ING Groep NV (ING) 2008 Q2 法說會逐字稿

Ladies and gentlemen, thank you for standing by. Welcome to the ING Groep second-quarter results 2008 conference call on August 13, 2008. Throughout today's recorded presentation, all participants will be in a listen-only mode. After the presentation, there will be an opportunity to ask questions. (Operator Instructions).

I would now like to hand the conference over to [Gemma Backs]. Please go ahead.

Good morning, good afternoon. This is Gemma Backs for ING Groep welcoming you to ING Groep's second-quarter 2008 results conference call. Before turning this over to John Hele, CFO of ING Groep; Koos Timmermans, Chief Risk Officer of ING Groep; and Tom McInerney, Member of the Executive Board and CEO of ING US Financial Services.

Let me first say that any forward-looking statements in today's comments are subject to a number of variables. They include interest rates; foreign exchange rates; inflation rates; movements in securities markets, including equity markets; and underlying economic health and changes. The realization of forward-looking statements could be materially altered by unexpected movements in any or all of these and other variables.

That said, good morning and good afternoon again, John, Koos and Tom. John, over to you.

Okay. Good afternoon to everybody. We will just give a very quick introduction and then turn it over to questions. I'm here, again, with Koos Timmermans, the Chief Risk Officer, and Tom McInerney, Executive Board Member for Insurance Americas.

Just in terms of highlights for the quarter, we had underlying net profit of EUR1.9 billion, down 29% from a year ago, basically because of lower investment returns and investment income. Our earnings per share was down 20%, and that is because of the EUR5 billion share buyback that we did over the last year that was completed in May. We had a limited direct impact from the credit and liquidity crisis in the second quarter. However, we're not immune to the challenging environment around us, and it has affected, to some extent, some of our business units.

All of our capital and leverage ratios are well within target, which a spare leverage capacity at the end of the quarter of EUR3.9 billion. And we continue to focus on our commercial growth, adding almost EUR30 billion of client balances in the quarter to be almost near EUR1.5 trillion client balances, from which we receive fees and margins from. We will continue to focus on our core business while keeping mind of the challenging circumstances in which we are operating.

With that, I would like to turn it over for questions.

(Operator Instructions). Bruno Paulson, Bernstein.

Three questions, if I may. Firstly, looking at your pretty sharp loan growth in Q1. I think the risk-weighted assets went up by 4.5% or something of that order. With that, how much of this was, if you like, voluntary, and how much of this was companies taking up lines of credit which have been advanced. Is there any evidence of procyclicality, i.e. the risk-weighted assets rising relative to assets as loan losses rise? And are there capital implications? Is this kind of loan rate sustainable going forward, and what is your planned growth in risk-weighted assets over the next few quarters?

Secondly, just a very quick mechanical question. What is the impact of the already-announced acquisitions and disposables on the spare leverage over the next few quarters?

And finally, a deeply obscure one -- on page 67 of the fiscal supplement, the realized gains and losses on equities in European insurance dropped from EUR70 million to EUR20 million. My understanding was that this was a fixed proportion, the 3% annual of the equity holdings. I was wondering why it dropped so dramatically. Thank you.

Why don't we turn over the first question on the loan growth to our Chief Risk Officer, who personally looks at all these loans, I think, on an ongoing basis. So, Koos?

I think on the question of the growth of risk-weighted assets, and is this due to the fact of the usage of either commercial paper or backstop lines, one of the things is, the answer in short there is no. It is not a sort of involuntary growth of credit against the precrisis level of these limits.

I think we do have some of these backup lines for corporates available. At the same time, we always have to make sure that we have liquidity for it. But actually, we don't see the utilization there. What we did see, however, is selective growth in the wholesale portfolio because of the fact that there is a bit of a bond market working less than it was in the past. You see a bit more club deals. So in that sense, you're growing something. And the other aspect is there's always the occasional changes of methodologies and things. But I would say a part of it is indeed more usage of bank balance sheet, but not the involuntary part.

And the question on if we're going to spare leverage over the next few quarters, we have some acquisitions we have announced that we have closed, as well as some we've ended up selling, and suffice to say that these are all factored in and we don't expect really material, massive changes in our overall spare leverage capacity from these. We will be paying a dividend, though, of course, the interim dividend, which will have an impact on it. That's factored in for the next quarter.

Your last question, I think, is on the gains and losses. And if it is perhaps regarding in Insurance Europe, the reduction of gains and losses on impairments and equity securities in Insurance Europe?

Yes, that's it.

Yes, being EUR20 million versus EUR70 million, which is on page 67 of 188 of the statistical supplement.

Indeed, (multiple speakers) 67, but I found that.

That is an impairment of -- it is actually an asset-backed debt fund that is pooled, and it is recorded as an equity security. So some of the underlying credits were actually impaired, and it is disclosed. We spoke about it in the Insurance Europe pages. It is a EUR49 million impairment. But it is classified as an equity, even though it is debt securities within the overall equity structure. It is pooled between a few units, and that is why it's structured that way. That is why that is in the business line versus normally impairments would be in the corporate line.

Duncan Russell, JPMorgan.

Just a quick question. Slide 56, on the impairment test for the Alt-A, just wanted to make sure I'm reading this right. Is it correct to say that you will only impair the Alt-A book when the estimated pipeline loss is greater than the credit enhancement? Is that what this slide is saying? Thanks.

I'll start this and then Koos can add to it. No, this is a screen by which we review these. And if you are less than 1, then we clearly look at it to impair, unless there is compelling evidence to the contrary, such as perhaps prepayments, because of a senior structure, might give you higher credit protection than what this initial screen does. But anything 1 to 2 is also very carefully reviewed and is done with a detailed cash flow and structure as we review that, to be on our watch list. And we use a model that takes into account prepayments and other factors to review, and also look at the structural credit protection versus this sort of a point-in-time structure. Koos?

Yes, I think you are right, John, in the sense that if you only look at credit enhancement and billed losses, then the one thing what you, for instance, miss is how far along is this bump already on the road. I mean, if it is close to maturity, then they're supposed to have eaten up their credit enhancement, so then it is less likely an issue. But in essence, it's just like John says, below 1 it should be less, and above 1, it is more suspect.

But indeed, there is quite a few criteria which play a role. It is credit enhancement versus losses. It is the excess credit which is still on the bump and the level of prepayment which you get there. But at least filter one works this way, and then you do the thorough analysis, and then you impair based on that.

Just to be clear, just on the accounting policy again, if it is in category 3, i.e. coverage ratio greater than 2, it will definitely not get impaired. Is that what this is saying? And if it is in category 1 and 2, you review it and then you might impair it. But category 3 definitely does not get impaired through the P&L?

In essence, what it says is, this is with our first filter, category 3, a bump does not come out at this moment as something which is likely to be impaired, given the current assessment which we are making. So in essence, it is level 5 is more an absolute no. Level 3, yes, you start to pay already a closer look at it. And level -- the first and the second column are clearly the ones which you have in an impairment committee and you look at very closely.

And what would happen if your security got downgraded materially by S&P, but it was still in level 5?

To be honest, I think we should spend the most time on looking at our own credit assessment there. Nevertheless, if at a certain moment a bond gets downgraded, then that gets done, because there are revised expectations of ultimate losses. And one way or the other, if that is materializing, something will fall through the crack of these filters, and something will migrate from level -- the bigger than 5, 2 to 5.

Ultimately, I think what you see happening is downgrades are happening based on loss estimates. Ultimately, loss estimates ultimately are fed by data on the current pipeline and this current pipeline, times a multiplier, is what you have here. So the things are very much linked, and it might be a matter of quarters or something like that in between. But things in general progress in sync. But a downgrade as such is not an automatic recategorization.

But if it was a dramatic downgrade of several notches, we would review that security very carefully.

And then estimated pipeline losses, again, just to put it into simple terms, what does that assume about house prices, as it were? Is it house prices as they are today, or has it factored in, so when you've got 5 times covered, EUR14.9 billion, is that assuming further deterioration in house prices? What does it assume for house prices?

I think what you've seen under there is we take a loss given default for subprime of 50% and for Alt-A of 35%. And if we then work our way backwards, what you could say is you have an Alt-A security with a 70% loan to value, but, okay, there is a 30% cushion there. But that will be eaten up basically by foreclosure costs and all the other lot because of foreselling. So in essence, from now on, you could say there's a 35% house price decline which you anticipate in the security there.

This formula seems to be saying 60% times 60-plus delinquencies, 90% times -- so that's, as of today, the credit outlook as of today doesn't factor in -- so when you say estimated pipeline loss, that's just as of today. What you're already seeing today is not factoring in further delinquencies. It's not an estimate (multiple speakers)

We started with a 70% loan to value in general on the securities. And house prices are down, depending upon the area, 15% or 20%. And if you use a 35% loss given default, you already have factored in some future elements of call it house price declines, foreclosure costs, time value of money before you can ever have an ultimate value. But we are down half of the -- under half the value, ultimately, that you are selling from when these were originally issued.

No, I agree with the loss given default. What I meant was the delinquencies is going to go up, isn't it, surely?

I think what you can see is that is a bit -- but then you come to the point of the forecasting. Delinquencies is going to go up. If we look at the S&P's scenario, what they say is they have doubled their ultimate losses, what they expect. And that is the sum of two things. I think they see a slight deterioration of the house prices, so they increase the loss given default, and that is a bit in line with the house price decline which we already had. So they made that factor of LGD bigger. And at the same time, they significantly increased in their scenario the amount of PDs. So that is why they came up with their doubling of this estimate.

So there is no doubt there is a deterioration, and that is something which is being factored in in these revised outlooks at the moment.

I won't take up any more time, but I just wanted to get one thing off this chart again. What if you put in the S&P estimated delinquencies in? What does this chart look like?

I'm sorry that the S&P's estimated delinquencies is something which I don't have over here. The only thing what I have is the ultimate losses per the different vintages. The way how they build it up with delinquencies, if I take the example on the neg-ams, I think ultimately they are going to end up with a 15% factor of ultimate losses for the 2007 and 2008. Given the fact that it is a market practice to take at this moment a 50% LGD on neg-ams, that means that they assume delinquencies in order to approach approximately 30%.

But it is very much a product class and vintage-driven way how they analyze it. So that means, like, I would have to come up with basically more than -- well, actually, over the different rating classes and products, 50 different delinquencies and house price declines. But that is why I gave you the most important, the '07 neg-am, 30% delinquency, 50% LGDs. That is something which is factored in their scenario.

That's the same scenario as applied to our Alt-A, gives us the EUR200 million in credit losses, or at these market prices at June 30 would have been a EUR400 million impairment.

Farooq Hanif, Morgan Stanley.

I had a few questions. Firstly, I'm afraid I'm going to dart all over the place, but if you go to slide 85, which is looking at your capital position in group and how it is downstream to bank and holding, one thing that has confused me that I'm sure you have explained before, but I can see how the equity from the group is injected into ING Bank and ING Insurance and how the core data as well goes into that.

But I notice in ING Bank you've got this Tier 2 capital of EUR12 billion, which doesn't seem to appear in the consolidated balance sheet. I was just wondering where that is in the consolidated balance sheet. That is question number one.

Question number two is, when you look at your loan provisions against the kind of provision loans, the nonperforming loans, that ratio now is about 35%. And I've noticed historically it never -- it hasn't really fallen to that level, as far as I remember, for a long time. So I was just wondering, how come that is low? Is that normal? Should we expect it to go towards the 50% or 45% level in our modeling?

The third question, to tie you up in knots even more, the CDO and CLOs that you have been buying through derivatives, am I right in thinking that that will basically move, using fair value accounting, to the Insurance Americas and the Wholesale Banking line?

Let me know if you want me to repeat any of those questions, by the way.

You wonder where the Tier 2 consolidated is, and I believe it is in subordinated loans when you consolidate across as eliminations. Of course, we have subordinated loans on both sides, bank and insurance. I'm seeing the Tier 2 show up as subordinated loans in the balance sheet.

But you don't show it -- the reason I ask is that you're not showing it in ING Groep Holding in that chart anywhere, so I just got a bit confused.

Oh, yes, right. It is because that is the parent. That is the holding parent. So this is not consolidated; this is on the balance sheet of the holding, on page 85.

Right. So we will -- I mean, if I look in your total balance sheet in the actual books, I should find that EUR12 billion in the proper consolidated balance sheet somewhere? (multiple speakers)

Yes, it is EUR9.6 billion or it could be split in the debt securities and issuance, depending upon the categorization of the Tier 2.

Okay, as long as it is not double-counted, that is the main thing.

(multiple speakers)

And sorry, the question about provisions versus the provision loans?

I will turn that to Koos for the second question.

If you talk about, well, the first thing, just to make sure, if you talk about provisions as compared to credit risk-weighted assets, there you could say ours doubled. But that is because the credit risk-weighted assets were cut in half. So that is the Basel I, Basel II effect. But what you are alluding to more is the amount of provision versus nonperforming loans.

Yes.

So in other words, the level of provisioning for a nonperforming loan is lower at this moment. Is that what you are --

Yes. And it just seems -- but I know it fluctuates, and there are probably very good reasons for that. But the ratio of 35% just seems well below where it has been since I've looked at that for the last two years.

Maybe I can ask, just on this one, I can ask [Payler Stoll], who is the head of credit here, to comment on that one.

That ratio is very much determined on what your profile looks like. If you have a very much secured portfolio, then your loss given default is actually less. So if you are very, very secured, let's say you have a EUR100 million loan with a EUR90 million security and they're outstanding, and you have a provision of 15, then your ratio is 150.

Now, we are a very, very collateralized balance sheet, both in our real estate operations and [we've got all this also] in the home retail operations, where we have underlying mortgages. So for a traditional wholesale bank with more unsecured loans at holding companies, you would see that ratio much higher than with us.

On the other hand, it is also impacted on what your write-off policy is. If I had a EUR100 million loan against -- with a EUR100 million provision, and I write it off on day one, I basically influence my ratio tremendously, because if I keep it on, then I keep that 100% on the books. And we are very, very conservative in actually writing off provisions. Our write-off policy is very strict.

Your third question?

Sorry, my third question was -- sorry, if I just follow up on that one point, the only reason I ask is that the ratio looked low compared to where you yourself have been, Q1 and for all of 2008 and 2007. So I just wondered what the reason was specifically in Q2 for that ratio to fall.

On the gross additions versus the release?

No, the loan provisions divided by provision loans within your -- for example, this is on page 139 of your --

I guess that has to do with the higher percentage of releases that are in there as compared to the actual growth and additions to that book.

Yes, so we're basically adding to it more now. I think [it's the assent], the cycle is kind of turning that we see happening now.

And then the last question was just on the protection contracts on CDOs and CLOs that you are buying to get access to those spreads. I mean, that's going to be fair value accounting, right?

Yes, the ones we wrote. I would turn it over to Tom McInerney.

Yes, those would be in the Insurance Americas, general account portfolio. It's well-diversified, investment-grade corporate credits, with very good spreads [as far as spreads]. So we took the opportunity, and we think the returns, as was stated in the call this morning, will be in the 20% range. And they added roughly in the order of EUR10 million of additional profits which would have been booked, if you look at the statistical supplement in the Financial Products segment of the Asset Management Group.

And those are recorded at fair value.

Right.

[Yap Meyer], Dresdner.

On your CDOs, you only provided a net exposure rather than a gross exposure. Could you provide that, and then let's say split up in corporate, multisector and ABS? We know the subprime, but those three sectors will be handy.

In terms of monoline exposures, could you provide the underlying assets, the valuation metric of those underlying assets, the valid risk and the provisions you have taken for the monoline exposures?

And finally, on your spare leverage, it's EUR3.9 billion, but if I remember well, it includes about close to EUR5 billion benefits from Basel II. So I'm just wondering how rating agencies are looking at this. They already recognize Basel II because initially they were not really willing to embrace Basel II. Thanks.

Why don't I start with the last question while everyone looks up the details for the first two question. The EUR3.9 billion, well, we are formally on Basel II. We file with our regulator. It is being fully reflected. We have had meetings with our rating agencies. We have no fundamental arguments about it moving forward.

So they certainly have never said to us that they don't believe in Basel II. And we, with our regulator, have been validating and even audited all of our calculations across the board. So it does include this benefit, but we are fully on it today. And we have set our ratios and communicated that to the rating agencies, and that is how we are operating. And I don't expect any problem from that.

Koos, do you have the gross and net exposures?

Sure. I think if you look at gross and net, that is especially applicable to the bank notes from our insurance company, because they are a net investor. I think on the banking side, you have some arbitrage type of business. And if we follow that portfolio in the cash CDOs, it is only a very small amount, I think EUR83 million. And that is net sold protection. That is the cash part.

But the bigger numbers are in the synthetics. There, you can say that at this moment -- first take index-related. Index-related sold protection, EUR1.9 billion; bought protection, EUR1 billion. So this is typically iTraxx tranches [offset] each other. If we take the bespokes, there I have EUR1.4 billion on each side, bought and sold protection. And if you look at -- so that's the synthetic CDOs. If I look at the CLOs, that is EUR100 million or something like that sold protection, so a small amount. So that is the totals.

You also had a question on our monoline exposures, but could you say what that is again?

What are the underlying assets? What is the valuation metric of the underlying assets, and therefore, what is a valid risk relating to the monoline sector, and what provisions have you taken on the monoline sector as a whole?

Sure. If we look at the monoline exposure, first, overall, the amount of insured money is EUR2.9 billion. Now, again, the interesting thing is it takes a double type of default before you get into a problem. The first is the monoline and monoline insurer itself, and secondly is the underlying asset.

So let's first focus on that one. What do we have with the monoliners? We have a few things. We have on the insurance side some wrapped exposure. We have on the bank Mont Blanc, so that is our customer-sponsored conduit. We have some wrapped exposure. We have some bond wrapped AFS books and ING Direct.

Now, overall, you could argue that, especially on Mont Blanc and the wrapped bonds, that is EUR1.5 billion out of the EUR2.9 billion. Those are assets shipping in that conduit or bonds underlying where if at a certain moment the insurer would go bust, in that case, we would not lose directly any money because the ultimate assets underlying it are held at cost. So in other words, they are not bonds in [the trading world].

Secondly, on the insurance side, we do have assets which are wrapped by an insurer, and that is something like below EUR500 million. That part, in case the monoliner defaults, you will find that the security gets a lower value. In other words, then you will get a EUR50 million to EUR60 million revaluation. That is something what we see happening at that moment because of downgrades.

The third element is ING Direct. There we do have some exposure which is wrapped as well, and that part would revalue, but only with relatively small amounts. So in other words, in your held-to-maturity part, there you don't see a lot. On the available-for-sale, there you will see in total an amount of -- let me see -- EUR75 million.

And then we still have one other part, and that is on the baking side, the derivatives. We have of EUR630 million of derivatives outstanding, CDSs in the trading book. And there, I would say the revaluation gain or losses, we are talking at the moment about EUR44 million.

So underlying, in a nutshell, if something happens with the monoliner, then the effect goes either directly via P&L, because we have some derivatives which cease to exist, and those derivatives, they have EUR44 million value for us. We have some loans in the portfolio. And there, you could say, well, slightly the loan in terms of repayment, yes, that has a slightly lower quality, but not a direct P&L impact. And you have the available-for-sale segment, so where you have wraps on bonds. And that is where you have in total the EUR75 million of value decline anticipated once these monoliners don't exist.

So how much are the underlying assets, then, underwater currently? So all the three together, even held to maturity (multiple speakers) value?

EUR75 million plus EUR44 million, essentially.

They gave US amounts, is more the value declines once this happens and the wrap is out. I don't have a precise number of the current value of these assets.

Yes, this is the impact of the wrap on it, is if all monolines would go bust tomorrow, we'd have a negative pretax P&L of EUR44 million and at the same time a negative pretax revaluation of EUR75 million.

I think the most important part is what we are ensuring is underlying assets there, and those underlying assets are performing assets.

Excellent. One question, actually, on Belgium. It is performing quite poorly, I think, the Belgian retail unit. Cost to income ratio is going up; revenue is coming down. And probably the outlook doesn't look very good for the second half because of the war for savings, and very little repricing in the mortgage side. Any plans there to address the situation?

Absolutely. In fact, some of the higher costs in Belgium in the quarter is some of the repositioning of the whole banking system there, where we are revitalizing the branches. We are shrinking some branches down to be more self-service, and also building more products to be sold through the other branches.

And in fact, there is quite a whole system to work to bring through next year -- some of the technology being built this year in the Netherlands will be applied next year throughout Belgium. So we believe that with the savings wars going on now in Belgium that we have to address the structural costs across the board. We already started a retro transformation program. We plan to accelerate that in the coming quarters. It does take time, though, obviously in the Belgium economy to restructure these things.

Is that a region you want to add on to increase the footprint to collect synergies, or is that not an option for you, if let's say one of the players were to sell their retail unit?

It depends what is available, obviously.

Farquhar Murray, FPK.

Four quick questions, and I do actually promise they are quick, all related to slide 56, so following on from Duncan, really. The first question is, would you have the unrealized loss associated with each of the bars, just in order to help us gauge the impairment risk attached to each? And then secondly, could you help us reconcile between the EUR35 million impairment in the second quarter and this graphic? What I'm trying to understand is whether all the impairments came from the less than one times cover bar. Presumably not, because it would imply a par valuation of 24%.

And then finally, with regards to the EUR400 million scenario, would those impairments stretch beyond the less than one and one to two times bar? and obviously that is similar to Duncan's question. And that is it. Thank you.

If you want, maybe let me start with the part -- well, let me start with the last question. If you look at the EUR400 million, in essence what we have done with this, as a B scenario, we have said, like, recalculate ultimate losses, but we do it a bit in a funny way, because we say, like we calculate those ultimate losses that they are going to happen over time, but we are not going to benefit from any of the credit spreads or any of the prepayments. Let's see what happens to our portfolio.

In such cases, what would happen is, in total, we fell EUR200 million of credit enhancement short. So in essence, EUR200 million of credit enhancement short would mean my bucket smaller than 1 would fill up. But here indeed, one of the points is, is that 20 tickets of EUR10 million, or is that 40 tickets of EUR5 million? That is one of the things which is a bit of a difficult one to determine, although what you can see is, with the S&P scenario, we looked at the total amount of bonds, 126 of the RMBSs, they would reach a credit enhancement.

I think, so in essence, it is a bit of a difficult one to forecast precisely, the EUR200 million credit loss in case S&P's materializes, that is a certain. But at the same time, would it materialize into EUR400 million impairments or EUR300 million or EUR500 million, that is a bit more difficult because it is the distribution of the bonds which play a role. But I don't have precisely over here how big in notional amount bar number one would be.

If I look at the total impairments, what you see is that most of the second-quarter impairments come from bucket two. So in other words, we have some from bucket one, but we also have -- the main part of it comes from bucket two. So you are right in saying that doesn't come from the bucket one only.

And then there was a last question, which was the total division of unrealized losses over the different buckets. And that is one number which I owe to you at a later stage, because I don't have that. What we did here is a credit impairment test, but then I would have to slice the bonds and take the realized losses or the unrealized losses of this and select it over these credit enhancements. I don't have this breakdown directly here. But I would be more than happy to send it later.

Marc Thiele, UBS.

Two questions for Tom, if I may. On variable annuities, can you give us a bit of flavor on what you think about the outlook? We had a number of US (technical difficulty) a very competitive environment. Do you anticipate more pressure on sales as well as the IRR? And could you also give us an update on what you think about the hedging costs going forward and if you think you are in a position to hand this over to the policyholders as well.

My second question is regarding the composite margins. Again, talking about competition, credit and rates slightly moving up, do you expect that to continue? And maybe you could also give us a bit of flavor regarding the outlook for credit losses.

I think on the variable annuity book, I think if you look in the second quarter, our monthly sales were about $950 million -- these are in dollars, in the fiscal supplement on page 8. You know, I think with our sales over the last few quarters, that would be roughly my expectation, although as we pointed out in the press release, it is softening modestly from where it has been. I think the IRRs, we have been pricing for 14% IRR, with the increase in the cost of the hedging. We have been earning -- we earned in the second quarter 12.7%. So we will look at the hedging costs and may pass that back through to consumers. I think all the competitors are probably looking at that.

In terms of the margins on crediting rates, if you look at that sheet again in the statistical supplement, our crediting rates have been roughly in the 400, 405 basis point range if you look over the last six quarters. I think it has been -- it's varied to within 4 or 5 basis points. So I would tell you on our book, given the fixed annuity crediting rates on the individual annuities and in our [retirement] services, I think those crediting rates roughly stay where they are, because obviously it is averaged over the entire portfolio.

And where do you expect corporate bond defaults to go from here?

You know, I think that we obviously go through with Koos and Koos' team, and my sense is overall, as is pointed out in the S&P work that we've done, we would think there probably would be some increase in defaults over time. I think the sheets that Koos reviewed are probably the best indication of what we're seeing.

Yes, I think that maybe to add there, in general, like on the banking side with loans, you do expect a bit of an overall increase to a long-term average of the credit risk-weighted, you know, base point credit risk-weighted assets. You see no reason to say that on the corporate bond diversified side, I mean, a slight increase of risk is probably what you will see there as well.

A correction -- I have the technical answer to the balance sheet on where the bank Tier 2 goes to. It's on page 12 of our consolidated balance sheet in our Quarterly Report. The Tier 2 is shown as subordinated loans in the bank. It has been eliminated on consolidation and moved from subordinated loans to other borrowed funds, because it is considered subordinated Tier 2 for ING Bank NV, but at ING Groep it is considered other borrowed funds. And you can see it is -- the EUR12 billion is within the EUR16 billion, roughly EUR16 billion that is eliminated and added between the two categories. So that is where the money is going to. And It is fully recorded.

We will take the next question.

(Operator Instructions). Bruno Paulson.

On the US, what I was wondering about, it is not so much the credit losses that you were just discussing There was also the $120 million interest and other market-related investment losses from page 7 of the US fiscal supplement, which obviously was a bit of a turnaround from quite significant positive numbers in the two previous quarters. I was wondering if you could tell us what that is, what drives it, and how we could think about what that is likely to be going forward.

I think that breaks down into credit-related gains and losses net of DAC, of $48 million, and then trading gains and losses were $22 million in the second quarter. And then on our CMO portfolio, we have derivatives, and based on the steepness of the yield curve, there is what we call cumulative catch-up provisions. So based on the value of those derivatives, and that in the second quarter was $58 million. That will bounce around; in last year's second quarter, it was a positive $7 million. So that was a big, if you look at quarter over quarter, a big driver of the difference.

So that $58 million, that costs you when the yield curve steepens?

No. When it steepens, it actually helps that.

Right, so it's a flattening yield curve (inaudible).

Correct, yes.

(Operator Instructions). There are no further questions at this time. Please continue with any further points you wish to raise.

Okay. We would like to thank everyone for their attention, the outstanding questions. We will make sure Investor Relations has communicated to everyone as we gather that number for people. Thank you for your attention, and we look forward to talking with you next quarter. Thank you.

Ladies and gentlemen, this concludes the ING Groep second-quarter 2008 results conference call. Thanks for participating. You may now disconnect.