State Space representation

[sulong]

Hi Tom ,

I have been trying to replicate this paper ( see below ) for quite some time, but I am have some difficulties.
Like them, I want to use the Kalman Filter -smooth to get a the resiliency series; then do some sort of regression.

I suspect the state and observation equation in the paper is not quite correct, can you have a look and give me
some pointers on how to use the DLM instruction to get the time series data for stock resiliency .
Their model is the Local Trend Model, am I correct?

appreciate your help Tom,

Thanks


( Resiliency, the Neglected Dimension of Market Liquidity: Empirical Evidence from the New York Stock Exchange
http://www.google.com/url?sa=t&rct=j&q= ... 2593,d.bmk)

[TomDoan]

That's actually OK. It might look a bit odd to have one of the state equations being 1=1+0, but that is a common practice if you don't have Z and MU options (as RATS does) to handle constant shifts. There is no "measurement error" (that is SV is left out of the state-space model) because those have been shifted into the state error, and the observable is included in the state vector.

If you haven't gotten it to work, you'll have to post the program so we can look at it.

[sulong]

ok

Thanks Tom,
I will give it a try.