In my recent Christmas article I touched on a topic that provoked some of you to comment – the financing of higher LTV mortgages via securitisations. Thanks for the comments, which make me think that this discussion needs to be explored further.
My point was that many in the covered bond market argue that one of the reasons for its superiority over securitisations is the stricter limits on risky mortgages. This is valid in the case of many of the more, shall we say, exotic lending practices underpinning mortgages in some pre-crisis securitisations. It’s always amusing to recite the various acronyms and euphemisms from the heady days of that market – NINJA, meaning no-income, no-job borrowers, for example. But more or less pointless, the stricter rules in the general mortgage lending business and in the securitisation directive have put a stop to that sort of thing (for bank lenders in Europe at least).
But it is not always valid to say that the greater risk tolerance in securitisation cover pools is a bad thing, particularly in the case of their ability to fund higher loan to value mortgages. We covered bond people wear 80% LTV caps as a badge of pride, evidence of our more sober approach to housing finance, whilst securitisations recklessly head into the 90s and beyond.
Clearly from a point of view of the stability of the instrument we are correct. All other things being equal a bond backed by 80% LTV mortgages is safer than one backed by 90% loans. But that isn’t necessarily the point of view that we should be considering. We are so pre-occupied with the stability of the financial system that we are frequently blinded to its role. That is a much bigger comment on post-crisis financial regulation in general, but for our purposes I mean that it is valid that some people with smaller deposits should be able to buy houses.
High loan to value mortgages are correlated with greater labour mobility. A more diligent person than I would prove it empirically so I will just have to say it anecdotally; in those countries with higher LTV mortgages, people, particularly younger and less affluent people buy homes earlier and move to where the jobs are more readily.
Whether this is correlated to higher or (as I argued at Christmas) lower levels of homelessness in society, I can’t say (although I’d be curious to know).
A key point in this argument is that ‘some’ people should be able to lend with smaller deposits. And this is an area where the securitisation market is unexpectedly bad at blowing its own trumpet – its more sophisticated analysis of risk than ours. Debt service to income ratios are a vital risk metric that we more or less ignore in covered bond laws and regulations (yes I know, with some exceptions). This is, or rather should be, more of a pro-securitisation argument than it is currently. Long periods of very low interest rates have two affects: they inflate house prices – making loan-to-value ratios optically better but more vulnerable to house price crashes – and they increase the ability of borrowers to service debt and (in some countries) repay faster. Both of which factors argue for debt-service to be a better measure of credit than LTV.
A traditional covered bond riposte – that there is nothing stopping higher LTV lending, just not using it as the basis for stable bonds, leads to the inevitable happy compromise – that covered bonds should be about funding the safe bit, securitisations about risk managing the riskier bit.
A theme which I suspect I will return to.