Lag Selection

[opticchickenmcnugget]

Dear E Views Forum Members,

I am currently trying to test whether treasury yields to see if they are a martingale/random walk series, as a precursor leading up to my error correction model.

Say i've taken quarterly averages of monthly data (wanted to test at shorter intervals), and i am testing for ARCH.

How many lags would you suggest I use?

Thanks in advance for any help.

PS: Feel free to drop me a message if you have some experience with ECM's and treasury yields.

[startz]

Don’t take quarterly averages. Use last day of quarter if possible. Taking averages screws up autocorrelations.

[opticchickenmcnugget]

Ok, will do - how many lags would you suggest for the ARCH test with quarterly data?

[startz]

No idea. Sorry.