FI’s Director General participated in the seminar Evolution of Mortgage Finance arranged by Stabelo for a broad group of institutional investors.
A well-functioning mortgage market is of great importance for both Swedish households and the economy. For a very long time, the structure of the Swedish mortgage market was quite stable, but the market is now changing in several ways. Today, I would like to talk about how FI views these changes as they relate to the provision of mortgage loans and the funding of mortgages. First, let's set the scene.
The residential mortgage market in Sweden currently amounts to around SEK 3,300 billion – corresponding to approximately 70 per cent of GDP.
So far, domestic banks have provided almost all mortgages to households in Sweden. There are several categories of banks: major banks, medium-sized banks, branches of foreign banks, independent savings banks and smaller, more niched banks. Traditionally, the Swedish mortgage market has been highly concentrated. If we count Nordea as a Swedish bank for the purposes of Swedish mortgages, the four major banks account for close to 75 per cent of the mortgage volume. However, the major Swedish banks have been gradually losing market share to smaller banks and foreign bank branches when it comes to mortgages.
Residential mortgages are a very important product for the Swedish banks. They form the basis of many customer relationships and enable banks to sell more products to the customer. Mortgages are far from being a loss-leader. The banks have been able to make good and, in fact, growing profits on mortgages. This is to a large extent due to the fact that Swedish banks are relatively cost-effective compared to their European peers. The banks have also been able to compensate low interest rates and extensive regulation by widening the lending margins. This has been possible due to high household demand for loans, partly driven by very low mortgage rates. The banks' profits also reflect traditionally weak competition on the mortgage market compared to many other financial products. But this may be changing.
In the traditional Swedish bank-based business model for mortgages, banks extend mortgage loans to customers and carry the loans on their balance sheets. These loans are to a large extent funded by the issuance of covered bonds, which in turn are secured by the bank's pool of mortgage loans and protected by special legislation.
In December 2018, the covered bond market amounted to close to SEK 2,180 billion, financing around 66 percent of the mortgage loan stock. Swedish covered bonds are generally AAA rated and are considered an attractive low-risk investment by fixed-income investors in Sweden and abroad, offering a yield pick-up to government bonds.
The banks issue covered bonds both in SEK and foreign currency. Foreign investors hold around 40 per cent of Swedish covered bonds. Among Swedish investors, who own the remaining 60 per cent, the two largest categories are insurance companies, which own 24 per cent of covered bonds, and banks, which own 15 per cent. Covered bonds issued by other banks play an important role in banks' liquidity reserves.
The funding costs for Swedish covered bonds have historically been low due to strong confidence in the banks and low credit risk premiums. All of the Swedish covered bond programs currently have the highest possible credit rating from credit rating institutions. An indication that investors assess the risk as low is the low interest rate differential (risk premium) with regard to government bonds.
Interestingly, covered bonds as a share of mortgages has declined for the major banks. In other words, banks have not increased their issuance of covered bonds at the same rate as mortgage lending has expanded. Why is this? The trend has been pointing downwards for quite some time. One reason appears to be that deposit growth has been strong over the past three years or so. Since covered bonds normally carry maturities of five years or more at issuance, there is a term premium to be paid for the issuer compared to deposits and short-term funding. Banks will obviously try to find an optimal financing mix, striking a balance between cost and risk. In the current ultra-low interest rate environment, ample supply of very cheap and typically stable deposits have allowed banks to hold back their term funding in the form of covered bond issuance despite the low funding cost.
The Swedish banks comply with the relevant liquidity risk regulations, but at the same time banks continue to have relatively short liabilities compared to the average in the EU. Since a high share of long-term lending (mostly mortgages with a typical contractual maturity of 30–50 years) is financed by relatively short-term liabilities, which need to be rolled-over frequently, the banks engage in maturity transformation in the mortgage market as well. This means that banks are exposed to refinancing risks and thus sensitive to disruptions during times of financial turmoil when risk aversion among depositors and investors goes up.
In order to protect the important functions that banks perform, the banks are subject to extensive capital and liquidity requirements and certain other arrangements, including deposit guarantees and resolution. FI has also introduced regulations specifically aimed at managing systemic risks that relate to mortgage credit expansion. These measures include increasing the banks' risk weights for mortgages to 25 per cent. Sweden has also activated the countercyclical capital buffer requirement and gradually increased this buffer to its current level of 2.5 per cent.
There is no denying that bank regulation has an influence on business models and profitability. Indeed: that is the intention. For example, in the absence of capital requirements, a bank could have leveraged its balance sheet further by making more loans and generating additional earnings. Alternatively, more capital could be paid out as dividends to banks' shareholders.
The effect of the capital requirements can be expressed as an economic cost to banks that amounts to 40-50 basis points if expressed in relation to the mortgage volumes. One way of thinking about this is to say that bank mortgage margins would be 40 or so basis points higher in the absence of these requirements. Alternatively, one might speculate that banks would have transferred this cost to their customers, in which case the customers would be charged loan rates that are 40 basis points higher due to the regulations. Or anything in between.
Summarising the recent developments in the traditional mortgage market, we note, firstly, that the market is becoming less concentrated as smaller banks and foreign bank branches are capturing market share from the four large banks. Secondly, the current trend is for bank mortgages to be financed by covered bonds to a decreasing degree. Thirdly, the mortgage activities incur economic costs as a result of specific policy decisions in the field of capital regulation.
A new trend is that non-bank newcomers are entering the mortgage market. This is partly a consequence of new regulation and technological development, which have lowered the barriers to entry and created more favourable conditions for market entrants.
While their business models differ, the new entrants share a common trait: they are not banks and they operate, with regard to both origination of loans and funding, outside the banking system. Unlike the banks, which hold mortgages on their balance sheets and raise funding through deposits or covered bonds, the credit risk in the market entrants' business models is transferred directly to end-investors.
Targeted investors are primarily pension funds and insurance companies, which already play an important role in financing Swedish mortgage loans by investing in covered bonds issued by the banks. Think back to the earlier slide, which showed Swedish insurance companies owning close to one-fourth of outstanding covered bonds. The market entrants are now offering direct investments in the form of new types of financial instruments backed by mortgages. The risk transfer can occur in several different ways, for example through a mortgage fund or securitisation.
There seems to be some demand from institutional investors for this. Some of the Swedish insurance companies and pension funds have now invested directly in mortgages. The amounts so far are small in absolute terms and miniscule in relation to their covered bond holdings, but today's conference indicates that there may well be more interest in these kind of investments in the not-too-distant future.
The arrival of the new players on the mortgage market has begun to shift some of the mortgage business, and, hence, some of the risk associated with mortgages, to entities outside the banking system. And we may see more of the same in the future. Is this a potentially dangerous journey into unchartered shadow banking territory? Is it part of a transformation of the bank-centred Swedish credit market into something more like the Anglo-Saxon models of financial and credit markets, where capital markets play a large role in the provision of finance to the real economy?
I believe that to objectively assess the implications of the new mortgage structures one has to start with whether the current regulation for new mortgage structures is sufficient from a consumer protection and financial stability perspective.
On the one hand, the legislator has not deemed it necessary to establish capital and liquidity regulations for the new market participants similar to that which applies to banks.
This may be in part due to the critical functions that the banks provide and the identity of those who ultimately provide the bank with funding. Banks fund mortgages, and other lending, by taking deposits or other so-called repayable funds from the public. This means that the provision of deposit and savings services is a critical function that must be safe-guarded and upheld. To avoid the risk that consumers and society as a whole will have to bear any negative consequences, prudential regulation manages the vulnerability in banks' business models. Since the new participants on the mortgage market do not finance mortgages through consumer deposits, they are not subject to the same rules.
At the same time, predictable and reliable access to mortgage financing is a fundamental service for both individual households and the economy as a whole. This credit function must also be safeguarded.
Regulation and supervision manages the refinancing risks for banks that arise due to maturity transformation. When mortgages are held outside the banking system, there are no corresponding rules and tools. Under stressed market conditions, refinancing problems for the new entrants could therefore lead to sharply deteriorating terms and conditions for borrowers or even termination of the loans. This would violate general consumer protection rules, which apply regardless of the creditor. Moreover, if a significant share of the mortgage market were to shift outside the banking system, the supply of credit for the entire economy could become less stable.
Consequently, consumers must be protected in two capacities: as both borrowers and depositors. And by doing so, we are able to enhance the stability of the economy as a whole.
In FI's opinion, a credit supply that flows partly outside the banking system, but through other regulated entities under supervision, can be a good thing. Still, it needs to be managed properly.
For borrowers, greater competition on the mortgage market can create higher customer value through lower loan rates, more alternatives and better services.
Given their role as large participants in the financial market, insurance undertakings and occupational pension institutions can contribute to financial stability through their investment behaviour. They have significantly longer investment horizons than what covered bonds currently offer. Therefore, it should be possible via the new business models to finance mortgages through investors with a longer horizon, corresponding more to the expected maturity of a mortgage loan.
By investing in new mortgage structures, the insurance and pensions sector could take over parts of the credit risks for mortgages that are currently borne almost entirely by the banks. And reduce liquidity risks in the system as whole, since insurance and pension firms are less subject to withdrawal and funding risks than traditional banks. This would lead to better overall risk distribution (more loss-absorption capacity thanks to different entities) and decreased structural imbalance (less maturity transformation risk due to better matching between asset and liability maturities), both of which increase the resilience of the financial system as a whole.
However, when mortgage financing takes place outside the banks' balance sheets, regulation and supervision need to adapt to ensure that market participants behave in a safe and sound manner and consumer interests are protected.
FI has therefore developed a comprehensive supervisory framework for all market participants who provide mortgages as well as those who invest in mortgages. The guiding principles are the same as those for banks with a traditional financing model for mortgages.
Firstly, investors need sufficient liquidity to be able to accept that their investments in the new mortgage structures cannot be viewed as a source of contingent liquidity. The underlying credit risk exposure may be the same as in covered bonds, but the liquidity is not as good in a mortgage fund investment. This puts constraints on the business model and asset-liability structure of investors.
Secondly, actors with new mortgage structures need to take the borrowers' long-term needs of uninterrupted credit provision into consideration. A decision to repay the loan early should be the borrower's right and never an obligation.
FI believes that the maturity for the financing of mortgages in new mortgage structures should be at least ten years and that funding should show a well-spread maturity distribution over time. The investments should essentially be considered "buy-and-hold". Investors may not be given a contractual right to redeem their investments early.
It is worth noting that there are other rules in place aiming to safeguard the interests of individual borrowers, thereby preventing borrowers from being pressured into early repayment of the mortgage. In particular, the Consumer Credit Act lays down conditions that limit a lender's possibilities for unilaterally raising the interest rate in conjunction with interest rate adjustments. This applies regardless of the creditor and funding model.
Thirdly, investors need to have sufficient capital to ensure that they will be able to absorb any losses arising from the mortgage investments. While the emerging mortgage market in Sweden is currently focusing on very low credit risk, investors must nevertheless have an adequate level of capital in order to withstand any future losses.
Finally, investors must have a high awareness of risk in order to accurately assess risks in the investments, including the credit risk associated with the mortgages. Again, the current credit risks appear very low, but in the unlikely event of losses arising, the investor may be under no illusion: the risk is for the investor's own account, not the fund's, the manager's or the originator's. This is important to emphasise since, unlike banks, insurance companies tend not to have a strong tradition of assessing credit risks in mortgages.
Banks themselves are beginning to show an interest in establishing new mortgage structures alongside their traditional on-balance sheet mortgage business. The combination of increased competition on the mortgage market and high economic costs resulting from the current banking regulations have created incentives for banks to consider adding mortgage structures that are similar to those of the new market participants. For these structures, some special issues and risks will arise that need to be considered by both individual banks and FI.
One aspect is whether these new structures should be consolidated at the bank group level, which has an impact on capital and other regulatory requirements. Another aspect is the potential flowback risks associated with mortgages the bank has originated and then transferred to the new mortgage structure. Flowback risk refers to the risk that the bank will feel obliged to bring the loan back on to its own balance sheet during periods of stress in order to help investors out, prevent credit losses and protect their reputation among investors in the market, even though it has no obligation to do so.
Whenever new players or products enter a market, there is a temptation in the media and in the general debate to either hail or condemn the new entrants. It is not FI's mission to protect or promote either existing or emerging business models; FI's focus is on whether they are sound. We support financial innovation if business models can be designed and regulated so as to be beneficial for both customers and financial stability.
In FI's view, the emerging new mortgage structures can be fundamentally positive if they are properly designed. This applies regardless of whether a new mortgage structure is owned by a new market participant or a bank.
The new structures can contribute to greater customer value through more choice and increased competition, increasing pressure on loan rates. In addition to improvements for individual borrowers, there are important potential benefits for financial stability if the overall demand for credit in the country can be met by a broader, more stable and more diversified base of capital and funding sources – instead of just by business models that keep all risks within the banking system.
The new structures are still in their infancy and different scenarios are possible. FI will closely follow the developments and monitor the market's scope and risk level carefully.
Thank you for your interest.