How to build the short end of a zero coupon curve for non-core Eurozone countries?

Question

I am in the process of building zero coupon curves for some countries in the Eurozone.

I have the following data sets:

1. Euribor and EONIA
2. Swap rates
3. Bond price and yields

The bond prices (and thus yields) reflect the relative credit worthiness of the sovereign issuer (a point that has come to the fore most recently). For the longer dated time buckets therefore, the yields "make sense". At the shorter end however, the official rates seems to be Euribor and EONIA (for s/n-o/n) - I can't see how this makes sense - since these are the same data points I would be using to evaluate "high quality" debt from Germany (for example). I can't see how it makes sense to use these same points for constructing yield curves for any of the peripheral Euro-zone countries - as the credit risk does not seem to be reflected at the short end - am I missing something?

What data points do professionals out there use to construct yield curves for Eurozone countries (with the exception of Germany)?

Answer 1

The risk implied by Euribor or EONIA (or their swaps) is for lending to another prime rated bank. These rate indexes represent where contributor banks are offering funds to each other in the interbank market. Contributing banks are mostly rated P-1 (Moody’s) or A-1 (S&P). You wouldn’t use these rates for govt discount curves because the risk doesn’t match.

If there are no short term govt securities you would have only a constant forward rate up to the first data point. Even when there are short term securities, you don’t necessarily want them in every application. A parametric curve fitting might do a poor job with them. Spreads between short term govt and other short term rates might be ill-behaved.