Message-ID: <18161242.1075843953732.JavaMail.evans@thyme> Date: Mon, 21 May 2001 18:13:00 -0700 (PDT) From: david.hoog@enron.com To: jeffrey.shankman@enron.com, mike.mcconnell@enron.com Subject: hedging for outage options Cc: per.sekse@enron.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Bcc: per.sekse@enron.com X-From: David Hoog X-To: Jeffrey A Shankman, Mike McConnell X-cc: Per Sekse X-bcc: X-Folder: \Mark_McConnell_June2001\Notes Folders\Notes inbox X-Origin: MCCONNELL-M X-FileName: mmcconn.nsf as per has mentioned to you, there have been a number o factors working against us recently in obtaining insurance for the book. i was expecting to have a positive response by tomorrow, but the person responsible had a baby on saturday, so this may be delayed until later in the week. (i had my computer with me in the delivery room for our first child, so its possible, (but i didnt for the second)). one unexpected factor is that aquila has signed exclusive arrangements with some of the insurers who could have backed us. these agreements prevent them from doing business with us. another primary market, ace, is still pissed at me for leaving and going to enron, whom they consider a serious threat to their franchise. this was somewhat expected, and i am already starting the groveling process to get them to transact with us. another unexpected situation is the supply/demand imbalance. given the fact that forced outage insurance policies have paid out no more than a few million in claims while collecting over $100 million in profits, the common wisdom in the insurance industry was that supply would increase and demand would decrease. this year has surprised even me -- demand is as strong as ever and supply has not increased. this is evidenced by the much higher margins we have collected on the few selected deals we have done. we also got started too late in the season, so we did not have a sufficient portfolio to attract insurance partners earlier. i tried to build the business too quickly from this position. of course, my job is not to explain the reasons for things turning out differently from what we expected. the current problem is that i need to hedge out our book in a difficult environment. first, we agreed that we would take action on may 15 if we did not obtain reinsurance by that date. i have halted origination activity in this area, which will result in the loss of ~75% of our business. we are currently in final negotiations on 4 transactions, which will then make up our final portfolio for this year. these are among the most profitable deals we have seen this year and should be the easiest to hedge. this will leave us with $15 million in revenue and $250 million in total theoretical exposure (P99 = $30 million). second, we agreed that we would not be sitting on a "high gamma" risk going into june. i am still hoping that we get the insurance offer next week, but if it is not in place by june 1, we will hedge the book in the underlying commodity to gain time for placing the insurance. this should buy us 2-3 more weeks to finalize the strategy without sitting on significant risk. this may also be the best time to talk about our newest hedging strategy. when backed into a corner, and deprived of sleep (and food), i usually come up with my best business ideas. since ive had little of either for the past 2 months, i have been working with alex and larry to develop a generalized hedging theory for outage risk. without going into much detail here, the idea is that there is a first derivative with respect to MW (i'm currently calling it omega (w = watt)), which is comparable to the delta and gamma hedging greeks. on a single unit, this has no value, but for a portfolio, you can compute this omega and apply a combination of static and dynamic hedges to equalize your option position, much in the same way the black-scholes formula hedges an option with respect to changes in the underlying commodity. for example, when you experience a single outage on the system, you increase your position so that you are prepared for the next outage. our preliminary thoughts on this are that it can manage the tail risk of a portfolio as well, if not better, than insurance. the other benefit of this approach is that it is much more consistent with the theory used by all other traded books to manage risk. it may also create an entire new way of pricing unit-contingent power based on the hedging cost (just like B-S is used to value options). i'm not suggesting that this solves all of our problems in the short term, but this development will be significant in power trading. i am working with tim belden this week to apply this theory to the massive pacificorp deal, because it is the only way to effectively manage that risk. the total hedging cost using this approach should be less than $30 million, compared to the offered premium of $140m. we can talk about this on thursday, then i'd like to discuss with vince & vasant to get their thoughts.