Financial Stability Review – September 2013 Box C: Lenders Mortgage Insurance
Mortgage insurance is a specialist type of insurance that protects the mortgage lender in the event that a borrower cannot repay their loan. In Australia, mortgage insurance is offered by prudentially regulated institutions known as lenders mortgage insurers (LMIs). These institutions charge an upfront premium to lenders, which usually pass on the cost to borrowers. Lenders generally use mortgage insurance for loans originated with a loan-to-valuation ratio (LVR) of 80 per cent or greater, given the higher risk profile of these loans. LMIs also provide credit enhancement for mortgage-backed securities – either at the individual mortgage level or on the underlying mortgage pool – to reduce the likelihood of losses for investors. Overall, more than one-quarter of Australian housing loans are estimated to be covered by mortgage insurance.
LMIs have liabilities that are concentrated in highly correlated risks. This exposes them to significant insurance risk as they can experience a heightened number of policy claims during economic downturns. This is different from other general insurers: many of their policyholders are insured against losses from relatively unrelated physical events (e.g. accident or theft), with multiple policyholders affected by the same event only in infrequent cases (e.g. natural disasters). The Australian Prudential Regulation Authority (APRA) therefore requires providers of mortgage insurance to be ‘monoline’ insurers (i.e. they can write only one type of insurance) in order to ring-fence mortgage insurance from other insurance activities.
LMIs and Financial Stability
LMIs can influence financial stability given their involvement in the credit creation process and linkages with the banking system. LMIs can help promote financial stability to the extent that they dampen swings in lending standards and maintain sufficient capital to withstand any housing market and economic downturn. As insuring riskier mortgages is the primary business of LMIs, they may take a longer-term view of mortgage risk than some (marginal) lenders. During buoyant times when risk appetite among lenders rises, LMIs could limit the extent that lending standards weaken because they provide a ‘second set of eyes’ in the loan origination process. Conversely, when risk appetite subsides during downturns, a well-capitalised LMI industry could increase at least some lenders' willingness to continue writing high-LVR loans, helping to smooth changes in lending standards.
The use of mortgage insurance will not necessarily moderate the amplitude of the housing credit cycle, however. Lenders may respond by relaxing standards because they believe the LMI is assessing the risk – an unintended consequence of having a ‘second set of eyes’ – or because they believe that any loss is an LMI loss. In theory, this type of behaviour would be more likely to occur in situations where LMIs fully insure lenders' losses, because lenders then have very little ‘skin in the game’. The use of LMI could also lead to adverse selection problems, whereby lenders, having superior information on borrowers' repayment capacity, only insure loans that are higher risk than they appear, and thus expose LMIs to greater risk than they realise. However, industry practices have developed to substantially mitigate these problems, including pre-approval standards, monitoring processes and claims management practices. More generally, as the mortgage insurance market is cyclical – like most financial activities – the risk appetite of LMIs will not necessarily be less procyclical than lenders in all circumstances. For instance, US bond insurers started insuring riskier financial products in the years leading up to the global financial crisis, contributing to procyclicality in the financial system.
The correlated nature of mortgage insurance policies means that a severe downturn in the housing market and the economy would likely result in substantial claims on LMIs, potentially weakening their creditworthiness. Distress in the LMI sector could hinder the payment of claims to lenders, although the direct financial impact of this on the Australian banking system is unlikely to be substantial.
There could be more significant indirect effects from distress in the LMI sector. Without the ability to transfer credit risk to LMIs, some lenders may be reluctant to offer new mortgages with high LVRs, for example to first home buyers. Any such pull-back from providing housing credit would in turn affect the broader economy and confidence in the financial system. Confidence effects stemming from LMI distress could be exacerbated by the likelihood that it would occur in the context of a weak economy, when lenders would presumably be incurring significant losses in their non-housing loan portfolios. A further relevant factor is the structure of the LMI industry in a given market, including, for example, the degree of substitutability and barriers to entry for new players.
APRA's prudential settings take into account that an LMI's business is concentrated in correlated risks and is closely linked to the banking sector. To ensure they are resilient to the key tail risk they face (a very severe housing market downturn), Australian LMIs hold a substantial amount of capital against ‘insurance concentration risk’ (a component of their total capital requirement). Furthermore, APRA's stress-testing of Australian banks also considers their interconnections with LMIs. More broadly, like all general insurers, LMIs are subject to intensive supervision, including ongoing monitoring of risks and financial condition, scenario analysis and detailed on-site supervisory reviews, followed by supervisory responses where appropriate.
International Comparison of Mortgage Insurance
Mortgage insurance is available in many jurisdictions but extensively used in only a small number, including Australia, Canada, Hong Kong, the Netherlands and the United States. The structure of the mortgage insurance industry across these and other countries varies considerably and is affected by the domestic regulatory landscape and the extent of government participation in each jurisdiction (Table C1).
Regulatory arrangements can support the use of LMI. In a number of countries, insured mortgages have lower capital requirements, creating an incentive for banks to use mortgage insurance. In other cases, mortgage insurance is mandatory: for example, high-LVR mortgages originated by regulated deposit-taking institutions in Canada and Hong Kong, as well as those purchased by the government-sponsored enterprises in the United States, must be insured. In Australia, mortgage insurance is not mandatory, but APRA's prudential framework includes lower capital requirements for insured higher-risk mortgages of (smaller) deposit-taking institutions operating on the standardised approach to capital adequacy. Even though the (larger) deposit-taking institutions operating on the advanced approach to capital adequacy have quite limited capital incentives to do so, they still use insurance extensively for high-LVR mortgages, given the credit risk transfer and other benefits of LMI.
In those countries where mortgage insurance is used extensively, governments often participate in the mortgage insurance market in one form or another. The structure of the Australian LMI industry differs, however, in that the mortgage insurers are all privately owned and operate without government guarantees. In Canada, the LMI industry consists of one large government-owned mortgage insurer and a number of smaller private LMIs, whose liabilities are largely guaranteed by the government. In the United States, the government owns one mortgage insurer which provides ‘socially targeted’ mortgage insurance, while mortgage insurance in the Netherlands is provided by a government guarantee to the lender.
Government financial support of the mortgage insurance industry can support social policy goals, for example by subsidising the provision of affordable housing credit for low-income households. These benefits must be balanced against the potential costs, including the cost to the taxpayer if the mortgage risk transferred from the financial sector's balance sheet results in significant losses. In addition, if mortgage insurance is subsidised and therefore under-priced, it could distort lending towards housing credit, particularly higher-risk mortgages.
For further information on the interaction between lenders and mortgage insurers, see Joint Forum (2013), ‘Mortgage Insurance: Market Structure, Underwriting Cycle and Policy Implications’, Bank for International Settlements, August. 
The explicit regulatory incentive for Australian banks to use LMI has, to a significant extent, been reduced for banks approved to use internal models because APRA requires a minimum 20 per cent loss given default assumption in these models irrespective of LMI. This floor was imposed as a substitute for the limited downturn experience in Australia over the past few decades. 
The Australian Government exited the mortgage insurance market in the 1990s when it restructured and sold the Housing Loans Insurance Corporation. The corporation was established in the 1960s to help low-income earners to obtain housing finance by insuring lenders against the costs of mortgage defaults, and sought to fill a ‘market gap’ that existed at the time. The government's exit in the 1990s was justified on the grounds that, among others, the private sector had a demonstrated capacity to provide mortgage insurance and the government's continued involvement placed a financial burden on the public sector. For further information, see Housing Loans Insurance Corporation (Transfer of Pre-transfer Contracts) Bill 2006 and Housing Loans Insurance Corporation (Transfer of Assets and Abolition) Repeal Bill 2006, Explanatory Memorandum.