3 Secrets to reading housing data for the real estate investor: Understanding Margin of Error

Note: this is part 2 in a series entitled “3 Secrets to Reading Housing data for the real estate investor”.  Part 1 is about Understanding Margin of Error. Part 3 is about Looking Beyond the Trend.

Part 2 – Understanding how to interpret Margin of Error

Different sampling choices and methods yield results with different margins of error. This is like saying that the actual answer could be higher or lower by a certain (also calculatable) amount.

For example, when the news media reports on the horse race of political campaigns, they might say that 40% of people surveyed said they would vote for candidate A, and 36% of people surveyed said they would vote for candidate B, and the results have a 3% margin of error.

What this actually means is that their highest level of confidence is that 37%-43% will vote for candidate A, and that 33%-39% will vote for candidate B. So, while at first reading it may seem that candidate A has a clear advantage, in fact there’s a range where the predictions overlap, and it may actually be the (unlikely, but possible) case that more people intend to vote for candidate B than candidate A.

How does all this apply to real estate investing in Silicon Valley? It matters because data on new home sales (or unemployment, jobless claims, home prices, etc.) is released each month, only to be revised up or down weeks later .   Not only are these  numbers important in themselves, the rate of change in these numbers often draw even more scrutiny.  Meaning, people are often more interested in the fact that statistic A (e.g. new home sales) grew by x% (e.g. 5%) last period, than in what statistics A actually is.  Needless to say, when the margin of error percentage is greater than the actual reported change in sales, the released figure becomes meaningless.

In addition, since the markets are forward looking, few people actually look back to see the revised numbers, relying purely on the first-reported estimates.

Bottom line:  1) compare the margin of error with the degree of change being reported to gauge how meaningful the data really is, and 2) don’t neglect revisions.

Related links on this blog recommended by Arkayne:

• Secrets to reading Housing Data for the Real Estate Investor: Looking beyond the trend | Silicon Valley Real Estate Investor
• Twitter Weekly Updates for 2010-01-31 | Silicon Valley Real Estate Investor
• Silicon Valley real estate news – week in review January 31, 2010 | Silicon Valley Real Estate Investor