Interview: Casper Rahbek Andersen, Standard & Poor’s

Casper Rahbek Andersen, director in Standard & Poor’s covered bond team, spoke with Nordic Financial Institutions & Covered about the rating agency’s perspective on the Nordic jurisdictions in light of recent regulatory developments and the economic challenges in Europe.

What do you make of the Norwegian authorities’ consideration of some form of curbs on covered bond issuance?

Casper Rahbek Andersen
Director in S&P’s covered bond team

First of all we have not been involved in the discussions, but from what we have seen, it seems that at the moment they are very much discussing among themselves how to go about doing this and whether you need limits at all. Let’s see what happens. It is a bit premature to start talking about any impact on covered bond ratings, which I wouldn’t expect. It will probably have more of an effect in terms of issuer ratings and our colleagues in the financial institutions team have published an article about how they view asset encumbrance issues.

Are there any other regulatory initiatives in the Nordics that are affecting covered bonds?

Again, we have not been involved in the changes but to our understanding there have been various clarifications and the latest was from Sweden. To our understanding this was basically a clarification of concepts that were already in the law, although some new things were introduced — ratings limits on counterparties, for example. We have our own counterparty criteria that we follow and that is the backbone of our rating approach. It’s interesting to see that the regulators are moving in a similar direction by putting limits on the ratings of counterparties.

The other thing in Sweden was the introduction of the stress tests on valuations. Again, it’s something that I believe is considered good practice in many jurisdictions anyway, so I wouldn’t expect it to have any kind of impact on ratings or on performance for that matter. It remains to be seen just what impact it will have because it’s not clear what will happen if issuers fail the tests or if it can be failed at all. We would expect that to be clarified.

That very much represents what we are seeing from the legislators and regulators in all of Scandinavia, not a change in legislation, but more of an adjustment and clarification of the existing law. And that can happen in different ways. In Denmark, it’s been through political signals that banks are then interpreting in their own ways, changing the way they originate, for example. In Sweden we’ve seen more written clarification of the law and we have also seen the same in Norway in relation to summer houses, where it was clarified that summer houses should be interpreted as a certain kind of property type for the purpose of LTVs. That’s very much what we are seeing in Finland as well. We’ve seen the regulators there look at asset encumbrance again. And I think asset encumbrance has been the issue that we have all been interested in understanding better.

Norway and Finland are in Category 2 under your covered bond methodology but Denmark and Sweden are in Category 1. How might Norway and Finland get into Category 1? For example, Sweden’s early move from 2 to 1 was aided by changes to legislation allowing better access to central bank funding, something that is being considered in Norway.

The S&P ALMM criteria is quite clear on what is required to be a Category 1 country, and of course funding and funding options are an important part of this categorisation. However, they are not the only part. If the Norwegian issuing institutions gain access to central bank funding, it will be another funding option that is open to them, and it is clear that we would see that as a very positive move. That said, you could also question whether they do not already have quasi-access to funding through the parent banks, but that’s another issue.

However, in our criteria, it’s quite clear that we also look at the history of the market. How long has the market been in place? How long has it been stable? Well, we consider Norway and Finland to be relatively new jurisdictions in the covered bond world.

The other thing is the size of the market: is it big enough to be strategically and systemically important for a government? You could say right now with the Norway oil income — and the general good finances — Norway would probably have limited problems supporting a covered bond market of the current size. However, we also consider the systemic importance of the covered bond market; is it a market that due to its importance would receive extraordinary support from the government?

Covered bond markets need to have a combination of four features for us to view them as systemically important. Norway is close, but Finland does not meet the first feature of the criteria that outstanding covered bonds in a jurisdiction total more than Eu100bn.

The relative size of the covered bond market as a percentage of bank funding in a country should in our view be greater than 20% over many years. “Many years” is of course an unknown, but, again, it is only recently that the covered bond markets in Finland and Norway have come about. In Norway it is clearly growing in importance as a funding tool, but Finland is still in our view a new market in terms of importance and they have a lot of other ways of funding their mortgage loans.

Then there is the need for covered bonds to be integral to the provision of housing finance as evidenced by mortgage backed covered bonds as a percentage of GDP being greater than 20% over many years. Again, it’s something that’s fulfilled for Sweden and Denmark, but in Finland it’s not, and Norway is just about to get there, but yet has to show it over years.

The final factor is that covered bond finance has been a long term feature of the funding market, typically for 50 years or more, and has operated with no history of covered bond defaults. That’s definitely no in both cases. Norway had had a system of mortgage bonds, not covered bonds, and they had some serious issues in the 90s when the Norwegian government had to bail-out more or less the whole banking sector.

We are comfortable with the current categorisations of Norway and Finland as Category 2. It’s not that the covered bond markets are unimportant markets for these countries, but they are not, in our view, comparable to the markets in Sweden and Denmark, where covered bonds are literally a substitute for government bonds.

Is there anything else you would highlight about Denmark and Sweden?

The point about the covered bonds acting as a clear substitute for government bonds — with a small margin pick-up, of course — is an important factor that is maybe not clear from the four points where we talk about history and size. It’s also that you don’t suddenly have investors disappearing overnight. In Sweden and Denmark there is a huge domestic investor base, which is a big strength. Pension funds will have to pay out pensions in Danish kroner and they need to invest their money in something paying Danish kroner — unless they want to get into a lot of swaps, which we don’t see them at the moment being interested in. The same is true in Sweden. This big internal demand for covered bonds in these currencies is not established in the same way in Norway and Finland. And in Finland, with the euro in place, will there ever be such a relationship? A pension fund in Finland can of course look at any euro currency country and find investments without exchange rate risk.

Do strong domestic markets play into covered bond ratings in any other ways?

Those four points are actually a clarification of what makes up the systemic importance of covered bonds. But we do look at funding as an input in this country categorisation, again, what kind of options issuers have to fund, to create liquidity. And if the Norwegians create access to the central bank, that would be considered a strength in that regard.

One of the reasons we are quite comfortable with, say, the ARMs (adjustable rate mortgages) refinancings in Denmark is because we see the same investors coming back and buying the same bonds again and again. A lot of it has to do with the currency — if you want to buy something relatively secure and highly rated in Danish kroner, government bonds are not really an option at the moment given the low debt issuance of the country. Pension funds are also involved in covered bonds because of the pick-up, of course. And we do consider that support as a strength for the funding of these issuers.

You can say the same of the domestic Swedish market, where large amounts of covered bonds are rolled on a very regular basis, and investors, basically the same investors, take up the bonds. However, in our opinion, the Swedish system does allow for a degree of potential interest rate and currency risk.

The Danish system has come in for some criticism. Do you think these are justified and are the Danes taking appropriate measures?

It’s clear what the general covered bond market perceives as a risk in Denmark, and it’s also clear what the issuers and partly the local investors perceive as a risk in Denmark — and they are not necessarily the same things.

Coming back to the discussion about asset encumbrance: in some jurisdictions that might be an issue; in others it might not. You can say the same about refinancing — and again we have to be careful what we understand by refinancing risk. The term might need to be defined in a better way.

If you have a refinancing or resetting of interest rates once a year or four times a year, with the same amount being reset, we don’t really see a big difference. Only in terms of the new LCR requirements will that have an impact: under the latest changes the next 30 days have to be covered, so the Danish issuers will not have one big chunk, but four smaller chunks to cover.

But we should decide whether that resetting of interest is a risk in the first place. We’ve stated in our arguments regarding the risks around the ARMs system that it is a system that, in our opinion, is clearly regulated.

You could say that if you issue a one year bond against a 30 year mortgage, there must be refinancing risk. And if a bank needs to fund a 30 year mortgage they may one year later pay a higher interest rate. But when a borrower signs a 30 year mortgage agreement, the borrower agrees to take whatever interest rate the new refinancing is giving him or her, which means that the bank could as such be more or less neutral on what the new interest rate would be because it is simply passed on to the borrower (they have an interest in the potential linked credit risk, of course). So what we are saying is, we are going to stress the borrower for this risk, meaning that the borrower has a higher potential to default if he has a loan where there is a potential for a sudden increase in the interest rate.

Regarding the actual feat of going to the market and rolling these bonds, we have also been quite clear about this: our approach does not say that there may be no market access at all. Even in the crisis we saw that the issuers had access to the market. So our approach does not limit that; it simply stresses it. Finally, there is in the supportive system other ways of getting around potential shortcomings in market access, for example there is substantial repo access for the issuers with the central bank should the need arise.

We follow the behaviour of investors and monitor whether there is a sudden shift in the nationality of investors, for example if Danish investors are losing interest in the product. But again, we haven’t seen anything of the like. The latest refinancing achieved the lowest rates on record. Of course interest rates will fluctuate, they will go up at some point. But we are seeing the core investors being very faithful to the product and the system.

Regarding regulatory and legislative initiatives, the Danish government has said that a combination of ARM loans and IO (interest only) loans are adding perhaps too much risk to what they would like to see — that’s what I meant earlier by signals coming from politicians in Denmark. And issuers have responded in different ways. We believe it is fair to say that there has been a repricing of risk, so if you want an IO loan and an ARM loan together now it would cost you quite a bit more than before and some issuers won’t even offer it to you if the loan is in a certain LTV bracket. So there has been action in those respects. I saw a headline that for the first time the one year ARM loan is not the most popular in the Danish system this year — it reflects an interesting development.

So some things have changed. But it’s an old system and our view is they are not going to let external pressures change the system overnight. However, they might make adjustments as they go.

Do you have any concerns about house prices? Are there any other macroeconomic developments to watch out for?

Generally for continental Europe — with the possible exception of Germany — Europe’s recession is still dragging down house prices in most markets, and we wouldn’t expect any different performance in the Scandinavian countries, except Norway, perhaps. If you look at the development of house prices over time, it’s clear that the countries are in different positions in the cycle: the Danes have experienced a fall in house prices; the Swedes are probably seeing the beginning of a correction; and even in Norway in the last quarter there was a small fall in prices, although this could just be a blip.

It’s clear from all the signals we are getting from the central banks and FSAs — in particular Sweden and Norway — that there is a degree of discomfort with the current level of house prices. That’s not necessarily because they are unsustainable; it’s more because they are aware that sudden shocks to the economy — such as changes to interest rates, unemployment or, in the case of Norway, oil prices — could have a substantial effect on house prices, and that could perhaps snowball the economy into a deeper recession than might otherwise occur. Not forgetting Finland, because they have had substantial increases in prices, although they are coming from a very low price levels.

We continue to follow house price developments in all the countries and we will make adjustments to our assumptions if we consider them necessary. In Denmark, for example, we haircut some valuations made in the period 2005-2008 by 10%.

Besides the oil price in Norway — which doesn’t seem to be an issue at the moment — we believe that the main thing to watch out for is unemployment, because it has been relatively stable in Denmark, Sweden and Norway. Should unemployment increase considerably, that could be the biggest driver of any potential problems.

Looking at the interest rate effect, we believe that customers are relatively unaffected by increases in interest rates because they are simply so low at the moment, and people would not have a problem paying higher interest rates. But they will have a problem if they don’t have a job and then need to pay higher interest rates, so of course that combination would be the worst case scenario. Finally, in Denmark in particular, but also Sweden, certain benefits have been cut back, so whereas in the past we would consider that as an added layer of security, it is now less certain that borrowers in distress can rely on the support of the state.