QUESTION AND ANSWER

Ladies and gentlemen, to ask a question on the phone lines today, press *, 1 on your touch-tone phone. Again, *, 1 for any questions or comments.

Taking a drink out of a fire hose with a lot of this stuff that we’ve given you. We’ve given you a lot of information today. We’ve give you a lot of new information that you haven’t had before. But hopefully, we’ve been successful today in giving you enough information to understand the positions that we have, talked about the risks around each one of those — each one of our regions, how our curves compare to others, the sensitivities to spark spread movement in each one of those regions.

Andrew ran through the complexities of the accounting that we have and any outstanding legal issues. There’s a lot of things that we’ve went — we’ve gone through.

The key concepts today, just to review, it is our stated intention to exit this business. We’ve been fairly successful in getting good value up until now for the $600 million that we have liquidated. But the market conditions that we see today make it difficult to exit quickly from this business right now, but the efforts will continue.

But in the interim, we will continue to manage the liquidity and the risk around the business. You can see the bullet points here. And I won’t go through them. We talked about the accounting, earnings and valuation and the variations there.

We now will open it up for questions. Let me just say we’ve got plenty of time for any questions that we have. And we want to make sure that we exhaust all the questions that people may have. We have questions from people here in the room. We’ve got people on the phones and we have people on the web. So we’ll sort of move those around.

Those in the room for questions, please wait ‘til you get a mike so that people on the phones and the webcast can hear your questions. And then, we’ll move from there. So question?

Exhibit 99.2

Yes, two questions. First, you’ve provided the cash flow forecast on an [Inaudible] basis. Could you discuss the cash flows on a discounted basis?

And second, for your estimated hedging tolling revenues, what are the pricing assumptions behind that? Are you using current forward numbers or you know, model numbers?

Bill or Andrew? Probably Andrew on those?

Sure. As far as the discounted numbers, there would be two things we would do off the undiscounted, as we’ve always talked about in the past. We get to what we call a risk adjusted number. We showed that number in the third quarter. Our total portfolio risk adjusted was a little over $1 billion.

So what we do there is we take those cash flows. And the first thing we do is we discount them at LIBOR, because that is the risk free rate. So you can go out and in your models, run a LIBOR curve.

The second thing we do is for any periods that are not hedged, we apply what we call a market risk premium. That’s the price that the market would require to enter into a contract and take on that risk. And we use the capital asset pricing model, which is spread out in many finance textbooks to do that with. And we discount back, based upon an equity requirement or an equity based return.

Historically in the equity markets, it’s been about 35 percent. So what you would see off those gross numbers is about $1.2 billion in what we call risk adjustments, over those time periods. The further out, the greater the risk adjustment because it’s unhedged. And that brings you back to your $1 billion of what — of really what we would call current value. If someone were to come in and that would, I guess, be a starting point for negotiation.

So hopefully, that answers your first question. Can you repeat your second question, because I forgot it as I was talking?

The estimated hedge tolling revenues, what the pricing assumptions behind that?

First of all, those revenues will happen, even if spark spreads — right now the pricing assumptions are market where there’s a

market. And then we extrapolate to our current — to our forward models, where there’s not a market.

However, it really doesn’t matter, because those values are based upon the fact that even if all of those dollars went away, since we have hedges in place, the value of the hedge would go up. So all we’re saying there is that to the extent my tolling revenues are in the money still, versus either market or my model, I’m going to get those revenues. But if the market goes to zero, I’ll just get those revenues up in my hedges. It’s just really more of a geographic location within the cash flow slide.

Did you have some more questions? I think — it looked like she had another question.

Just to follow up the — could you break down in terms of the sensitivity of the pricing that you would — or excuse me, the amount that you would really get in terms of the market versus what you are assuming is the model years out?

OK, I think, once again, I think simplistically, let’s just assume that those revenues are associated with the sparks spread of $1.00. The spark spread goes to zero, all of that revenue line would go away.

However, the hedge against that would increase in value by the equivalent dollars. So we really don’t feel there’s any price sensitivity in those revenues because they are hedged. The reason we don’t — the reason we — it’s kind of weird the way we show it that way. That value is really associated with the sale itself, that’s called the hedge. It’s just that the hedge is marked against market. So you have the value that the hedge is in the money against the market. And you have the value that the toll is in money against the market.

Well, if the market goes like that, you just see that the value of the hedge increases. So it’s really a geography on the income statement the way we showed it. The sensitivities would be equivalent to the sensitivities we showed only on the unhedged megawatts, though.

So there’s certainly going to be some credit sensitivity, which we do not model for you. I think that’s fairly simply to model. There’s also going to be some option sensitivity, based on the closer it comes to expiration. But the vast majority, the price risk has all been hedged out.

Other questions? Up here in the front? Right?

I have two questions. I believe you mentioned that for your standard OTC contract, some of them can be hourly and some can go out a year or two. Is there any way you could approximate what percentage might be hourly, versus relatively longer term? And the other one is on your — let’s see, this is slide 34, the AES 4000 tolling agreement, there’s a mention that 68 percent of that is hedged until 2010. Could you break out for me how you get there? Maybe I’m just not adding some numbers correctly, but as I add up the resale of the toll, that’s about 1175 megawatts. And the forward power sales of A, B, and C combine to about 425 megawatts, plus the 60 megawatts at Milagro.

And yet, you’ve got 3956 of the AES. 4,000. How are you getting to that 68 percent? Thank you.

Sure. First of all, on the hourly versus longer term, you’re going to have a lot more shorter terms. It’s going to skew heavily towards your shorter term. You’re trading hourly more. You’re trading a day ahead more. You’re trading a month ahead more than you’re trading two years out.

I don’t have an exact breakdown, but you’re probably easily 80 percent in — let’s say under a year. And then maybe 20 percent between one and five years.

Regarding why you’re adding megawatts up, you know, when we talk hedge, what we’re really focused on is the cash flows. The product D does not have any megawatts associated with it. Really, it’s really a capacity sale.

So you have to actually add up the cash flows of each of those associated agreements, compared to the demand payment. And that’s where we get our coverage ratio.

You know, if you’ll notice on some of those other regions, we’re 100 percent hedged, but we’re cash flow negative. And that’s because we thought the market was going to get worse. And we went ahead and sold into it, in effect, locking in a loss, but preventing a much greater loss.

So there, I think if you’ll add up the cash flows bought for each of those contracts, I think you get back to the coverage ratio.

The other thing to add to that is the 3956 megawatts is the capacity of the plant, but it sits in an intermediate pocket [Inaudible] . It’s not a base load, so it’s not going to run all the

time. It’s only going to run when it’s economically viable to do so. So you may have the ability to generate 4,000 megawatts a day, but it’s only in the money 2,000 of those.

If we were going to tell you what percentage we were hedged on volume, we would only do it against the expected output, not against the capacity. These plants are not designed to run 70 to 80 and 90 percent capacity. They’re in the 30 to 40 percent capacity. So that’s also why you see...

About 884 megs is the expected output in ‘03.

Yes, and that’s just to clear that up, that’d be the fourth quarter of ‘03 only, because we’re only doing forward looking. So that’s only the fourth quarter as of September 30th.

Let me take one off the web. And then Ann, why don’t we take some calls from the phone after I do this one?

This one talks particularly about slide 74. And they just want to make sure that they understand what we’re saying. It says slide 74 shows portfolio sensitivity to spark spreads for years after 2011, you’re using a spread that’s $10 to $20 less than Sears estimates.

So if spark spreads in 2011 are only half of Sears number, then your cash flows could be as much as $4 billion higher than your anticipating. Is that right? Want to comment on that?

I will say that assuming that directionally the — I just can’t calculate that off the top of my head, but assuming the number you just calculated is half of what zero says, it’s still above where our line is at. Then we would expect the number to be higher than our current estimate, absolutely.

But I can’t comment on whether it’s $4 billion or $2 billion.

We had some questions from the phone, Jan?

We’ll go first to Scott Siller, Morgan Stanley.

Hi, good morning. A question for you all. And I guess maybe Phil and Bill, probably the best two people to ask this question out, but I think you all are doing a very good job of laying things out. And I think people are starting to understand things better.

Because of the fact that you’ve got what looks like very positive cash flow, it seems as though — what really is holding you guys up from selling to these tolls, because I mean, is it the fact that perhaps public utility companies are hesitant to now move into California so soon? Or is there perhaps private equity people and financial people that would probably be less worried about that? And therefore, might be actually talking to you all as we speak?

And I know you probably can’t say certain things, but can you give people a better feel for — I think most people who follow your company on the buy and sell side are getting more comfortable with how you’re managing this business. And so, I’m just wanting to understand the — what the log jam and actually selling some of these tolls?

Yes, Scott, I’m going to turn it over to Phil in a sec. The — clearly, one thing we’re focused on, those contracts we’re selling, we’re getting pretty much 100 cents on the dollar for. And they’re really risk reducing in general. That’s a big focus. We could sell other pieces of our book, but we’re actually introducing more risk.

So we’re looking for total solutions. And as we’ve had folks come through, I think one of the biggest obstacles has really been the rating agencies in general, that a lot of utility are really struggling when taking on that burden.

But Phil can tell you, we still have folks coming and talking to us. And...

But Bill, they’re on a rating agencies. I mean, I would think that they have obviously much better access to information than most people who follow your stock, you know. And so, what do they need to better understand or what metrics do they need to become more comfortable with, before they would change their tone?

Well, I think in and Andrew and Don can weigh in, but I think what they’re looking at is they’re taking the demand payments and candidly giving us very little credit for the offsetting cash flows.

But you know, Andrew, I don’t know if — I think that’s a fair statement. And I don’t want to speak for the CFO of the company, but I think quite frankly, all we know about is what they’re saying Williams has to do. And from a consolidated point of view, we’ve made it very clear where our metrics have to be to do what Williams wants to do. How that fits within another company’s portfolio is their issue with the rating agency. So I can’t really comment on it.

Oh, Phil said he really didn’t have anything to add.

I guess I would say one thing. Bill’s accurate. A lot of the utilities have probably demurred because of the implications of the merchant risk, that they don’t want to take on in the out years.

I think where we’ve been successful is marketing to match up positions. When we sold Jack into progress, that matched very nicely to a long position that they had in that region. There are other parties are very keenly interested in our positions specifically in around California. And there are also structural solutions that may make sense, wherein we can take the hedged in cash flows and create value out of that.

But I do believe that over time, as the — it is going to take a long time. I think my colleague, Don Chapel, said well, in the quarterly call, it’s going to take a while for us to work out of this. And we’re going to continue to pursue it aggressively.

I just have another real quick question. Then I’ll turn it over. On — and you may have already covered this, even if you have, but just very simplistically, when you’re talking about hedging your hedges on your tolls, is — how much of your tolls are basically on the power — on the power price side are essentially hedged because you basically have agreed with a party, who will contract to buy power at a fixed price per megawatt hour or within a very tight range? So that when people look at this, and even if people don’t understand derivatives, they can understand well, simplistically, they do have a certain percentage just by virtue of having a real physical contract to deliver power to a utility at a specific price.

Well, Scott, I guess — I think what you’re saying is do we have contracts with real counterparts — or utilities or whatever that in effect whatever we would expect to produce, we forward sold to them. You know, California would be the — what we used down

there is we’re 68 percent hedged through 2010. 80 percent of that is with long term contracts with CDWR.

And you look at our Tanaska position, that’s long term contracts with the four Georgia utilities. You look at Ironwood and Red Oak, that’s long term contracts with Allegheny.

So I feel a very significant portion of it. And actually, I guess we could do the math going through the cash flow slides, but very significant portion of it is with long term customers. More than trading counterparties.

And Bill, an offshutting that on the gas side, you’re procuring gas at prices that you’ve hedged through mostly probably swaps in the long term stuff. But it doesn’t seem like there’s a great deal of variance to the actual cash flow on the — you know, on the fuel side and then on the output is all I’m trying to get at. So...

Right, no, Scott, you’re right. I mean, we’re primarily when we’re hedging gas, we’re buying from counterparties, versus through a Nymex. It’s just given our credit situation, we’re allowed to offset. So it makes a lot more sense for us to buy from, you know, whoever in the market, but that to be clear, yes, when we’re selling power and we’re doing a long term contract, we know what our gas needs are and go out and cover that in the market.

OK, that’s it. Thanks a lot.

We’ll go now to Maureen Hale, RBC Capital Markets. Ms. Hale, your line is open?

Thank you very much. On slide 72, I know you’ve addressed a couple questions on the hedge tolling revenue for the estimated hedge tolling revenues. But I’m just trying to understand. You say, if I understand you correctly, that the price to risk is all hedged out. But there is quite a variation just between 2003 and 2004. And I’m just wondering if you can just provide a little more explanation around that line and around what is moving that line as we go forward?

Sure, I’ll try to speak to that. If you’ll go back and recall in our 10-K, that’s the place I’ll start with, because of the credit situation that Williams found itself in mid last year, up ‘til about mid this year, the marketing and trading company was actively liquidating contracts, oftentimes for cash and oftentimes not for the best value. That would not be necessarily a large part of the $600 million, but there was contracts that were liquidated, primarily in the OTC and Nymex markets, as our credit was cut. Well, we had to liquidate for, you know, out of money positions or good positions for less than we wanted to, just to be able to conserve liquidity, and for Williams to be able to survive.

So if you look at the 10-K that we presented at the end of 2002, you clearly see, I think the number was negative $83 million was the next 12 month cash flows. That’s before SG&A and everything. So we clearly knew, going into 2003, that even in the markets didn’t move, that we would be cash flow negative just because of everything we found ourselves having to do, starting in the second quarter of last year. That’s the primary reason why you see a big change from ‘03 to ‘04.

Secondly, many of the long term sales that Bill has outlined, the California Department of Water primarily, the new products A, B, and C and product D, those were the result of a settlement that we did at the end of last year, that the volumes really kicked up and began coming in, in 2004 and forward. The Georgia EMCs, as Bill outlined, don’t come in until really 2005 and forward.

So that’s why we began seeing a real pick-up. And that’s really consistent with the business model that we started in 1998, where timing of the market we decided we wanted to get — to go out and do tolling, and get what we would call long. We’d just get power long. And then we sold it in the peak of the market, so that the cash flows forward would realize all — as the market turned and was in its down cycle, which is where we find ourselves.

It’s just unfortunate that in 2003, we had to take actions to help salvage liquidity from the company that kind of locked us into a loss situation for ‘03. That’s the primary reason for the difference.

Well, then just in terms of further clarification, if there is no price risk on — and if it’s all hedged out, and we are talking about long term sales contracts going forward, why are we calling it estimated hedge tolling revenues?

The primary reason is because two fold. First of all, there’s still credit risk. So if a counterparty defaults, those cash flows will not happen.

But there’s credit risk on all of those items above it. And we don’t see the word estimate used in those cases?

And I agree. And I was going to get to the second point. The second point is if the market goes down, if you’ll recall from the earlier explanations, the value of the tolling agreement will go down, but the value of the hedge will go up.

So the line that you see called estimated tolling revenues, if the markets were to continue to decline, let’s just take 2004, for example, you could see that line, that $175 million in cash flow go away from being called estimated tolling revenue, and be called full performance revenue because it is a hedge. And the value would shift from being recognized in the tolling agreement to being recognized in the sale itself.

But so, so the word estimate only refers to the word, hedge, but it doesn’t refer to the value on the line?

It refers to both the value and the word hedge, because they were talking cash flows here. And the cash flows right now are associated with the fact that our tolls are in the money, but they are also hedged.

If those tolls were to go out of the money, that $175 million of value would shift and be recognized in the hedge, because the hedge is still in place. So it’s called estimated because of where it may end up residing when it’s actually settled.

And that’s just if you look at the example of the hedge, I forget which slide it was on, let’s see if I can find — okay, slide 29 — if you look at the example number 3, that’s like saying that in example 1, the $5 or the $10 sits in estimated tolling revenues. If it goes to example 3 and the market goes down, that $10 shift will shift over and become revenues from the hedge.

Mm-hmm. So I’m with you on that. I understand it’s like $29. So I’m just wondering then does the value go another line on this chart?

Correct.

And that line would be where?

It would go to one of those three or four hedge lines above that, because the tolling underlying is associated with all of the lines above it. It could go to the refill of tolling. It could go to the fixed forward power sales. It could go to the full requirements, any of those lines.

So just final question, to make sure I’m understanding this, even though you’re calling that estimated, other than credit risk, these numbers and I suppose any contractual changes you might make, and you know, I don’t know, monetizing contracts or whatever, but that aside, the sub total is pretty much locked in, other than credit risk?

There’s still a little bit of I referred to earlier in my presentation, not to get too technical, but since these are options, there is still some risk associated with what we call the volatility of the options. As it gets closer to expiration, you can’t — there’s — it’s very difficult to hedge that off.

A perfect hedge of that is a resale of the tolling. You can also do volatility options in the market. It’s not a very liquid market. So there’s some risk associated with that value as well, but it’s minimal at this point.

Any estimate of sort of what percentage risk around — is around the sub total line?

No, but if you’d let me look at some information, I’ll definitely have Travis get back with you with an answer to your question.

That’s great. Thanks very much.

Okay.

Let’s take a couple from the floor here. Sam, once — get that mike coming behind you, Sam. Oh, wait one — that’s OK. Go ahead.

Yes, at the end of second quarter, you had given guidance for operating income of zero to $300 million for power. And then that was increased to $150 million to $300 million in the third quarter. But you’ve been pretty much break even year to date. So I’m wondering what in the fourth quarter is going to be driving this increase in the baseline?

I think if you look at our financials, I think year to date, we’re at the third quarter, we were in the $240, $250 million in earnings range. If you say break even, that probably is because you’re backing out maybe the asset sales.

I don’t know where you’re getting to a breakeven number. And I think I reported earnings at the end of the third quarter were in the $250 million range. So I would assume, just looking forward, that does not assume any further asset sales. And it does assume that some of the gaps mark to market goes against us to get to that range.

But I don’t — I’m not quite sure how to get to your break even number, because that’s not what we’ve reported. We’re reported clearly that we’re positive cash flow. And our earnings was about $250 million.

OK?

Yes, one general, one specific question. Strategically if you look at exiting this business, and clearly you see the value resides in the western portion of it, is there a scenario where you would consider

selling that off, and you could be stuck with the other stuff as near term is under the water?

And secondly, specifically, you alluded to some plant retirement. And we were just wondering if you were making any assumptions in your spark spread scenario about plant retirement in the regions that have not yet been announced?

Bill, why don’t you take the first half of that?

Yeah, I’ll take the first question. I think the direct answer is yes, we would consider selling the California or western position outright. And in fact, are pursuing alternatives to do just that. And it’s a question of value. And if we can get good value out of it, that’s what we will do.

And I think if you’ll run the math all the way through on that, on a reasonable expectation of value, the reduction in interest rates that we would otherwise have to pay if we use those proceeds to retire debt will offset some of the cash flow from operations impact of the residual book and substantially mitigate what might otherwise look like a fairly negative picture.

I think sad to your question on the price curves, we certainly reload, shape and build fundamentals. We’re factoring in things like what we would see as plant retirements. I don’t know, you know, how much of it is — how much — how many megawatts we’re planning to come off with.

I think, though, my understanding when I was getting some explanation as to why the serial curves were so much higher than ours was some of that was they’re just more aggressive on plants coming down than we are. So you know, I think — that can probably be something we could work towards getting you what we’re assuming in that forward curve as far as retirement, but I know it’ll have some impact.

Anitol? (ph)

A couple of questions. Can you give us a little color on the product D? We’ve got the pricing at $140 per kilowatt year, 1175

megawatts, which I think is about $165 million a year. And yet, the resale numbers are in the $120 range. And also, you know, that’s obviously to the DWR from a political standpoint. Is there potential for further negotiations on that, since that one toll, which covers, you know, roughly 30 percent of the capacity almost takes care of the full payments to the — to AES?

And then the second question on the sort of — what we would call in the past, the mark to market value of just over $500 million at the end of the third quarter, can you break out for us how much of that is perhaps gas hedges and how much of that is the A, B, and C contracts to the DWR? Thanks.

Well, I’ll let Andrew address the valuations on the first two issues. The D contract is the mirror — basically just a mirror of our tolls. And we do have what we’re paying AES. Probably most of you in the room know what we’re paying AES as far as demand payments. So the cash flows associated with that would be the result of the spread of those two and should work out to the numbers that we have here.

As far as renegotiating, we’re always willing to sit down with counterparties and renegotiate contracts. Always have. We — you know, we — when we had Hoosier, we renegotiated that three times. The way I view it with state of California is certainly utilities are going to get more involved. And you know, if they’re interested in sitting down and renegotiating contracts will do it, clearly, we’re going to do that with an idylworks (ph) preserving value.

Right. Is that addressing — before we go to Andrew or?

Absolutely. So you’re saying that $120 million is a net number? You’re allocating the portion of the capacity payments that goes to AES for that to net that out?

Right, right. As far as what part of the third quarter mark to market assets is basically gaps or other OTC or other Nymex hedges versus the California, I’m just going to — if I need a more specific answer, I’ll get you more specifics. When I look at the value of the west under the term long term physical forward power sale, that’s primarily products A, B, and C from a cash flow point of view.

So if I just — and at the top of my head look at that number, I’m going to say that probably about half of that total number is that. So this is going to be on a discounted basis. And about half is

going to be the gas hedges. If you need more specifics, I’ll be glad to get that, but I’m betting that’s about how it’s broken out.

Yeah, first of all...

You’re on, buddy.

First of all, I’d like to ask, you know, how wide is the bid ask spread on buying out obligations? You know, is it a couple percentage points of value or is it, you know, 20, 30?

On tolling or?

Tolling. I mean, we’ve seen some examples. People have bought out initially or buying out for 10 cents on a full obligation. And it got to 20 and 23 percent.

Right.

But I’m just curious what the bid ask is, whether there’s room here negotiate, to buy out some of those longer tail obligations or not?

There — it depends on the counterparty. Certainly in discussions with Cleco are going to be a lot different than discussions with Kender Morgan. And so, I would say the bid ask is fairly wide. I mean, you talk to counterparties like AES. You talk to counterparties like Kender, they’re very happy with the agreement. They’re not worried about our credit, see no reason to take a discount.

So not overly optimistic anything will come of that. Whereas with Cleco, we’re certainly working with those guys. And if it comes out a deal where we think it’s worth the value, and they think it’s worth the value, we’ll get out.

But you know, since their credit is improved, at least perception wise, it’s much more difficult to get out of those agreements, buy out.

Okay. One of the other things that we seem to get descriptions of value moving forward is it seems like the bulk of the conversations centered around spark spreads and kind of like what you would get from the energy piece of the portfolio. But the vast majority of people I talk to say that where the real value in plants is moving forward is they’re going to give the energy away on a pass through and its capacity and ancillary services, etceteras.

How do you view contextually the value of the plant, as we go to recontract in extracting the value, because I’m not so sure that it’s spark spread we’d have to focus on?

I think in general, we agree with that. We feel that ancillary services and capacity sales are going to increase in value. As we — as you do that, we think energy will increase in value. We think the net sum of the game will probably be about what we’re projecting. It’s just lower energy values, bigger ancillaries are going to offset that.

Certainly in our values, we do model ancillaries where there are ancillary services. We stick to the liquid points. And as far as the outer points, we’re not really modeling any value in there until that develops.

A good example’s Kender Morgan. We think a robust ancillary market will develop up in Michigan. But until it gets a lot further along, we’re not going to add that to our valuations. But conceptionally, Rick, you’re right. I mean, we think the ancillaries will go up, the capacity goes up, energy will come down. But in effect, the cash flows probably won’t materially change.

OK, and then just one last thing about dispatch. When we get to all the categories except the unhedged piece, which is small, but then grows, how much of that above the unhedged is a function of dispatch? If we don’t dispatch, we really don’t have a margin? So if we get, I don’t know, el Nino for six years in a row in California, what does it do?

Well, yeah, the megawatts that we have, what we call really merchant, but they’re also unhedged megawatts, that’s purely

dependent on dispatch, because we’re not going to dispatch the facilities unless those prices are there.

But as Andrew’s talked about under the hedged cash flows, those megawatts will be dispatched, because we have contractual commitments to honor. So I don’t know if that’s answering your question, Rick, but we’re clearly not going to dispatch when we’re not in the money, unless we have a contractual obligation to do so.

And probably to add to that, Rick, probably I would say product D, since it’s tied to our facilities, would be more critical from a dispatch point of view because like it’s not the kind of product where you can just meet it from the market.

Most of the others you can meet from the market. So dispatch is just our facilities. It’s probably the most critical to product D.

Let’s take one more from the floor. And then we’ll go back to the phones. Is there another on the floor here? Up front?

Thank you. I have two questions relating to your modeling. You mentioned 99 percent confidence interval as your model, your sensitivity. I’m not sure which chart it’s on. Can you discuss your holding period that you use in your value at risk model? Are you using the same holding period for all of your power contracts and in all of the same geographic regions? Or do you apply different holding periods? And what are they?

The second question applies to your mark to market. Do you find that you’re largely having to rely on an internal model? Or are there other sources that you can rely on with a good level of confidence, that are perhaps external as you do your mark to market?

OK, as far as holding period is concerned, what we report in the queue is pretty much not even an FDC mandate. It’s pretty much just an industry standard. We report a one day, 95 percent confidence interval. And I think Williams have been very vocal in the past, saying for our portfolio, that really is somewhat meaningless, because we know that we can’t get out of this portfolio in one day.

So internally, what we look at is we look at different holding periods up to one year. And in — it’s really mathematically, you can convert a one day to a one year [Inaudible] basically by taking

the square root of time. So square root’s 152 days. You can take the — whatever our last reported VAR was and multiply that times that number. And that gives you our one year VAR.

So statistically to get — when we’re looking at trading decisions, we look at the VAR over the tenor of the trading decision, clearly. So if we’re looking at a five year deal, we would look at the risk reduction to the portfolio over the tenor of the bill.

So we whatever holding period is necessary. One year being the least necessary. If we do that, just for 10-Q purposes. As far as the mark to market, because of the fact that most of what is now mark to market is actually settled every day through margins, I mean, you have prices. We’re exchanging cash on gas derivatives, on power derivatives, every day with counterparties or with the Nymex.

So it’s clearly the vast majority of that’s going to be market based. So — and to the extent it’s not market based, since it’s with the counterparty, you agree on a number. So you have not only valid proof by the fact that you settle every day, but then at the end of every quarter, we actually go out from the mid office, which reports to me and validate through broker quotes, through the market, things of that nature that we have the correct number.

So there’s a lot of active data. And most of these are in the period that Bill talked about. And one, two, three, five year periods. So.

OK, let’s go to the phones.

We’ll go next to Derrik Chris, Glenview Capital.

I’m sorry, guys, if — I know we touched on this already, but on page 72, I really want to focus on consolidated numbers going forward. If we look at the sub total line, and somebody asked us before, it appears to us that if you look at the sub total line through 2010, so ‘04 is 16205 is positive, ‘06 through ‘10 is negative, it looks like no matter what happens to power prices, we are still going to be free cash flow positive through 2010? Is that the correct take away?

As we said, there’s two risks that primarily exist in this within the way the numbers are presented, that are not hedged away in their entirety. The first is credit risk. So that can, as we seen over the past two years, greatly impact that number.

If you had another counterparty, let’s just say the size of an Enron, that we happen to have in our portfolio, and they go under, then anything where we were margining them at that time, we already have their cash. But anything from that point forward, obviously, we would be giving away because they’re no longer around.

The second part is the fact that these are options. There is part of the dollars associated with the volatility of an option. If you run a Black-Scholes model, the further from expiration an option is, the more valuable it is. So there is some risk associated with option volatility, but cannot be hedged away except through a mirror tolling agreement, somewhat of a full requirements deal, certain parts of full requirements deals hedge volatility, selling volatility, options in the market themselves, which there is a — not a very liquid market. Or the options basically just move to the spot market and expires either in the money or out of the money.

So I would say that — and I think the earlier pledge, and I think I said that we would try to get some order of magnitude around how much volatility there is in those numbers, because I don’t have that order of magnitude.

But materially, your statement is correct that the power portfolio will generate positive cash flows, excluding the two risks I just talked about, pre-tax, pre SG&A. I mean, that’s the number we’re looking at there. And also pre-working capital changes. So I want to make — and all of those are clearly outlined on that same schedule.

Through 2010.

Through 2010.

Right, and then after that point, we’re hoping that the power market improves. And then we’ll be able to hedge the rest of it.

That would be a fair statement.

Fine. And then my only other question is, as I look at page 72, and I know we have all these tolling demand payment obligations

there, is the gas that is associated with those tolling demand payment obligations, does that gas already bought?

Yes.

So those, you know, even though we have the toll — we have these tolling payments, even if the price of gas moves around, that’s not going to affect our running of these plants?

Only to the extent that we would have to go into the spot markets if say spark spreads were to go up, if we had to go into the sparks — into the spot market to sell the power, we’d obviously have to buy gas in the spot market, but we wouldn’t care because the spread has increased and it’s in the money. And we’d do that.

So those numbers from that point of view probably would only get better.

All right, perfect. Thank you, guys.

We’ll go next to Dalmetto Easy with Royalist Independent Equity Research.

Good morning and well afternoon up there in New York. Thank you for hosting this. Appreciate the detail, but you know, Bill, you guys have never been accused of being aggressive in terms of conservative approach to your business, but I’m wondering from an economic conditioned standpoint that the U.S. finds itself in, versus where your curves are, you know, are you in a best case, worst case, mid case scenario? I’m just curious to where you went there.

And the second question I have deals with — Steve, your cash balance in the company versus, you know, the risk that you see in the power book, etceteras. I mean, do you feel comfortable, uncomfortable, or you know, reasonably comfortable with that?

And lastly, with respect to the discontinued operations of the power business and the book itself, how long can you keep it out

there for accounting rule wise as a discontinued op? Is it, you know, is there a timeline that moves with the next transaction is done? Or is there just not a timeline as I recall? Thank you.

Donato, I’ll take the first one. Then I think Don has got a mike. He’ll take the next two. You know, certainly, we make certain assumptions around load growth, the retirements, etceteras, when we’re building our curve.

I don’t think we’re by any means predicting tremendous load growth or for that matter, tremendous retirement. So I tend to view our curves even given a potential downturn in the economy again as still on the conservative side.

But you know, we certainly aren’t forecasting a depression or you know, sustained recession. Just like we’re really not forecasting any tremendous growth. So you know, that’s about the best I can do to answer, as far as the economic outlook of the country.

It’s certainly changed. If one of those scenarios develop, our curves will change. There’s no doubt about it, as will the real value of the facilities.

Fair enough. Thank you.

This is Don Chappel. With respect to the company’s cash position, and perhaps it’s ability to weather some of the volatility of the power business, I think again Andrew has articulated pretty clearly the range of volatility that we see. And we believe it’s in the $300 billion range, excuse me, in terms of a potential margin requirement.

However, we’re retaining substantially more cash than the $300 million to provide us with that extra degree of assurance, as well as just some work-in-capital volatility.

At this point in time, we have a cash balance, unrestricted cash balance of about $2.7 billion. We have debt obligations in the fourth quarter of about $200 million. We have debt obligations in the first quarter in terms of debt that’s coming due of just a little bit less than $700 million. And then another $200 million, I believe, in the fourth quarter of next year. And then a very modest amount in 2005.

So you know, we have a substantial cash balance that would allow us to retire the debt in 2004, 2005, as it comes due. We have additional assets that are for sale. Some of which we’ve announced

and expect to close perhaps some in the fourth quarter this year. And certainly a very large set of assets expected to close next year.

And that will also add to our cash balance. And then finally, we have good reason to believe that the cash flow from operations and free cash flow will be net positive in 2004, 2005. And we’ve detailed that in our third quarter conference call. And that’s available on our Web site. So the net of all that we feel very confident that we’re in very good shape to handle whatever volatility that could be reasonably expected in this power business.

And the last selection as far as discontinued ops, from an actual GAAP financial statement point of view, the power company is not classified as a discontinued op. We don’t quality for hedge accounting, but that’s not the same as saying it’s a discontinued op. And the qualification to turn back to hedge accounting would be number one, either senior management says it is their stated intent to stay in this business and withdraws all action to try to sell the business. Or ultimately whoever buys this business gets the luxury of being able to take hedge accounting.

Thank you.

We’ll go next to Winfried Fruhoff, with National Bank.

Thanks very much. I have a question on, well it doesn’t really matter, slides 15, 16, and so on, that deal with sparks spread.

And I do appreciate that you’re endeavoring to help us model numbers for your company. But when it comes to calculating spark spread, there’s a great difficulty in determining the appropriate gas price to be used, because every plant has a specific — plant specific gas price, which is not identical to Nymex or anything related to Nymex, because there’s transportation involved. In some cases, maybe even distribution. So how much help is it really to have spark spread data, when we don’t really know the effective delivered cost of natural gas to each of the plants?

Well, Winfred, the reason we’re trying to show it the way we’re showing it is really so you can make your own assumptions around what you think gas will be and power will be. And you model it

yourself and come out with is our portfolio overvalued or undervalued in your view?

I mean, you’re absolutely right. Each one is a specific location, but they’re all fairly liquid basis locations that — where our facilities are, that you can — they’re city gates, basically, that you, you know, can look at basis relationship to the Nymex and then look at the power prices.

So I mean, I agree with what you’re saying. We do not give you prices by location. We didn’t give you power prices by location. We gave you spark spread assumptions.

So that could be something we could think about in future calls, or future disclosures, but I really — what we were hopeful is that we would give you enough data that you could model it yourself and say either the cash flows are less or more.

Well, given the kind of really poor publicly available data in every regard to your reference to price reporting a little bit earlier, what kind of guidance would you offer to us in general terms, as far as approach is concerned to trying to approximate plan specific to the gas prices?

The only advice I can give you, there are publications out there. Certainly CERA, among many, Pyra. Others who forecast by region forward gas prices, forward power prices. Some companies do their own fundamental analysis in their own house and come up with their own points of view.

But you know, really, that’s about the only thing I can tell you is go to an industry publication. Or you can use Department of Energy. They publish data, but go to one of those and you know, see if you tend to agree with them. And if so, then plug that into your assumptions.

What does your company use?

We use really a combination of fundamental analysis under Andrew’s new responsibilities, we call it point of view, but we certainly have a point of view, Williams does, to where gas prices are going, where power prices are going. NGLs, crude, etceteras, everything that affects our business.

We certainly look to outside publications, too, to see if our assumptions are reasonable, conservative, aggressive, and try to reconcile there. But really, we do our own in-house analysis for the most part.

All right, thanks very much.

I’ve got a couple more here from the web. And I think that’s all that we have on the phone. And then we’ll go to any left here in the room. First one is about spark spreads. Bill and Andrew, please comment further regarding forward spark spreads used in your cash flow assumptions relative to historical levels?

I can’t give you metrics. I could give you by region that there are forward spark spreads. It depends, again, what you want to call historical. Certainly take the anomaly of 2000 out. I would say we’re still very much on the lower end of where you’ve seen historical spark spreads on a go forward basis.

They do improve as regions come into in effect supply demand comes into a balance, you’re going to see spark spreads return to historical levels. But in the short term, and in the next five years or so, I would venture a guess, we’re very conservative versus historicals. And that’s throwing out 2000 because that’s purely an anomaly.

And this last question from the web, Andrew, would be for you. If you intend to get out of your power business entirely, can you and will you report the power business separately from your integrated gas and the new Williams?

The answer to that is yes, we can. Will we I think will be a decision made by Don Chapel and Gary Bailiff, our chief accounting officers. If they ask me for my input, I’ll be glad to give it, but I think the one thing I would say is that — the one thing we were careful of is anything we put out in the presentation today we knew would be at least a required minimum for every quarter going forward from an actuals and a forecast point of view, as we pass to whatever guidance we’d be giving.

So at a minimum, my team is prepared to begin preparing actuals and estimates in the same format you’ve seen here today for every

quarter going forward until we’re told differently. So I think that should provide a lot more flavor on a quarter basis going forward around the power company, which may preclude us from having to break it out separately, but there’s no legal reasons that I can think of not to break it out separately, as long as it’s consolidated.

OK, other questions from the room? I know we’ve had a long morning, what we’ve gone through. Well, it doesn’t look like there’s any, so let me just finish off by thanking everybody for coming. Those of you on the phones and on the web, we don’t have lunch for you. But those in the room, I think there is some lunch out. I don’t know if it’s ready, but it will be, if you want to hang around. And we appreciate everybody coming. Thank you very much.

Once again, ladies and gentlemen, that concludes today’s call. Thank you for your participation. You may disconnect.

DISCLAIMER

CCBN reserves the right to make changes to documents, content, or other information on this web site without obligation to notify any person of such changes.

In the conference calls upon which Event Transcripts are based, companies may make projections or other forward-looking statements regarding a variety of items. Such forward-looking statements are based upon current expectations and involve risks and uncertainties. Actual results may differ materially from those stated in any forward-looking statement based on a number of important factors and risks, which are more specifically identified in the companies’ most recent SEC filings. Although the companies may indicate and believe that the assumptions underlying the forward-looking statements are reasonable, any of the assumptions could prove inaccurate or incorrect and, therefore, there can be no assurance that the results contemplated in the forward-looking statements will be realized.

THE INFORMATION CONTAINED IN EVENT TRANSCRIPTS IS A TEXTUAL REPRESENTATION OF THE APPLICABLE COMPANY’S CONFERENCE CALL AND WHILE EFFORTS ARE MADE TO PROVIDE AN ACCURATE TRANSCRIPTION, THERE MAY BE MATERIAL ERRORS, OMISSIONS, OR INACCURACIES IN THE REPORTING OF THE SUBSTANCE OF THE CONFERENCE CALLS. IN NO WAY DOES CCBN OR THE APPLICABLE COMPANY ASSUME ANY RESPONSIBILITY FOR ANY INVESTMENT OR OTHER DECISIONS MADE BASED UPON THE INFORMATION PROVIDED ON THIS WEB SITE OR IN ANY EVENT TRANSCRIPT. USERS ARE ADVISED TO REVIEW THE APPLICABLE COMPANY’S CONFERENCE CALL ITSELF AND THE APPLICABLE COMPANY’S SEC FILINGS BEFORE MAKING ANY INVESTMENT OR OTHER DECISIONS.

© 2003, CCBN, Inc. All Rights Reserved.