How to Use Case-Shiller Indices to Predict Housing Trends

(Click on the image to enlarge)

Sonya, a fresh college graduate with a major in Economics, is interviewing for an Economic Analyst position with a major economic consulting firm.

Question # 1
Interviewer: Is there any difference between these two Case-Shiller* housing trends?

Sonya: They are very similar trends. In fact, even the monthly growth rates are almost in lockstep. By the way, am I looking at the seasonally adjusted data here?

Question # 2
Interviewer: Yes, you are. Are these month-over-month data comparable?

Sonya: Yes, since they are seasonally adjusted; otherwise, we would be comparing April, 2019 with April, 2018, etc. 

Question # 3

Interviewer: Why do you think the top-20 markets are moving in tandem with the top-10?

Sonya: Because the US housing market, overall, has returned to normalcy. Right after the last recession, known as the Great Recession, a number of major Wall Street companies started buying up the inventory, in large volume, creating a highly asymmetric market around the country. For the last 2-3 years that trend has significantly subsided, paving the way for a more normal market.

Question # 4
Interviewer: How did their involvement create an asymmetric market?

Sonya: Because they concentrated primarily on Sunbelt markets and as a result the growth in prices in those markets far exceeded the other markets.  

Question # 5
Interviewer: How would you characterize the health of the current market?

Sonya: It's still a healthy market considering 4.0% to 5.0% annual growth rates. Of course, these are muted rates compared to the prior run-up rates when the Wall Street investors were active. 

Question # 6
Interviewer: In terms of the index components, are there any duplications?

Sonya: Yes. The Composite-20 has all of the Composite-10 components, plus 10 more markets.

Question # 7
Interviewer: Can you name a few components that are mutually exclusive?

Sonya: Atlanta, Charlotte, Cleveland, Dallas, Detroit, Minneapolis, Phoenix, Portland, Seattle and Tampa.

Question # 8
Interviewer: Wow! That was sensational. Would you have graphed the data differently?

Sonya: Yes. Since these are seasonally-adjusted data, I would have bar-graphed the month-over-month (percent) changes.  

Question # 9
Interviewer: Would you use the NYC Case-Shiller data to show the Brooklyn trend? 

Sonya: Yes, to perform a quick-and-dirty trend analysis. If I were performing a trend analysis for a client, I would not use this MSA-level data which is quite broad in nature. I would start by collecting the parcel-level sales data from the Borough of Brooklyn itself.

* Case Shiller is a registered trademark of S&P CoreLogic.

Link to the Book