wilmott.com

Which method should I use in order to calculate/estimate EONIA? I need the result to simulate VaR. I have to admitt that i am not having too much experience with interest rate models. Which methods are employed by the banks ?Sorry, if the question sonds to easy or stupid for the most quants Thank in advance

Are you using EONIA as your spot interest rate? Or are you simulating other interest rates and want to estimate EONIA given those rates?If the former, are you simulated just EONIA or other rates?If you just want EONIA in a vacuum, it's not a hard problem. You've got a good history with which to calibrate a model. GARCH is one choice of model, or jump diffusion with mean reversion.If you're putting this in the context of more interest rates, you need to start by picking an interest rate model. Which one to use depends on what you're trying to do.

We have a energy contract, where the variable part is linked to EONIA (no other interest rates as benchmark). I woluld like to asses the the future spot rates (3-4 years) in order to calculate the price of the contract at any given time in the future (i have no access to Bloomberg or Reuters ). I have the historical interest rates of EONIA, so i think a simple simulation will be good enough for my purpose. A friend of mine suggest me to employ a CIR model as discribed in Jackson/Stauton book ( i am working only with excel and VBA )Why GARCH model- do you mean, i colud use the future volatility predicted by the GARCH directly as VaR in this case ?Thanks in advance, Aaron

Hi there,I'm still a student, it's my first message out there but let's try to answer.GARCH model can help you in that you estimate the volatility of the interest rate so its future values using the appropriate dynamic. You can use any interest rate model as you have the volatility (as you do with BlackScholes model). Some models assume that you can approach future spot rates considering forward rates (like LFM or LSM).That say, I don't really understand what you wanna do. Compute the VaR or just get your product price ?

QuoteOriginally posted by: VegaPlusHi there,I'm still a student, it's my first message out there but let's try to answer.GARCH model can help you in that you estimate the volatility of the interest rate so its future values using the appropriate dynamic. You can use any interest rate model as you have the volatility (as you do with BlackScholes model). Some models assume that you can approach future spot rates considering forward rates (like LFM or LSM).That say, I don't really understand what you wanna do. Compute the VaR or just get your product price ?Hi VegaPlus,first at all i want to run a simple VaR. I employed already a historical simulation, but I think a GARCH-model will be a better method for my calculation. Are the GARCH-calculation steps the same as by assesing volatilty of equity/indices or interest rates should be treated in other manner? What i am going to do, is to build the daily interest rate log returns and then to estimate a, b, w using MLE /Nelder-Mead algorithm.The second task i have to manage is to build a simple short interest rate model simulation. I dicided to employ the CIR-model. Frankly speaking, i do not have an idea how to calculate a and b in excel? As sigma i will take simply the GARCH-value?

I uploaded a GARCH file on EONIA. But i still having a problem with my model since i am not able to ensure the condition alpha+beta<1.Could someone help to adjust the model?

QuoteOriginally posted by: niki5I uploaded a GARCH file on EONIA. But i still having a problem with my model since i am not able to ensure the condition alpha+beta<1.Could someone help to adjust the model?I suppose my request is pretty ease for the quants and they are bored beyond belief, but I also know- you are good people , so give me a hint, what I am doing wrong in my calcualtion