The NAR's existing home sales report for the second quarter of 2007 is out today, and it shows a continuing slide in sales volume: existing home sales registered an annualized 5.91 million unit rate, down 10.8 percent from year-ago levels. Only six states showed an increase in sales volume. Median prices followed sales volume, dropping 1.5 percent from year-ago levels to come in at $223,800 -- but the NAR focused the majority of its analysis on the fact that 97 out of 149 MSAs registered year-over-year price growth. 83 MSAs had registered increases during the first quarter, something the NAR was more than happy to point to as signs that the worst might be behind us:
Lawrence Yun, NAR senior economist, said the price trends are encouraging. “Although home prices are relatively flat, more metro areas are showing price gains with general improvement since bottoming-out in the fourth quarter of 2006,� he said. “Recent mortgage disruptions will hold back sales temporarily, but the fundamental momentum clearly suggests stabilizing price trends in many local markets ..." “Since all real estate is local, this report on metro area home prices is more meaningful than our monthly data on national prices because metro areas are less subject to price distortion that can result from geographic changes in the composition of sales,� Yun said.
There is nothing like always being sunny-side-up, I suppose; and, of course, at some point the NAR will actually manage to be right. It's not as if the housing downturn is going to last for eternity. It's just that when they do eventually manage to be right, you'll want to remember that being correct doesn't always equate to market awareness. It might not even equate to a basic college education, if what I'm seeing from the NAR lately is any example. I usually try to avoid rants on this blog, but as someone who has studied for a PhD and understands econometric analysis at a very high level -- yes, really -- I simply can't take it anymore when I see a well-cited trade organization absolutely fail to do any sort of basic analysis to support its own rhetoric. HW readers should by now understand the awful truth -- that housing is still in some serious trouble. Prices aren't leading indicators of where a market is headed; so it's sort of silly, really, to see a senior economist at the NAR expounding on how some perceived "movement" in prices might signal better days ahead. What's even more unsettling here is the method of analysis used to create the claims being made. Since when does any self-respecting economist -- someone supposedly reasonably familiar with at least the most basic of statistical analysis techniques -- eyeball an aggregate median comparison and say that we're seeing more increases this quarter than last? From the data those of us in the public are provided, we would have no way of actually knowing which way prices really moved, because we don't have any way of knowing if the medians being compared differ in a statistically meaningful way. Is a 0.8% increase in Fort Wayne, IN really an increase? Is a -0.7% drop in Kansas City really a drop? Who knows? I realize that the NAR uses median data, but it could have just as easily calculated means as well. Either way, some sort of actual test should have been used to answer the question: is the difference in median values (or mean values) actually even meaningful? For medians, the most common test is something known as the Mann-Whitney. I won't bore you with why the NAR should be using what are known as parametric tests, but the point is that even a simple median comparison can be tested. If Lawrence Yun apparently can collect a paycheck by pretending to be an economist, why am I sitting here writing a blog for free? Why are you reading this blog? After all, I can pull my own opinions out of thin air and make up conclusions without actually doing any analysis. I'm pretty sure you could do the same, too.