Fit Vasicek model to market data

[xiloto]

Hey , guys. I'm an italian student and i have to do a work for a project in my study course.
The project is on the vasicek model
The part that i don't understand is the following.

1- Using datastream or Nyse-euronext (ex.liffe) find the cross-section of the euribor rates ( or the corrispondent ZCB prices) for different maturities and get the term structure for the closest available date.
2-Estimate the parameters of the vasicek model with OLS for the best fit at the term structure observed.
3-Repeat the previous point for the quotations of the last 12 months with weekly intervals ( get the wednesday quotations)

The first part is the part where i have more difficulty, beacuse i don't understand what data i should take
Thanks in advance. sorry for the english

 

[lucasnutz]

Hi Xiloto,

I understand that the EURIBOR curve is built just like the USD-Libor curve. The USD-Libor curve (which according to bloomberg is build just like the EURIBOR curve) is built as follows:

Short end:
You use the USD-Libor cash rate quotes published by the BBA.
Middle end:
You strip the 6m eurodollar contracts (FRAs) to infer the middle end
(eg: you have the 1 year cash rate, and you have a 12x18 FRA contract, ergo you can strip the 1.5y zero coupon rate)
Lond end:
Using the USD swap contracts (fixed-Libor swaps) you strip the rest of the libor curve.

Taking the latter as an example, you could use the Vasicek Interest Rate model to value zero coupon bonds, FRAs and swap contracts. Using the cash rates to set zero coupon bonds, and setting the valuation of the FRAs and Swap contracts to use a zero coupon yield curve to discount each cash flow, you can fit the Vasicek model. How? Move your Vasicek parameters (accounting for non-negative parameters) using OLS as instructed. This way you would build a EURIBOR curve starting from an O/N rate up to 50 years.

Hope this explanation serves you well.

Best regards,

Lucas