Today the Commerce Department's Bureau of the Census released its report on sales of new one-family houses during December and January. December annualized sales were 665,000, up 0.5 percent from November, and January sales were 693,000, 4.2 percent higher than December. Or, were they?

The Bureau qualified both the December and January sales totals by stating that they were only accurate within +/-12 percent. In an expanatory note that was part of the government report, the Bureau stated:

"These statistics are estimated from sample surveys. They are subject to sampling variability and errors of response and nonreporting. Whenever a statement such as 'an increase of 3 (+/-4) percent' appears in the text, this indicates the range (from -1 percent to +7 percent) in which the actual percent change is likely to have occured. All ranges given for percent changes are 90-percent confidence intervals. If the range contains zero, it is unclear whether there was an increase or decrease; that is, the change is not statistically significant."

Applying this to the numbers reported today, we conclude that the December sales change could have fallen anywhere between -11.5 percent and +12.5 percent. Similarly, the January sales change could have fallen anywhere between -7.8 percent and +16.2 percent. What's more, since both of these ranges contain zero, we don't know for sure whether the monthly changes were positive or negative, and neither of the changes was statistically significant.

So, if you saw any reports today that claimed that January's new home sales figures were yet another sign of a resurging economy, don't you believe it. According to the Bureau of the Census, there is a 24 percent range of uncertainty as to what today's numbers mean.

During 1995, 664,000 new houses were sold at an average sale price of $158,000.

The average sales price of new houses sold in December was $163,200. The average price in January was $147,500.

At the end of January, the seasonally adjusted estimate of new houses for sale was 381,000. This represented a supply of 6.7 months at the current sales rate. This unsold homes inventory figure has been on the rise in recent years. The yearly averages for 1993, 1994, and 1995 were 5.11 months, 5.78 months, and 6.50 months. So, it appears that the new home market is suffering from inventory build up. Just like other sectors of the economy.

Friday, the Columbia University Center for International Business Cycle Research (CIBCR) announced that its leading inflation index rose to 104.0 in February from a revised reading of 103.2 in January. This was the first uptick in this indicator in 14 months. Over the longer term, the year-over-year change in the index was -6.7 percent, the tenth consecutive negative reading. The lead time of the CIBCR index is approximately six months.

One data point doesn't define a trend. So, we need to see more data on this indicator before we can draw any conclusions about the February uptick. If the February data point marks a turning point in the direction of the index, this would be an alert that, sometime around August, the general price level may begin to drift upward.

Byline: Lafferty (MF Merlin)