With SE's fantastic efforts to construct an index, we can now answer all the questions that have been vexing us for years. One discussion that seems to come up all the time is whether the Case-Shiller index for the New York metropolitan area (NYXR) and the NYC market (SECMI) have strong correlation. To answer this question, we can look at the year-over-year returns of of each and see what the... [more]
With SE's fantastic efforts to construct an index, we can now answer all the questions that have been vexing us for years.
One discussion that seems to come up all the time is whether the Case-Shiller index for the New York metropolitan area (NYXR) and the NYC market (SECMI) have strong correlation. To answer this question, we can look at the year-over-year returns of of each and see what the correlations are. The answer is 66%. If we compare against the Case-Shiller New York metropolitan area condo index, it's 76%. So, the Case-Shiller indices for New York are actually pretty good proxies for NYC.
Another interesting thing to look at are correlations between SECMI and the S&P 500. The correlation of year-over-year returns is more modest: 24%. You can also look at the correlation of the year-over-year return in the S&P 500 one year against the same for SECMI in the next year. In effect, does a change in S&P 500 in one year predict the change in SECMI in the next year? Curiously, this correlation is much higher: 55%. Correlations of the S&P 500 to the Case-Shiller indices are appreciably lower.
[less]
Response by w67thstreet
about 15 years ago
Posts: 9003
Member since: Dec 2008
I don't care what the data shows!!!!! Co-op force field, ON!
Ignored comment.
Unhide
Response by w67thstreet
about 15 years ago
Posts: 9003
Member since: Dec 2008
Good mkt, good bonus, more nyc re purchase. ;) didn't need the 'data' to know it gut wise.
Ignored comment.
Unhide
Response by Riversider
about 15 years ago
Posts: 13572
Member since: Apr 2009
Well we know we're dealing with real estate in both . And both are in the same relative geography. Are you not surprised the correlation is not higher?
I think it is important to note, though, that the correlations you mention are still generally relatively low ones. In addition, I think it is generally better to look at "R-Squareds." (Simply multiply the correlation - R - by itself.) That figure shows the percentage of New York real estate's price movements that are explained by movements in the stock market.
So in your last paragraph, 24% (R) becomes 6% (R-Squared) and 55% (R) becomes 30% (R-Squared). Fairly low numbers - but interesting nonetheless.
The stock market has said very little about concurrent movements in New York City real estate prices and a modest amount about next year real estate prices. Lots of other factors are presumably more important.
Hope this hasn't sounded too pointy-headed...
Ignored comment.
Unhide
Response by Sunday
about 15 years ago
Posts: 1607
Member since: Sep 2009
What about vs. nyc unemployment rate?
Ignored comment.
Unhide
Response by inonada
about 15 years ago
Posts: 7952
Member since: Oct 2008
Topper, not too pointy-headed. I know the deal between r-squared and correlation, and I hear what you're saying.
Ignored comment.
Unhide
Response by inonada
about 15 years ago
Posts: 7952
Member since: Oct 2008
Sunday, corrrelation of year-over-year SECMI vs. year-over-year NYC unemployment rate is -72%.
Ignored comment.
Unhide
Response by Topper
about 15 years ago
Posts: 1335
Member since: May 2008
Pretty interesting, inonada. Pretty high!
Ignored comment.
Unhide
Response by jason10006
about 15 years ago
Posts: 5257
Member since: Jan 2009
R2 is not as meaningful because the analysis is NOT about causation
'/%3e%3c/g%3e%3c/svg%3e)
I don't care what the data shows!!!!! Co-op force field, ON!
Good mkt, good bonus, more nyc re purchase. ;) didn't need the 'data' to know it gut wise.
Well we know we're dealing with real estate in both . And both are in the same relative geography. Are you not surprised the correlation is not higher?
http://streeteasy.com/nyc/talk/discussion/23298
Interesting comments, inonada.
I think it is important to note, though, that the correlations you mention are still generally relatively low ones. In addition, I think it is generally better to look at "R-Squareds." (Simply multiply the correlation - R - by itself.) That figure shows the percentage of New York real estate's price movements that are explained by movements in the stock market.
So in your last paragraph, 24% (R) becomes 6% (R-Squared) and 55% (R) becomes 30% (R-Squared). Fairly low numbers - but interesting nonetheless.
The stock market has said very little about concurrent movements in New York City real estate prices and a modest amount about next year real estate prices. Lots of other factors are presumably more important.
Hope this hasn't sounded too pointy-headed...
What about vs. nyc unemployment rate?
Topper, not too pointy-headed. I know the deal between r-squared and correlation, and I hear what you're saying.
Sunday, corrrelation of year-over-year SECMI vs. year-over-year NYC unemployment rate is -72%.
Pretty interesting, inonada. Pretty high!
R2 is not as meaningful because the analysis is NOT about causation
For the housing indices I mean