The federal budget due March 29 almost certainly will unveil the government’s thinking on the future of the Canadian covered bond market. This may prove strangely important, because covered bonds lie at the heart of one of the most contentious public policy issues in the Canadian financial system.
The bonds are “covered” because they are backed, normally, by high quality, or low loan-to-value ratio, mortgage loans. Canadian banks originate mortgage loans, and sell them to a bankruptcy-remote special purpose vehicle, which in turn sells bonds, typically with five-year terms, to foreign and domestic investors. The stream of interest and principal payments on the mortgage loans backs these contractual covered bonds.
The cover pool is always overcollateralized, meaning that the face value of the mortgage assets exceeds the value of the bonds they back; Canadian practice limits overcollateralization to 10% of the bond issue. The issuing bank’s full faith and credit gives the bonds dual cover.
This belt-and-suspenders approach to security strengthens the confidence of bond buyers. For instance, when the Royal Bank of Canada first came to market in 2007, with a cover pool comprising uninsured residential mortgages, the bonds sold at a low yield spread over comparable duration sovereigns. And when in February this year Moody’s Investors Service announced a potential downgrade of the parent company’s credit rating, it later clarified that RBC’s covered bonds would retain, even in the event of a bank downgrade, their Aaa rating.
So far, so good — the covered bond structure ensures that banks are exposed to the risks of the loans they write, letting them sell the bonds they back at very low yields. This securitization model, with its short intermediation chain — in contrast to “originate-to-distribute” mortgages and the structured securitizations and derivatives built around them, which went so wrong in 2008 — distributes risks to informed investors who are able and willing to bear them, rather than centralizing risks within the financial system. That is a good outcome, which also shows how markets can evolve to deliver a sound housing finance system.
Yet, owing to government-backed mortgage insurance, the future might continue to look rather like the past, with rising taxpayer exposure to bank lending and borrowing.
Since 2007, Canadian banks have increasingly come to the covered bond market with bonds backed, in whole or in part, by mortgages individually insured by the Canada Mortgage and Housing Corporation. This insurance cover boosts the surety of the bond pool, and marginally lowers the banks’ cost of capital and, arguably, perhaps lowers the cost of homebuyers’ mortgages. But an otherwise functioning financial market also gains government and taxpayer participation, and risk exposure, to uncertain net benefit.
There is more. CMHC also offers, and many banks buy, portfolio or “bulk” insurance for bundles of otherwise uninsured mortgages, which in turn are placed in bond cover pools. This cheap insurance is an additional guarantee, and banks save typically 10 basis points or more on their insured bond issues.
The popularity of bulk cover is one reason why CMHC is now pushing toward its $600-billion legislated cap on the extent to which the agency can back debt with a government guarantee. And the belt-plus-suspenders-plus-insurance raises questions, such as the extent to which taxpayers should guarantee bank lending to foreign bond buyers who might otherwise be satisfied to bid for uninsured covered bonds. Bondbuyers seeking a sovereign guarantee have access to government of Canada treasuries and long bonds, which are a similar, competing investment.
In turn, that is why Ottawa will be considering limiting — perhaps even to zero — the extent to which covered bond pools may be stocked with portfolio-insured mortgage bundles.
Other components of a legislated, as opposed to contractual, covered bond framework may be warranted, such as standards for replenishing bond pools. Over the life of, say, a five-year covered bond, some of the mortgages that back it will go bad, and others will age out of the pool — they need to be replaced by mortgages of similar or higher quality.
And that is where things can go bad. An example from the 1890s makes the point, when one of several farm mortgage crises rolled across the U.S. Mortgage companies began to fail, and as their capital came under pressure, they restocked cover pools with bad mortgages, and removed good ones, violating their bond programs’ trust agreements, and spreading their financial illnesses to investors at home and abroad, who thought they had secure collateral.
No one would suggest such a scenario among domestic banks today, but the incident sufficiently resembles the 2007-08 collapse of mortgage securitizations that it deserves attention. A sound regulatory and legislative framework, including disclosure standards for asset replenishment, common practices and audit standards, can all help the Canadian covered bond market develop.
Even better, housing finance has a route that may help it evolve toward lesser reliance on taxpayer guarantees, and with confidence. And that would be a good thing.
Finn Poschmann is vice-president, research, at the C.D. Howe Institute.