Forum Archive Index - November 2000

["Jeremy" <jeremy@electrosilk.net>]

> i wasn't meaning to imply that the climate's getting wetter, but that the
> weather 'patterns' of decades ago are of little use in forecasts for a
> financial enterprise.  more numerous and more extreme weather events (hot,
> cold, wet, dry, whatever) have characterised the 1990s more than
historical
> records would predict.  prudent managers of weather-dependent enterprises
> should be at niwa's gates.

Unfortunately, this is a one of the few things I am qualified to talk about,
seeing as I spent a lot of years analysing weather extremes and setting
design parameters for offshore oil rigs (none of which have fallen over yet,
fingers crossed)

There is a basic problem with most peoples understanding of climate and how
it relates to business or engineering activities.  Things like 100 year
storms or 1000 year storms are simply statistical estimates of the likely
hood of such an event, using observed patterns and frequency of storms in
historical times.  You can have a 1000 years storm tomorrow and also have
one next week and nothing will be unusual.

The stats allow an organisation dependant on weather to make a logical
assesment of risk, and an expected loss or return.  e.g.  Building a
structure to survive particular levels of storm versus cost versus loss
versus economic life.

A typical calculation using made up figures and stats is :

It costs $100m to build a structure with a design life of 20 years  The
structure is designed to survive a 100 years storm, .  If it fails it will
cost $500m in colateral damage.  While it is running it generates $50m per
year nett revenue.   Probability of a destructive storm during the design
life is roughly 4%.

Increasing the frequency of the storms so that a 50 year+ storm will now
make it fail increases the probability of a damaging storm to roughly 20%
during the design life.

Factoring these in (really roughly) gives an expected return of $800M with
the 100 years storm and $495M with a 50 year storm.

Do you as a prudent manager accept the lower expected return?  Or do you
spend another $200M beefing up the structure? Or do you say the expected
return on the project is acceptable at both levels of risk?  Or do you take
out insurance to cover the potential difference and pass the risk on to
someone else ?

Personally I'd say that if the project is viable at one level of risk and
not at another, then it is too marginal.  A good financial design should
work over all levels of risk - which is perfectly possible without
over-engineering anything.

If your company has been caught out by bad weather, then it refelects bad
financial planning and bad risk assesment by the management, rather than
anything so stupid as 'the wrong kind of weather' or 'bad forecasts'

Jeremy





----------------------------------------------------------------------------
http://www.sharechat.co.nz/          New Zealand's home for market investors
http://www.netbroker.co.nz/        Trade on Credit, Low Brokerage. Join now.
----------------------------------------------------------------------------
To remove yourself from this list, please use the form at
http://www.sharechat.co.nz/forum.shtml.