---------------------- Forwarded by Vince J Kaminski/HOU/ECT on 10/13/2000 
04:24 PM ---------------------------


Zimin Lu
10/13/2000 11:59 AM
To: Brad Horn/HOU/ECT@ECT
cc: Vince J Kaminski/HOU/ECT@ECT, Stinson Gibner/HOU/ECT@ECT, Robert 
Shiring/HOU/ECT@ECT, Jay Knoblauh/HOU/ECT@ECT 
Subject: Re: Option P&L  


Brad,

I was extreamly busy yesterday.  Sorry for answing your question till now.

Although I am not exactly sure how  the system handle gamma, this is what I 
think the system is doing:


Curve shift = today's price - yesterady's price

P/L due to curve shift = today's market value using today's price curve (with 
everything esle the same as yesterday's) - yesterday's market value using 
yesterday's price curve.

So P/L due to curve shift contains both delta and gamma and higher order 
terms.
We then use theoretical gamma (meaning option model gamma: 0.5*gamma * (price 
change)2 ) for gamma contribution and?define delta = curve shift - theoretical gamma.??Therefore, the gamma may not be very accurate to explain the delta change, ?especially when you have big price change due to higher order contribution. ??Let me know your thoughts on this.???Best wishes,??Zimin?????????   Brad Horn                10/12/2000 07:11 AM??To: Zimin Lu/HOU/ECT@ECT, Stinson Gibner/HOU/ECT@ECT?cc: Vince J Kaminski/HOU/ECT@ECT, Vladimir Gorny/HOU/ECT@ECT, Robert ?Shiring/HOU/ECT@ECT, Jay Knoblauh/HOU/ECT@ECT ?Subject: Option P&L??Gentleman:?        The ERMS system, as you know, has an excellent capability for ?decomposing option P&L into the following components:??new deals?curve shift?gamma?vega?theta?rho?drift?2nd order adjustments??What i dont understand is the gamma component which is reported in dollars.  ?The unit of measure suggests that incremental changes in a contract position ?is being associated with specific prices.  These prices are the effective buy ?or sell prices associated with the dynamic delta position.  ??Stated differently, the standard taylor expansion has incorporated a price ?variable in such a way as to convert the unit of measure from gamma's ?standard contract count to total gamma dolalrs.  This is something I dont ?understand.  To date, inquiries to the risk management accounting group has ?further revealed that the gamma component of P&L is not well understood.  ??This is what concerns me: Bridgeline has 2 books with option exposures (NYMEX ?and Gas Daily).  Both books dynamically hedged its positions during ?yesterdays large price move and, through anticipitory hedging in advance or ?during the large price move, secured sufficient coverage to neutralize ?expected changes in delta.  However, our P&L from our underlying position did ?not offset our gamma P&L.  Consequently, I have to ask WHY?  Im hoping that a ?brief look at the why gamma dollars are calculated may reveal something which ?will better guide our hedging decisions.??Any help is appreciated????