whatsoever that you have produced something capable of working in the future. Past performance may not predict future results, but it certainly captures the assets required to obtain future management fees.
Futures on the Past
This little diversion into the nature of past, present and futures markets was prompted by a review of the Chicago Mercantile Exchange's housing futures. All futures markets are based on the principle of indifference. If interest rates are at 5% and storage costs amount to 1% of the underlying asset's price, then a one-year future should be priced 6% over the current cash market price. You should be indifferent to the choice of buying the asset now and storing it yourself or buying it in the futures market for delivery a year from now. Futures markets also have a large measure of insurance built into them. Producers sell futures to lock in a price to be received, and consumers buy them to lock in a price to be paid. Risk management is understood, almost without saying, to involve events that will happen in the future. This is not the case with housing futures. Each of the contracts is based on the S&P/Case-Shiller (CSI) home price indices. They cover metropolitan areas of Boston, Miami, New York, San Diego, San Francisco, Washington, D.C., Chicago, Las Vegas, Denver and Los Angeles, as well as a composite national index. That in itself does not present a problem; we have close to 25 years of experience trading index-based, cash-settled futures on things such as stock indices.