Exhibit 99.401

                            CALIFORNIA POWER EXCHANGE
                        MARKET EFFICIENCY RECOMMENDATIONS

From a California Power Exchange view of the market, the shift of trading out of
the PX markets into the ISO real time market has been the single most dominant
fact that appears to have reduced supply significantly in our day ahead market.
This is problematic in that it burdens ISO with procuring significant amounts of
energy along with capacity in the most volatile and time sensitive period, and
offers an inter-market arbitrage opportunity that was outside the design
intentions of the California market founders.

In May and June of 1999 the real time market accounted for 2 to 3 % of the ISO
grid. In the summer of 2000 it is averaging 8 or 9% and at times has exceeded 20
%. There has been a corresponding decrease in day ahead supply. The PX day
ahead, in these times of supply tightness, now serves as the "floor" with day of
and real time usually priced higher than day ahead. The sellers are going to say
this has nothing to do we them. They have been complaining about the IOUs
shifting demand from day-ahead to real time as a cost savings strategy for more
then a year. We know that load shifting does make sense. And, we know that the
only way for the supply to arbitrage the price spread is for the supply to
follow the demand to real time.

To help correct this situation, to uncomplicated the California marketplace
somewhat, and to reduce the strain on ISO resources to meet real time market
requirement that was never intended, the California Power Exchange recommends:

1. The ISO real time price that is paid for supply will be capped at the actual
ISO real time price for any supply deviations up to 5 % outside day ahead
forecasts and shall be no higher than the PX day ahead price for all other real
time supply procured by the ISO. Conversely load deviations greater than 5 %
shall pay the greater of the ISO real time price or the PX day ahead. The first
5 % will be at the PX day ahead. The benefit for the California markets is that
this would eliminate economic incentives without undue penalties to
under-schedule load or withhold supply from day ahead and avoid price/time
pressure of real time marketplace. It would also allow utilization of real time
for intended purpose of close in adjustments due to load/weather changes or loss
of generation capability. How does a generator know when he is bidding
supplemental energy whether he is bidding to the first 5% market or the greater
then 5% market. Suppose a generator offers to increase output for X dollars and
the ISO accepts to meet real time demand. Then after the fact you tell the
generator "oops, we took your energy to meet demand that was outside the 5%
band, so we're not going to pay you your bid price, we're going to give you the
PX DA price which is half of your bid price"? This is contrary to the FERC
policy that the ISO can set a cap on what it is willing to pay, not it can not
set the price that the suppliers must bid.

2. The PX Board will review the market implications of amending its tariff to
provide for publishing the daily supply offered into the PX day ahead market at
various prices or release aggregate daily curves instead of the current three
month lag policy. Publishing the supply bids offered in the UMCP market
accomplishes what? These bids are not binding except to the extent they are
accepted in the UMCP auction. In fact they have nothing at all to do with the
market price except in the hours when there is not congestion. In any hour when
there is congestion, the price is determined by the schedule adjustment bids
submitted after the PX runs its market, not the bids in the PX market. The PX
does not have aggregate curves for Schedule Adjustment bids (SAB). If you were
to aggregate them, you would have to do so in a way that you aggregate those on
the same side of the congested

transmission line separately from those on theothe side, but until you send them
up to the ISO you don't know if there is congestion, so how could you do this?
Also, the bids in the PX are piece wise linear curves, while the SABs are step
function bids. So you would need different software then presently used to build
the aggregate.

Have you given any thought to requiring that the same bids be given for the UMCP
market and SABs? This would not be without its own set of problems, but I have
always suspected that there is some gaming going on with the SABs. The SABs that
set the price usually come from the divested generators.

In addition to lessening real time proportions, the following should also be
considered:

3. The PX and the ISO will work to explore the introduction of electronic
tagging from source to zone for all in-state production. This will provide an
audit trail to track the exact routes of generation from within California. What
does track the routes of generation mean? How does within California come into
it?

4. Provide daily to local newspapers the hourly PX prices for the day ahead in
areas where the rate freeze has ended. This valuable market information could
trigger potential demand response opportunities. However, our attempts to
accomplish this goal have met with stiff resistance from the media.

5. IOU hedge limits should be increased further.

6. Explore jointly with the ISO the benefit of the ISO utilizing PX markets (day
ahead and day of and even CTS daily block products and potentially a new PX
capacity option market) to reduce out of market purchases and minimize
procurement resource requirements.

The PX has reviewed these concepts with the ISO and believes they represent some
logical corrections to some of the near term market inefficiencies. The detail
implementation will need to be worked out between the two staffs and market
stakeholders.