Enron email

Algorithmic Indexing

25 Oct 2001, 18:22 UTC

RE: Models

From
Harry Arora harry.arora@enron.com
To
Iris Mack
Date
Thu, 25 Oct 2001 11:22:25 -0700 (PDT)
Iris

what is the difference between this spark spread function call and the one being used in the spreadsheets we use ?


Harry
Quoted or forwarded from earlier messages
-----Original Message-----
From: 	Mack, Iris  
Sent:	Wednesday, October 24, 2001 10:34 AM
To:	Danilov, Viacheslav; Issler, Paulo
Cc:	Arora, Harry
Subject:	FW: Models

 << File: CrackSpreadOptions.xls >> 

Hi Slava,

	Paulo says you have some questions about what we're doing with spark spreads.

	He did not give me any specifics as to what you needed.  Give me at call at X3-6711 for more details.

	In the mean time, I have been developing exotic options modules/tutorials for the power trading floor.  One of the modules is on spark spreads  - please see attached file.

	
Regards,
Iris





 -----Original Message-----
From: 	Issler, Paulo  
Sent:	Wednesday, October 24, 2001 10:26 AM
To:	Mack, Iris
Cc:	Danilov, Viacheslav
Subject:	FW: Models


Iris:
I think you are the expert on how we are trading spark spreads (if we are doing any). Slava wants to know more about it. I appreciate if you contact him directly. 
Many Thanks.
Paulo Issler

 -----Original Message-----
From: 	Danilov, Viacheslav  
Sent:	Tuesday, October 23, 2001 2:05 AM
To:	Issler, Paulo
Subject:	RE: Models

Paulo,

thank you very much. 
we will look at all your work.

just to remind you on spark spread options.

many thanks,

Slava


 -----Original Message-----
From: 	Issler, Paulo  
Sent:	23 October 2001 01:00
To:	Danilov, Viacheslav
Subject:	Models

Slava:
Here are the files related to the models I mentioned in our phone coversation.

FwdFwd Vols:

Here is a spreadshet implementing a technique for calculating fwd fwd volatilities for seasonal vol curves. The approach avoids the issue of geting negative fwd fwd variances when using the traditional approach. It makes necessary, however, to develop a specific fwd fwd vol curve for each contract month. 

Here is how it works:
	1) Create a variance curve: Var = vol*vol*(Ti-t)
	2) Fit a desoasonalized curve with functional form: y = a*(tb)?	3) calculate fwd-fwd variances for the desoasonal curve:?	4) Calculate fwd-fwd vols for each month by multiplying the ?	    fwd-fwd variance curve by a factor equal to:?			SQRT(Var(Contract Month)/Var(fit))??? << File: FwdFwd.xls >> ?I implemented a C code for calculating that. This, with the omicron vols may be used to model the gas curve evolution in a HJM framework.?? << File: fwdfwd.c >>  << File: FwdFwdw.c >> ??Swing Options:?Here is the code and documentation for the swing option:?? << File: XLCALL32.LIB >>  << File: bico.c >>  << File: factln.c >>  << File: FRAMEWRK.C >>  << File: FRAMEWRK.H >>  << File: FSSwing.c >>  << File: FSSwingOpt.c >>  << File: FXSwing.c >>  << File: FXSwingOpt.c >>  << File: gammln.c >>  << File: Interpolate.c >>  << File: nrutil.c >>  << File: nrutil.h >>  << File: SwingOpt.def >>  << File: SwingOptW.c >>  << File: SwingW.c >>  << File: XLCALL.H >>  << File: Bdt.c >>  << File: Swing Model.doc >> ?Here are the spreadshets examples:?? << File: FSSwingOpt.xls >>  << File: FXSwingOpt.xls >> ?LookBack:?Here is my implementation for the lookback option with fixed strike. It uses an analytical solution in which prices are sampled continiously over time. It still provides a good approximation for a discrete daily sampling:?? << File: XLCALL32.LIB >>  << File: Cumnorm.c >>  << File: framewrk.c >>  << File: framewrk.h >>  << File: FxLKBK.c >>  << File: FxLKBK.def >>  << File: Integration.c >>  << File: lkbkTEST.xls >>  << File: nrutil.c >>  << File: nrutil.h >>  << File: put.c >>  << File: xFxLKBK.c >>  << File: XLCALL.H >>  << File: call.c >> ?Let me know if you got them all.?Give me a call with your questions/concerns.?Thanks.?Paulo Issler?