Get Blog Updates on your iGoogle homepage:
| Demonstrated vs. Estimated Demand |
| Written by Jonathan Smoke | |
| 12.19.2007 | |
|
Discuss this article on the forums. (0 posts) In my last post I described the distinction between estimated demand and demonstrated demand and how strictly looking at demonstrated demand can lead researchers and planners to miss clear opportunities in a market. There’s nothing quite like a real example to show the difference. So I am sharing with you some research we recently completed in a submarket of Atlanta. The above chart plots the estimated and demonstrated annual demand by price point in a single submarket in Atlanta. The estimated demand numbers are calculated by a proprietary method we have built that relies on consumer segmentation estimates of household growth and factors in ownership rates, forecasted changes in those rates, and adjusts for existing household turnover and replacement homes required for aging housing stock. Our demand model estimates the number of new homes expected to be in demand by price point over an annual time frame. Demonstrated demand in this example has been pulled from actual closings. It is composed of the last four completed quarters of new home closings in this submarket, so it too represents an annual view of demonstrated demand. It is demonstrated because it is what the market has demonstrated, i.e., real transactions. When you compare them you can make some interesting observations. For example, most price points in this submarket show that demonstrated demand has not lived up to expectations over the last four quarters. This is likely a reflection of the suppressed demand caused by the credit crunch and concerns about future home price declines. Even though there is limited chance of real home price declines in Atlanta, the media-amplified fear factor has had an impact. There are two price points where demonstrated demand exceeded estimated demand ($450,000 to $550,000). The data just shows the discrepancies—it doesn’t tell you why. But I’d be willing to bet that the demonstrated demand was higher in these price points because the price points below and above were undersupplied instead of believing buyers in these price points threw caution in the wind and had no problem getting jumbo loans. It’s also pretty clear that there is enormous unmet opportunity for product in the $200,000 to $350,000 range. And it can’t be impossible to deliver those price points because new home product was sold under $200,000. I will be the first to admit that our estimated demand techniques aren’t perfect. Our demand model is built upon demographic projections of household growth. At its core, the demographic models used rely on Census figures, which are then expanded by survey inputs to derive household information both by detailed consumer segments and by more discrete geographies. Historical research has shown that such Census-based estimates are reliable but tend to undercount growth. At higher geographic levels such as MSAs and cou  nties, there is ample data to make the estimates very reliable. As these estimates are applied to discrete consumer segments and smaller geographic areas, there will be more distortion, especially if rapid changes are happening in the segments or geographies that the population and survey models do not take into account.For example, if an area has grown rapidly in the past, these demographic models will tend to show continued growth even if it is impossible for infrastructure or housing stock to keep pace with such growth. Likewise, in areas where there has been limited historical growth, the models have no foundation to project a sudden surge in growth caused by new infrastructure or developments. That is why this sort of information, while very powerful, needs to be interpreted by experts in the local market who can identify areas in which the projections are likely to be understated or overstated. From our experience, we have found that the errors in the projections tend to level out across a county or a complete MSA, as when one area exhausts its capacity for growth, the growth will move into new areas where the market supplies the needed capacity. I hope you can see from this simple real example that estimating demand using consumer models can provide some very valuable insights in local housing market dynamics and opportunities. We aren’t experts in local markets, but we do have the ability to produce this kind of powerful intelligence for almost any area of the country. 
We're always looking for feedback from our readers.
|
))
| < Prev | Next > |
|---|
