Ratings

It is true that some mortgage originators have withdrawn from the market and others have been forced to scale back their activities. This is the result of the inability of these market participants to either fund their activities or do so at a cost that allows mortgage lending to be profitable, at the present time. It is also true that there has been no primary issuance of Australian RMBS into public markets so far in 2008. Those mortgage originators that substantially relied on this market to fund their activities have been adversely affected.

However, it cannot be said that consumers looking for mortgage finance have been similarly affected. There is no evidence that prime mortgages are not readily available from other lenders, although it may be that potential mortgagors who are less than prime could be finding it more difficult to obtain a mortgage. Nonetheless, while growth has moderated, home lending is still increasing, although the median house price across Australia has plateaued as rising interest rates take the heat out of the market.

The advent of mortgage originators in the Australian market has brought considerable benefits through the competition that has been engendered. There has been considerable innovation that has introduced redraw facilities, offset accounts, fixed-interest, and interest-only borrowings. The mortgage originators have also facilitated easier access to mortgage finance, with increased loan-to-valuation ratios being acceptable, along with the development of non-conforming low-doc, no-doc, and even reverse mortgages. And of course, the credit margins applied to mortgages have reduced considerably, so the cost of borrowing has reduced.

Not surprisingly, this has largely coincided with the boom in house prices. Data from the Reserve Bank of Australia (RBA) shows that in the decade to the end of 2007 housing finance outstandings increased by some 325%, or a compound annual growth rate of more than 15%. At the same time, according to Australian Bureau of Statistics data the median price of an established house across eight capital cities in Australia increased at a compound annual growth rate of almost 10%.

If contraction among mortgage originators now leads to a slowdown in the housing market, and vice versa (as these effects tend to be cyclical), it will not necessarily be an adverse development. The RBA has been steadily raising interest rates over the past year or so to fight inflation, and this in part has also been aimed at slowing booming house prices. Indeed, rising house prices and increasing interest rates have sufficiently eroded housing affordability to the extent that growth in housing finance has been in decline since peaking around 2004.

Recent purchases of RMBS on a repo basis by the RBA for longer-than-usual periods and in sizeable volumes may give the impression that the RBA is now becoming concerned about the state of the RMBS market and the absence of issuance in 2008. However, it is understood that the RBA's actions are aimed squarely at providing liquidity to those financial institutions that need it and are not intended to refinance RMBS holdings in order to restart the market. The Bank of England has implemented similar measures in the U.K. for banks and building societies that are eligible for the Banks' Standing Facilities.

The RBA is taking steps to address short-term liquidity problems that are arising from the dislocation that is currently being experienced in debt capital markets globally and locally, as the excesses that developed in the long period of accommodative credit conditions are cleared. To further illustrate the point, the debt capital market in Australia has not been overly receptive of any debt issuance since the start of this year.

The Australian banks have issued A$15.0 billion of bonds for the year to date (at the time of publishing) and foreign banks operating in Australia have issued A$2.5 billion. The banks have paid dearly for this funding, with five-year bonds recently issued by ANZ Bank yielding 128bps over bank bills/swap. A year ago such funding could have been raised at less than 15bps over.

The supranational and agency 'kangaroo' issuers have been the only other group to issue bonds in the domestic market this year. This group has not had to pay more for the privilege; in fact relative to the swap rate (and this is the key indicator for their cost of funds swapped back into their operating currencies) their credit margin has narrowed to around 15-20bps below swap, from 10-15bps below, previously.

This speaks volumes about the anxiety of investors and the extent of their flight to quality, when domestic banks are shunned and 'AAA' rated bonds from foreign government-backed issuers are highly sought after.

Nevertheless, it is Standard & Poor's view that there is not a crisis in our residential mortgage market but there is a temporary dislocation in credit markets. We note that Reserve Bank governor Glenn Stevens recently conceded to the House of Representatives economics committee that the credit crisis had damaged competition from non-bank mortgage lenders but said it would only be temporary.

While the dislocation in credit markets is expected to be temporary, it has exposed the structural imbalances that exist in the provision of housing finance and even the housing market in Australia. Arguably the greatest imbalance is in the almost exclusive reliance of some mortgage originators on securitization and the issuance of RMBS to fund their businesses. It is this reliance that is the key reason for most of the departures of mortgage originators from the market to date.

A lack of diversified funding sources is a serious flaw in any business model and while this was recognized before the credit crunch, competitive pressures on mortgage originators reduced their flexibility to seek a more diversified and robust business model; the alternatives were considered too expensive. Given recent events, and the likelihood that credit spreads will not return to early 2007 levels, it is expected that some non-bank originators will develop more diversified and robust funding platforms. This will likely necessitate a level of consolidation among some participants.

This structural flaw is also a concern for the smaller banks and other approved deposit-taking institutions (ADIs) that have relied on RMBS issuance to fund at least some of their mortgage portfolios. It is of less concern though, by virtue of the fact that these organizations have equity capital, deposit bases, and may even be able to issue bonds in their own right to fund their assets, of which mortgages comprise a substantial proportion. It is these other sources of funding that will be relied on while RMBS cannot be issued but it may be that mortgage lending by these institutions will still need to be constrained.

(A side issue is the competitive position of all ADIs in terms of their ability to attract deposits when increasingly sophisticated households have been diverting their savings away from deposits and into cash trusts, unit trusts, equities, and superannuation etc. However this has become less of an issue during this period of financial market turmoil, as the trend has been reversed and there has been a significant increase in deposits.

This suggests that there is a natural balancing mechanism in periods of market dislocation. As such, the move by households is consistent with the institutional investor flight to quality that has been observed over the same period, as discussed above.)

It may well be the stronger position of ADIs relative to other mortgage originators that ultimately forces the rationalization of the non-ADI mortgage originators and lead to the emergence of a better capitalized group.

Another structural imbalance highlighted is the heavy reliance on international investors and those pursuing a carry trade to buy Australian RMBS. In the period leading up to the credit crunch international investors accounted for more than 50% of the take-up of primary RMBS issuance, but prior to 2003 the balance was the other way around. The growth in the prominence of international investors coincides with the growth in the special-investment vehicles (SIVs) and other conduits.

This suggests that in recent years the supply of Australian RMBS was well in excess of the capacity of domestic investors to absorb, recognizing that domestic investors would also have included some arbitrage investors; this therefore points to reduced investor demand going forward. However, this structural imbalance may be more easily and quickly addressed by the market.

It is fair to say that both international and domestic real money investors were effectively crowded out of participating in primary issues of Australian RMBS during this period, as SIVs and other conduits drove the credit spreads down to below where the real money investors were prepared to participate. There is no reason now to believe that the real money investors will not return to the market once the current situation stabilizes.

Moreover, the supply of RMBS in the near future is unlikely to be as high as it was. While credit spreads were very narrow there was a significant volume of opportunistic issuance from the major banks, but at current levels the major banks are unlikely to continue this, relying instead on other funding options. Also, with housing credit growth slowing, the funding needs of the more traditional Australian RMBS issuers, the regional banks, and other smaller ADIs as well, as the non-ADI originators, will moderate.

Finally, there is the issue of any imbalance that may have been imposed on the Australian housing market, as it was in the U.S. In a world that was awash with cash, there was a race to buy any assets, at almost any yield, and a race among mortgage originators and packagers to meet that demand. No doubt this was a contributing factor in the emergence of new mortgage originators, the greatly increased availability of mortgage finance, the boom in house prices, and the significant increase in household indebtedness.

These excesses are now working their way out of the system. Fortunately, a much stronger economy, tight labour markets, increasing immigration, and a lack of new housing make a collapse in house prices the likes of that seen in the U.S. a less likely event in Australia.

If it is considered desirable to investigate the establishment of such an agency, for whatever reason, then consideration should be given to the models that have been implemented in other countries as well as the Canadian model. Without being exhaustive, the models adopted by others in the Asia-Pacific region have been reviewed by Standard & Poor's: specifically Japan, Hong Kong, Malaysia, Korea, and India. We have not looked at models that are based on providing subsidized housing or housing finance for disadvantaged groups in the community, as was the original mandate for the Swedish National Housing Finance Corp. (SBAB).

Each of the models reviewed is said to be based on Fannie Mae in the U.S., to varying degrees. The one that appears the least similar is in Malaysia: Cagamas Berhad does not buy mortgages outright but enters into repo deals with the originating banks for periods of three-to-seven years.

However, the rationale for the establishment of each of the entities is, by and large, the same among the pool, with there typically being two policy objectives on the part of the government concerned:

The promotion of home ownership; and

The development of domestic bond markets (or secondary markets for mortgages)

In relation to the first policy objective, each country had relatively low levels of home ownership at the time of establishing the government-backed mortgage-finance agency. Moreover, the banking system in each country was generally reluctant to engage in mortgage lending because of the very long-term nature of such lending and the interest rate risk that this would necessarily involve. The mortgage-lending model being pursued tended to be the American model, with 30-year-plus terms and fixed-rate lending.

As for the second policy objective, and again at the time of implementation, the domestic bond markets were under developed and limited almost exclusively to government bonds. The markets had not developed to the point of accepting credit risk and therefore only a government-guaranteed issuer would be acceptable.

The Canadian model not surprisingly is based on similar policy objectives to those in the pool, and operates in a similar fashion, although it does have some peculiarities that may be unique to the Canadian market. Canada Mortgage and Housing Corporation (CMHC) was established in 1945 by the Canadian government to provide affordable housing to servicemen returning from WWII, via direct lending. At this time the Canadian banks were reluctant to provide housing finance.

In 1954 CMHC moved away from direct lending, encouraging the banks to undertake this function, but facilitated bank lending by providing lenders' mortgage insurance (LMI) for conforming mortgages. Reflecting this, CMHC's policy objective evolved to be one of facilitating affordable housing finance for all Canadians. In this capacity CMHC has developed into the second largest Crown corporation (in terms of revenue), ranking only behind Canada Post.

Until recently CMHC had a monopoly on the LMI business in Canada but lately mortgage insurers such as Genworth Mortgage Insurance Corp., PMI Mortgage Insurance Co., and AIG have entered the Canadian market. Oddly though, CMHC guarantees 90% recovery of the principal sum insured (plus unpaid interest and costs incurred) should one of these mortgage insurers fail to meet a claim.

CMHC supports mortgage lending by the Canadian banks but has not particularly facilitated non-bank mortgage origination. Participation in CMHC programs requires mortgage originators to be federally or provincially regulated and have a minimum net worth and, in most cases now, a credit rating as specified by the CMHC. In fact, it is anticipated that the mortgage originators will be financial institutions.

To this end, mortgages guaranteed by CMHC and held on bank balance sheets attract a zero weighting for capital-adequacy purposes, making them quite attractive for the banks to hold, if they cannot be securitized. In fact, mortgages account for as much as 45%-50% of total assets of the banks, and the mortgage market in Canada is dominated by the six chartered banks, which account for around 85% of the market.

Not surprisingly, given CMHC's requirements for guaranteeing and securitizing mortgages, mortgage lending is focused on prime mortgage origination with little innovation in the mortgage product. In recent years a non-conforming mortgage market was beginning to emerge in Canada but has all but vanished with the advent of the credit crisis. It should also be noted that the Canadian banks have tightened underwriting criteria for mortgage lending during this period despite the existence of CMHC.

To understand why the banks have reacted this way, it is necessary to look at how the CMHC operates. While the CMHC will guarantee conforming mortgages, this does not help the banks finance the holding of those assets. In this respect the CMHC has had, and continues to have, limited capacity to assist.

Initially, CMHC provided this assistance by bundling conforming mortgages into MBS known as National Housing Act MBS (NHA MBS), also guaranteed by CMHC, and then selling them to investors.

Canadian investors, however, were not enthusiastic buyers of NHA MBS, disliking their non-bond like characteristics of unpredictable and uneven amortising cash flows. As a result, there was a limit on how much NHA MBS would be absorbed by investors, thereby restricting the ability of the banks to securitise their mortgages. Foreign investors were not buyers of NHA MBS due to the application of interest withholding tax (IWT).

In 2001, CMHC moved to address investor concerns about NHA MBS and introduced the Canada Mortgage Bond (CMB) – a five-year bullet-maturity bond with regular coupons. The bonds are issued by a special-purpose vehicle, Canada Housing Trust, and again are guaranteed by CMHC.

The bonds represent a repacking of NHA MBS, with swaps intended to address the cash flow mismatches between the NHA MBS, and a normal bullet-maturity bond being provided by the Canadian banks. However, the cash flow mismatches are not as significant as may be first thought.

While Canadian mortgages incorporate amortization schedules calculated over 30 years or more, the mortgages technically have a maturity of only five years and then need to be renegotiated with the lender. (While this model has worked well in the past, with almost all mortgages being refinanced at maturity, it must be highlighted that ultimately households bear the refinancing risk, which, depending on how the market develops, may become a more pronounced risk in future.) In addition, prepayment of mortgages is limited to no more than 20% of the original principal sum per annum, without penalty.

CMBs now account for virtually all MBS issued in Canada – there is little or no MBS issued outside of the CMB program. (It is instructive to note that the Bank for International Settlements (BIS) said in a December 2006 report ["The role of government-supported housing finance agencies in Asia", BIS Quarterly Review, December 2006] that, with the exception of Japan, MBS markets have failed to develop in any of the Asian markets we examined, despite the efforts of the respective agencies.) CMBs have been enthusiastically embraced by Canadian investors and are included in Canada's domestic Universe Bond Index and also in the Lehman Bros Global Aggregate Index.

International investors are also buying CMBs following recent changes to Canada's IWT regime. This could introduce greater investor-driven volatility in demand for CMBs, and in pricing, as has been seen in Australia but it also underlines the existence of continuing international investor demand for mortgage-backed debt instruments.

By the end of 2007, 30 CMB issues had been undertaken in Canada following a quarterly issuance schedule, and outstandings totaled approximately CA$119 billion, accounting for around 15% of the Canadian mortgage market. Issuance has been contained at this level to ensure CMBs continue to attract very fine pricing relative to government bonds.

Australia's housing market and its bond or fixed-income markets have developed differently from those of Japan, Hong Kong, Malaysia, Korea, India, and even Canada. Historically, home ownership has featured strongly in the Australian psyche, and Australian governments have not found it necessary to directly intervene in markets to support home ownership, or home owners in general, since mortgage-lending rates were briefly capped at 13.5% per annum in the late 1980s.

Australia has one of the highest rates of home ownership in the world, and the banks and other lenders have been keen to support this. Mortgages typically are provided on a variable-rate basis (thereby removing interest rate risk) for somewhat shorter terms of around 25 years.

With large and stable deposit bases, the major banks have been content to hold mortgages on their balance sheets using securitisation only sporadically, primarily to demonstrate that markets for RMBS product can be accessed and used to supplement their funding mix. There is usually limited cost of capital or funding cost advantages for the major banks in utilizing these markets.

The regional banks and other ADIs on the other hand have made greater use of securitization, and other mortgage originators have used it extensively. It is the development of securitization markets both domestically and internationally that has allowed significant competition to develop in the Australian mortgage market.

Perhaps the critical indicator of the level of competition that developed is not the degree of innovation that has been seen in mortgage products, nor the increased availability of mortgages to more borrowers, but simply in the erosion of the lending margin of the banks. In the mid 1990s the margin between the bank home-lending rate and the official cash rate was 3.0%. In early 2006 this margin had eroded to less than 1.2% on a discounted mortgage, which itself was 0.6% below the standard mortgage rate, according to RBA statistics.

The fact that this occurred again highlights the success of securitization in the Australian fixed-income market. Since around the mid-1990s Australia has developed a sophisticated fixed-income market that is quite accustomed to accepting credit risk via corporate bonds and pass-through RMBS structures and the higher yields they offer. Corporate bond sales in Australia since 1999 total more than A$320 billion, and almost A$300 billion of RMBS has been sold in domestic and international markets.

It is difficult to imagine why the domestic market would now revert back to being a government-backed bond market with modest yields, and duration management being the key skill of bond-portfolio managers. Moreover, it is the significant investment in developing risk-management and credit skills that has contributed to the development of a dynamic funds-management industry that is now looking to export its skills to the world.

Adoption of the Canadian CMB model or the mortgage finance model of one of the other countries considered above may result in a lower cost of funds for mortgage borrowers, but this would be achieved by the government guaranteeing the mortgage bonds issued. By using its 'AAA' rating the Australian government would be effectively subsidizing mortgagors.

Experience has shown that the subsidization of any group of home buyers (e.g. first-home buyers) or any reduction in the cost of buying a house (e.g. falling interest rates) inevitably results in rising house prices. This could be exacerbated by artificially boosting the supply of mortgage finance to the housing sector. Overall, this would seem to be an undesirable outcome.

Moreover, with a funding cost as low as has been suggested in the ASF proposal, it could become uneconomic for banks to retain mortgages on their balance sheets unless the mortgage finance agency established was similarly restrained, like the CMHC with its CMB issuance. While unconstrained issuance is unlikely, given that funding costs would increase as the volume of mortgage bond issuance increases, just where the balance of these counteracting forces would be is unknown. Thus any sizeable decline in the proportion of mortgage assets on bank balance sheets relative to other assets could have negative implications for overall asset quality and therefore bank credit ratings.

There may also be negative implications for participants in Australia's fixed-income markets and for the Commonwealth government. There is currently around A$150 billion of Australian RMBS outstanding. If this RMBS were progressively replaced by government guaranteed-mortgage bonds, the amount outstanding would be around three times more than current Commonwealth government bonds on issue, a similar multiple of the supranational and agency bonds outstanding in the domestic market and about 1.5 times state government bonds outstanding.

The latter two are close substitutes for Commonwealth government bonds and would likely see their credit spreads move much wider in the face of such issuance and consequently their cost of debt would increase and/or their domestic issuance would decline. In other words, there could be a crowding out effect.

The Indian National Housing Bank and Hong Kong Mortgage Corporation are often the largest non-government bond and MBS issuers in their respective markets. Moreover, the Canadian Treasury in its 2007 review of the borrowing framework of major federal government-backed entities observed that in the case of the Australian bond market, the Australian Government has been allocating budget surpluses to repay government bonds "which has led to greater liquidity for the development of the corporate market".

The cost of funds for the Commonwealth government could also increase if it were perceived that its credit quality had declined with the acceptance of the contingent liability that the guaranteed mortgage bonds would bring. This should be an anathema to even the present government after more than a decade that has been spent selling-off government-owned enterprises, developing a pristine balance sheet and eliminating net debt.

Standard & Poor's recently highlighted the potential risks to the U.S. Government's 'AAA' credit rating posed by the increasing contingent liabilities in the likes of Fannie Mae, Freddie Mac, and other U.S. government-sponsored entities. See "For the U.S. ‘AAA’ Rating, Government-Sponsored Enterprises Pose Greater Fiscal Risks Than Brokers", published April 14, 2008.

Implementing now a mortgage funding model that typically has the twin objectives of promoting home ownership and developing domestic bond markets would be introducing to Australia's well-developed mortgage and fixed-income markets a solution to a problem that doesn't exist. In fact, the BIS has highlighted Australia as one market where a government-supported housing agency has not been needed to develop well-functioning housing finance markets or liquid mortgage bond markets.

Moreover, mortgage financing is expected to remain very competitive even with the loss of some mortgage originators. Technology and changes to industry structure have allowed the more regionally-based small banks and other ADIs to become national competitors in the mortgage market. As stated above, we do not believe there is a problem in Australia's mortgage market except for possibly marginal borrowers.

Unfortunately there is no temporary or even long-term solution for the structural imbalances in the market that have been highlighted above. The only solution is to let the imbalances work their way out of the system. In this respect, any action taken by the RBA, the Commonwealth government, and even other regulators such as the Australian Prudential Regulation Authority will need to be carefully coordinated and applied to ensure an orderly re-balancing of the market rather than a bust.

As for the fixed-income market, there is a temporary dislocation but we remain of the view that the market will correct itself. Securitisation markets globally have seen minimal or no primary issuance so far this year – Australia is not alone in this respect. This is a direct result of the global credit crunch and the uncertainty that has been created around underlying asset quality. This situation will only improve when investors feel sufficiently confident to start buying again: price volatility will need to reduce and a greater consistency in the application of valuation methodologies would also be helpful.

This process is already underway, with excess RMBS supply in the secondary market clearing and talk of new primary issuance emerging. Credit spreads on primary corporate bond issuance have been contracting recently, and the Aussie iTraxx, generic five-year CDS index has retraced more than 50% of its widening since the credit crunch began.

It is also worth noting that two successful asset-backed security issues have been completed in recent weeks, one by St.George Bank and one by Bank of Queensland, albeit at credit margins much wider than before, but nevertheless finally breaking the drought of issuance in the securitisation market. Both transactions attracted interest from domestic and international investors

If a solution is required to the market dislocation it should be a temporary one. Implementing a permanent solution runs the risk of establishing a government-owned and -guaranteed entity that, for the most part, won't have a role to play.

It should be noted that in the case of Hong Kong and Sweden (as the role of SBAB evolved to be one of promoting a competitive mortgage finance market) the government-backed mortgage-finance agencies have been so successful in developing competitive mortgage-finance markets that they are now losing market share and are looking for other business opportunities. It is likely that it will prove more difficult to scale back the involvement of these entities in their domestic markets than it was to introduce them.

Any temporary solution would ideally be one that can be implemented again in the future should the problem reoccur. In this respect, the AussieMac proposal may have merit, if its purpose is solely to act as a provider of liquidity in times of market dislocation. But who will it do this for? Clearly, it would only be for those mortgage originators that are not ADIs.

As noted above, the RBA is already providing RMBS repo facilities to ADIs for periods of up to almost a year. The Bank of England is extending similar repo facilities to its constituents with the additional flexibility of being able to rollover the facility for up to three years. This additional flexibility could be adopted by the RBA in need.

Is it necessary to have another government-backed entity undertaking the same task but only for non-ADIs? The U.S. Federal Reserve has possibly provided a precedent with the rescue package that it recently extended to Wall Street. In times of unprecedented market dislocation, liquidity support can be provided to all critical market participants.

If needed the RBA could consider similarly extending its RMBS repo facilities to non-ADIs to assist in returning liquidity to the RMBS market. This could of course, lead to a debate over which non-ADIs are critical market participants but on the other hand, it could provide the new Commonwealth government with the opportunity to bring this part of the market under its regulatory umbrella – a stated objective.