We are concluding our 5 part series, Reading Housing Data 101, to help you better understand quoted housing statistics and avoid taking others’ conclusions at face value. We have thus far dealt with the distinction between local vs. national data, seasonality, quarterly vs. annual comparisons, now let’s look at the last two: statistical significance and trends & conclusions.
Statistical Significance
Housing data is released each and every month, only to be revised up or down a few weeks later. Pay attention to that margin of error of +/- some percentage. When this margin exceeds the actual change in whatever is reported (housing starts, existing home sales, etc.) the data reported becomes meaningless. Example: housing starts were up 3% last month (+/- 5%). This means that housing starts could have been down 2% or up 8%, with the 10% range of actual possibility. In addition, everyone talks about the preliminary numbers and yet they are just that: preliminary. Pay attention to the revised numbers that come out a few weeks later; they rarely get the bold headlines of their earlier brethren.
Trends & Conclusions
This is a topic we’re very passionate about. Even perusing through the quarterly sales reports for Manhattan, the ease with which we all conclude “prices decreased from Q3 to Q4 in x neighborhood” makes it all too convenient to believe it, even though it’s flawed thinking. We’re not talking about prices for apples or gold; rather a basket of varied properties, with varied characteristics. The more accurate statement is therefore “the inventory that sold in Q3 in x neighborhood was % lower in price than the inventory that sold in Q4.” After all, the same apartment was not sold in both quarters to be able to cleanly say its price decreased.
The key lies in recognizing that there is no direct linearity to depend on. Yes, patterns emerge from this data and they help inform our analysis of the times in which we live. Just understand the underlying dynamics behind this data to guide your decision-making.
