Comments on Caveman Forecaster: Data Analysis, Graphing, and Forecasting: Case-Shiller: June '09 Data, Housing Price Forecas...

Dave, Great points and I totally agree. In my own...

2009-08-25T13:45:54.598-05:00

Dave,

Great points and I totally agree. In my own head I always keep in mind that these types of models are totally "ignorant" to the market fundamentals and any artificial distortions at play but I don't always mention that predominantly in my posts.

I am suspicious any time a sharp inflection is forecasted like this. Sometimes you will see a one or two point anomoly where the forecast shifts but then lines back up where it was and in those cases I would discount it as bad or temporary artificial "distortions" in the free market as you mention. But here, the shift seems to be slow and consistent... so far. And I do have some confidence in that I have seen this type of ARIMA model successfuly forecast an inflection like this many months out as this specific ARIMA model did forecast the top of the housing market back in 2005/2006.

But the real test will be later in the year as we get back to the negative seasonal time period. If the forecast for a recovery holds up then I will be much more confident in this forecast.

I still worry about the effects of recent governme...

2009-08-25T12:05:02.071-05:00

I still worry about the effects of recent government distortions on this series.

I mentioned the Freddie and Fannie foreclosure moratorium previously, and Calculated Risk had a few bullet points on the unsustainable impact of new home buyers on existing home sales.

Separately, he looks at the fundamentals which suggest that prices are still too high. Personally, when I look at those graphs and see unusual angles at the end of the series, I grow a bit suspicious, just because the rest of the series are relatively smooth. With the exception of the peak of the bubble, most other sharp angles tend to be temporary anomolies in the middle of much smoother trends. One possible explanation for the recent sharp uptick is the temporary distortions in the marketplace.

And Mish covered the fall in cure rates, which will lead to an upcoming wave of foreclosures.

Anyway, it'll be interesting to see how the model fares in the coming months.