Enron Mail

From:vince.kaminski@enron.com
To:erin.rice@enron.com
Subject:Re: Pre-submitted question on eSpeak
Cc:vince.kaminski@enron.com
Bcc:vince.kaminski@enron.com
Date:Wed, 12 Jul 2000 02:06:00 -0700 (PDT)

Erin,

A book by Martha Amram and Nalin Kulatilaka is an excellent
non-technical introduction to real options and it contains all the most
important
references to more advanced books and papers.
The book has an endorsement by Jeff Skilling on the cover page.

The Research Group offered several one-day seminars on real options
and applications to the energy industry. We still have a few binders with
the presentation materials available if you are interested.
We plan to repeat the seminar sometimes in the fall.
The first part of the seminar (about 4 hours in the morning) covers general
concept of real options
and their applicability to the energy business.
The second, more technical afternoon session, covers stochastic processes used
to model price uncertainty in the energy markets and specific case
studies (valuation of natural gas storage facilities and of peaking gas-fired
power plants).

The real options approach has been developed specifically to address the
problem of making
investment decisions under uncertainty. Nobody in this field claims that this
is a
perfect tool, but it represents a significant progress compared to other
techniques developed
earlier.

Discounted cash flow analysis that tries to incorporate uncertainty through
analysis of several, in most cases, arbitrary scenarios (most likely,
optimistic,
pessimistic). These scenarios don't identify explicitly the risk drivers
and don't specify the future proactive management decisions.

The real, option approach is very powerful because it allows to (1) capture
uncertainty in an explicit way and (2) to design investment projects that
allow to exploit future positive developments and reduce future exposures to
downside risk.
This approach allows also create a link between investment decisions and
future
operational decisions. Forward-looking investment decisions create options
that are exercised
in the future through active management of a project.

The real options technology relies heavily on advanced statistical tools to
come
up with the representation of future possible states of the world. The real
challenge is
to use these tools in a sensible way. I have seen in my career (almost 30
years
of applying mathematical tools to business and economic problems) many quants
armed with
powerful computers who reminded me of monkeys armed with hammers. The
challenge is not
to run mechanically thousands of simulations based on arbitrary assumptions
but to translate in a creative way the insights of people who understand
specific businesses
into parsimonious quantitative models. It is especially critical to
stress-test the assumptions
of any model and to ask the question if the outcome of a model depends
critically on any set of assumptions.
If this is the case one should use common sense to examine the underlying
assumptions.
I remember that in the early eighties quite a few models simulated the
dynamics of oil prices,
but all the stochastic scenarios represented fluctuations around a very
optimistic upward trend.
One would have been better off stepping back and asking a simple question
what Economics 101
teaches about cartels and the dynamics of supply and demand.







Enron North America Corp.

From: Erin Rice @ ENRON 06/27/2000 03:52 PM


To: Vince J Kaminski/HOU/ECT@ECT
cc:
Subject: Pre-submitted question on eSpeak

Good news! There is already a question waiting for your July 12 eSpeak
session. I have pasted it below, although you don't have to respond until
the actual event. Don't forget to send a bulleted list of discussion topics
if you would still like us to advertise the event on the elevator screens.

Thanks.

- er

submitted by breineck
I was doing some reading about the application of real options in the
evaluation of non-financial assets. Would you recommend any texts or articles
for a
more in-depth study of this area? Is quantification of risk or uncertainty
the major challenge in using this concept? Can statistical tools be used with
this to do
a sort of sensitivity analysis?