The Triumph of the Past?
John Dizard wrote a provocative article for the FT recently. He discussed what amount to little more than a rumour, albeit possibly a well informed one, about the views of leading regulators and central bankers concerning the future of securitisation.
This is worth examining in some detail. First, how could central bankers compel or at least strongly suggest to banks that they desist from securitisation? Presumably without any changes at some point the securitisation market will come back - indeed some MBS deals have been done already this year. So to rein in what they evidently see as the Gorgon, the central bankers will have to do something. The only available policy lever is regulation: capital charges will have to go up markedly to throttle off demand.
Note too that there are different reasons for banks to use the securitisation market, some of them less toxic than others. At the benign end of the spectrum selling the AAA to get sub-Libor funding on an opportunistic basis (and obviously not assuming that this can be done as part of your funding strategy) is natural for smaller banks. Without it, the barriers to entry get even higher and the benefits of consolidation more significant: do we really want more banks which are too big to fail?
On the other hand, clearly supervisors have a vested interest in preventing regulatory capital arbitrage via securitisation. Basel 2 does not do this: it simply introduces new arbitrages from the ones available under Basel 1. There are fewer of them perhaps but they are still there.
The final motivation for securitisation is pure arbitrage. Sometimes the assets really are worth more in securitised form. This may be because they are illiquid, hard to originate, or hard to assess. Institutions with an edge, be it origination or market knowledge/credibility, can legitimately exploit that. Remember too that securitisations are like blenders: they can liquidify previously illiquid assets. That in itself can make them more valuable.
Could one hope that the supervisors will continue to permit the good securitisation yet design rules which make the bad more difficult if not impossible? It seems unlikely. If there is any wiggle room at all in the rules the industry will find it. So to be sure of banning the bad kind of securitisation it is likely that the rules will end up disincentivising the good kind too. If Dizard is right, then, Basel 3 will tend to favour large banks with a low cost of funds and plenty of capital.
The next part of the problem is that there are not too many of those left. So most banks will be forced to scale back on risk taking while they rebuild capital and delever. The consequences of that for the G10 economies will be profoundly negative: credit in all forms will become much harder to obtain, market making will be less practiced and hence liquidity will decrease, and growth will slow. Banks are too addicted to securitisation to go cold turkey without the detox effecting the broader economy. Instead I suggest central bankers should be looking at something more like nicotine patches: get rid of the most harmful form and manage dependence without generating too much discomfort. An over-reaction, however, seems most likely.
The official [central banking] world, and those close to it, are anticipating that we're going back to an on-balance-sheet financial industry. That is, the extension of credit will be done, to a much greater degree, through direct lending by depository institutions rather than through the securitisation of structured products. The frenetic expansion of securitised and, it was supposed wisely distributed, risk turned out to be not quite so wise after all.
The problem with putting credit on bank balance sheets is that those balance sheets aren't big enough to cover the losses from past practices and to continue to expand credit at an adequate pace. Shareholders' equity and reserves aren't there, at least in the necessary size.
Very true, the central bankers will say. So we'll just have to get some new shareholders. The "real money investors" are there, and not just the sovereign wealth funds. What about the present shareholders, I ask, my face turning white? As Colbert memorably said: "A banker is a soldier in the service of the state." So perhaps the rally in bank shares might be a little premature. The central banking world is expecting a serious shake-out of individual and perhaps institutional participants.
This is worth examining in some detail. First, how could central bankers compel or at least strongly suggest to banks that they desist from securitisation? Presumably without any changes at some point the securitisation market will come back - indeed some MBS deals have been done already this year. So to rein in what they evidently see as the Gorgon, the central bankers will have to do something. The only available policy lever is regulation: capital charges will have to go up markedly to throttle off demand.
Note too that there are different reasons for banks to use the securitisation market, some of them less toxic than others. At the benign end of the spectrum selling the AAA to get sub-Libor funding on an opportunistic basis (and obviously not assuming that this can be done as part of your funding strategy) is natural for smaller banks. Without it, the barriers to entry get even higher and the benefits of consolidation more significant: do we really want more banks which are too big to fail?
On the other hand, clearly supervisors have a vested interest in preventing regulatory capital arbitrage via securitisation. Basel 2 does not do this: it simply introduces new arbitrages from the ones available under Basel 1. There are fewer of them perhaps but they are still there.
The final motivation for securitisation is pure arbitrage. Sometimes the assets really are worth more in securitised form. This may be because they are illiquid, hard to originate, or hard to assess. Institutions with an edge, be it origination or market knowledge/credibility, can legitimately exploit that. Remember too that securitisations are like blenders: they can liquidify previously illiquid assets. That in itself can make them more valuable.
Could one hope that the supervisors will continue to permit the good securitisation yet design rules which make the bad more difficult if not impossible? It seems unlikely. If there is any wiggle room at all in the rules the industry will find it. So to be sure of banning the bad kind of securitisation it is likely that the rules will end up disincentivising the good kind too. If Dizard is right, then, Basel 3 will tend to favour large banks with a low cost of funds and plenty of capital.
The next part of the problem is that there are not too many of those left. So most banks will be forced to scale back on risk taking while they rebuild capital and delever. The consequences of that for the G10 economies will be profoundly negative: credit in all forms will become much harder to obtain, market making will be less practiced and hence liquidity will decrease, and growth will slow. Banks are too addicted to securitisation to go cold turkey without the detox effecting the broader economy. Instead I suggest central bankers should be looking at something more like nicotine patches: get rid of the most harmful form and manage dependence without generating too much discomfort. An over-reaction, however, seems most likely.
Labels: Basel, Regulation, Securitisation
posted by David Murphy at 07:30
0 Comments:
Post a Comment
<< Home