The Expanding Universe
Mortgage and Reverse Mortgage rates clearly fluctuate all the time. Several articles today, however, focus on predicting rates and financial performance as a “science.” An article on mathematical models of mortgage rates was in the science section of the NYTimes, along with another much longer article on Quants, comparing stock market performance predictions to quantum physics and string-theory.
Putting predicting rates in the same hard science category as Einstein’s unifying theory seems slightly ridiculous. Although both are theoretical, the recent stock market crash has proven that its impossible to predict the behavior of the markets with a high degree of certainty. The bursting of the housing bubble came as a surprise, much like the bursting of the .com bubble but with much much more damaging consequences.
Yes, many people foresaw that the markets were overextended, but they did not know when, how, or how dramatically they would fail. Thus the models (especially the long-term models) were generally unable to prevent people from losing a large amount of money last year. While that does not make them worthless, it does prevent models from being seen as an accurate predictive measure.
Greater than 50% accuracy may be enough for a model to perform well on Wall Street, but a much higher degree of certainty is necessary for a theory to gain any support in the scientific world. Models prediciting mortgage rates and stock market performance may be scientific in nature and are certainly highly mathematical. However, they should not carry the same weight and are not in the same category as the science behind developing a new cancer drug or determining Einstein’s constant for the expansion of the universe.
Eventually the mortgage rates and the stock market will dramatically rebound. If the models actually met a scientific standard we’d know when that’s going to happen; sadly, we don’t.