Re: VaR problem with STORAGE

Winston,

Thanks a lot.

I talked to Tanya about testing VaR based on 20-day historical fwd-fwd vols.
It would be great if we had this done by the close of business today.

We should also push ahead to get the ability to use the fwd-fwd vols input
from
different sources (historical, trader's input, etc.)

Vince

Wenyao Jia
06/05/2000 08:36 AM
To: Grant Masson/HOU/ECT@ECT, Tanya Tamarchenko/HOU/ECT@ECT, Vince J
Kaminski/HOU/ECT@ECT
cc: Naveen Andrews/Corp/Enron@ENRON, Stephen Stock/HOU/ECT@ECT
Subject: VaR problem with STORAGE

Hi everyone,

By using simulated curves and positions, I manually calculated both days's
VAR for AGG-STORAGE in Excel spreadsheet. I got almost same results as the
ones produced by the current production VaR model.

I did found two things:

1. The positions in Storage-Prc changed although the total position did not
change much:
For July delivery: delta changed from 2,682,338 to 5,695,589
For Sep. delivery: delta changed from 4,629,638 to 1,682,515.

2. The forward forward NG volatility has a big change on Dec-01:
May 31 : 0.0122956
May 30 : 0.278
But the Original Vol curve don't change much.

I also changed code so that it will read NG forward forward vol curve from a
file. We can run some testing today.

Look like Jin found some problem in the factor loading program. If we can
produce new factor loadings we can also rerun VaR using the new factor
loadings.

If there is anything else we can do, please let me know.

Thanks!

Winston Jia