Economic forecasters aren’t quick to adjust their expectations, according to a new study by Goldman Sachs. In fact, if a forecaster overestimates an economic factor, they will continue to overestimate for several months before adjusting. Then, they will spend several months underestimating the factor.
Why do forecasters generally take so long to adjust their expectations? And how does first tuesday avoid over- or underestimating California’s housing market?
first tuesday insight
The short answer? Future expectations must adjust for changes in current trends.
Forecasts consider a variety of factors, each of which are constantly changing. A good forecast will consider the whole picture – not just a single trend.
Basing a forecast upon a single trend is called a projection. Projections are based on data from the recent past. So, if the trend in home prices this year is up 20% from the prior year, the projection would anticipate 2013 ending 20% higher than last year.
In contrast, a forecast takes 360 degrees of the market into consideration. For example, even though California mid-tier home prices are up by 20% as of May 2013, home sales volume has been slowing since November 2012. Historically, home sales volume movement precedes home price movement by about 12 months. Thus, home prices are likely to begin slipping by the end of 2013.
A multitude of other factors are also taken into consideration, including:
- job creation;
- interest rate movements;
- bond market rate movements;
- demographics with multiple facets;
- construction; and
- loan products.
Many forecasters assume the housing market is a fully rational equation, devoid of unpredictable influences. But forecasting is like keeping leashes on hundreds of hounds, all jerking in different directions, and trying to figure out where they’re taking you. Ignoring the animal nature of the hounds (the fluid changes in culture and economics) makes for a forecast that will diverge wildly from reality, sooner or later.
We know the market can never be fully rationalized for reason of our animal instincts. That’s why we take as many factors into consideration as possible when making each forecast. Consider first tuesday your expert tracker of California’s real estate market. We are, you understand, in our 35th year of playing this game.
Re: Forecasters are bad at forecasting, study finds from the Washington Post