Statistically Speaking, the Numbers Can Be Misleading
“How to Lie With Statistics.”
I remember reading this book several years ago and marveling at all the ways in which a few simple numbers can be manipulated to say anything we want them to. And nowhere are the principles of this book more alive than in the wonderful world of real estate.
Two months ago, it was reported that the inventory of homes for sale in the San Fernando Valley had reached an all-time high. Now, the number of homes for sale has dropped below what it was at the start of September, 1991, and September, 1990.
Statistics are supposed to help buyers and sellers understand the real estate market better. Much of the time, however, the numbers are just confusing or misleading. You have to know how the numbers were compiled and what may have caused the numbers to go up or down during a particular period.
Here’s a sampling of what could be construed as good news: There are more homes selling now at the asking price than last year. Despite the depressed market, the average time it takes for a home to sell in the Valley has increased by only 10 days. The ratio of final sales to new listings is actually higher today than it was two years ago.
And average residential sales prices, as reported by the San Fernando Valley Board of Realtors, have been climbing steadily over the last five years. In 1987, the average sale price in the Valley was $187,287. In 1988, the average was $222,426. The 1989 average was $255,425. The 1990 average was $257,192 and the 1991 average was $260,010.
Unfortunately, all these upbeat statistics don’t mean very much.
Consider other statistics that more accurately reflect the nose-dive in the local housing market:
The Board of Realtors reports that the average sale price was $247,900 in August, while the median sale price was $199,000. That compares with numbers of $277,100 and $215,000, respectively, for the same period last year. If that’s not enough to sink the hearts of home-sellers, consider this:
Total year-to-date sales volume reported by the board as of August, 1992, was $1.7 billion, down from $2.1 billion in 1991. Although the average selling time is currently only 10 days more than it was last year, this statistic only applies to homes that ended up selling. If your home sells, it takes an average of 138 days to find the buyer. The time period stretches to 215 days for residences between $100,000 and $110,000.
Another telling indicator is the increasing ratio of expired listings to new listings, which suggests that more sellers are simply pulling their properties off the market. This also helps explain why the percentage of homes selling at the asking price climbed slightly--people chose to stay put rather than slash the price on their existing residence.
In August, 1992, 1,672 property owners let their Valley listings expire. This compares with 1,607 expirations during the same period in 1991 and 1,746 expirations in August, 1990. Meanwhile, 2,615 residences were added to the Valley multiple listings in August, 1992. There were 2,843 added in August, 1991, and 3,418 added in August, 1990.
Even real estate brokers have mixed feelings about all the monthly sales statistics.
“The statistics are useful, but we generally know the trends before we read about them,” said Bobbi Miller, vice president and general manager at Fred Sands Realtors in Woodland Hills.
The numbers don’t tell all, Miller said. “There are some people who are deciding this is the wrong time to sell. Others are selling at a lower price and moving up to another home at a lower price.”
“Statistics can be used in many ways,” said Temmy Walker, president of the James R. Gary Co. Ltd. East real estate brokerage in Studio City. “It all depends on how you read them. The problem with averages and medians is that you can tip the scales with just a couple of million-dollar sales.”
The number of listings is down, Walker observed, but most of that decrease is a result of the usual decrease in listings during the fall and winter months, and “because sellers are taking their properties off the market.”
Using statistics to predict the future is a risky business, Walker said. For example, you’d think that an increase in interest rates would put a damper on home sales. Not so, Walker said. When interest rates start to climb, buyers get nervous and decide to take the plunge before rates go up even more. When rates are decreasing, she said, buyers seem to think they ought to wait for the market to hit bottom.
Another confusing element, Walker said, is the time lag in most statistics. August sales, for example, actually represent transactions started in June and July. To get a really current look at market activity, it may be better to look at pending sales rather than completed deals.
“In order to make a sound decision, a client needs to look at more than just compounded numbers,” advised Malcolm MacEwen, branch manager at the Jon Douglas Co. in Woodland Hills. “Unless you’re looking at your neighborhood, you may not be getting good information. There are dramatic differences just based on the location of the property.
“We put a lot of credence in numbers provided to us by the realtor associations,” MacEwen said, but it’s important to have a full range of statistics and an idea of how to interpret them. “A little bit of it is detective work,” he said.
So what do the numbers really say?
Consider: The Valley Board of Realtors reports 13,717 active listings of houses and condominiums for sale at the end of August. This compares with 14,330 at the same time last year. Meanwhile, only 852 single-family houses and condominiums were sold in August.
When you consider that August is one of the strongest months for housing sales, it figures that the remaining inventory won’t be sold until sometime in 1994. And how’s that for a statistic?
