SPEECH BY LORD ADAIR TURNER TO LENDIT EUROPE 2016
InterContinental Hotel, London: Tuesday 11 October 2016 (11am)
The speech can be watched at: http://www.lendit.com/europe/2016/videos/non-bank-lending-systemic-risk
(CHECK AGAINST DELIVERY)
It is a pleasure to be here at this LendIt conference and thank you Christine [Farnish] for inviting me.
As many of you will know, what prompted Christine was a press report of some comments of mine on peer-to-peer lending earlier this year: so I should begin by explaining what I said and why.
I didn’t intend to say anything about peer-to-peer lending, since the interview was actually about something completely different – and my comments were made when I assumed the interview was over
– but the journalist asked one of those “oh by the way” type questions, and I answered not thinking that my reply would be of interest to anyone.
And the question was something like “Will these crowd sourcing systems, these peer-to-peer lending processes in which individuals directly lend to other individuals or companies – will they entirely replace banks ?” to which I replied that if you think that you can replace good professional credit analysis with the individual judgements of non-professional investors, relying as it were on the wisdom of crowds – that is a delusion,
…and that, not immediately, but in say five to ten years’ time, when we face the next big upswing of irrational exuberance – individual peer-to-peer lending would probably produce losses which would make bankers look like geniuses.
And I still believe that that would be the case and would be a problem for the direct lending industry – if that was indeed the model of how direct lending worked. But in fact individual investor credit assessment now plays only a very minor role in the growing direct lending market…
…and having managed quite unintentionally to create a stir, the side benefit for me is that it has forced me to understand better what most of this industry is doing
And as a result, I am better placed to reflect this morning on what role the industry will play relative to banking, and whether its growth will make the credit supply system more or less risky.
1.Banks are dangerous things
Let me begin with the central problem with which prudential regulators struggle – that banks can be very dangerous institutions.
- Dangerous because they perform maturity transformation, lending out money for much longer term than their liabilities
- And dangerous because highly-leveraged, making promises to depositors who believe their investments are 100% safe , because of loss absorption by equity which even today, after all our post-crisis reforms, need only on average be some five per cent of total assets or liabilities .
- And dangerous because in aggregate banks don’t just take existing money and lend it on , they create credit, money and purchasing power which didn’t previously exist
…. enabling them to drive total economy leverage to excessive levels and to generate the cycles of credit and real estate price booms and busts which are not just part of the story financial instability in modern economies, but again and again the whole story .
So banks are dangerous – but also economically useful . For there’s a good argument that without maturity transformation -which makes it possible for people to fund long-term investments while holding short-term financial assets – it would be more difficult to generate the credit and investment needed within our economy.
Which is why our public policy approach to banks is not to prohibit them, but to impose capital and liquidity requirements – requirements which sadly were far too lax in the years before 2008, which we have tightened significantly since then, but which I believe should ideally be tightened much more.
So banks are useful, but create major risks.
2.Direct non –bank lending
Direct non-bank lending would seem to solve these problems. Because with non-bank lending, there is no maturity transformation, and there is no leveraged institution sitting between the end investors and end borrowers, but a direct contractual link, with each investor directly owning claims on real borrowers.
And it’s important to remember that non-bank lending has been very significant part of the credit system for a long time .
- We have always had trade credit extended from one business to another, and that is effectively a form of non-bank lending
- And for over a century, large companies have directly issued single name bonds which are bought by end investors. And its notable that since 2008, sluggish bank lending in both the US and Europe has been offset by corporate bond issues, with by far the largest increase in non-bank lending arising from such plain old-fashioned corporate bonds rather than from newly innovated forms of direct lending
- And after the 1970s, there was also an explosion of securitised credit – in particular in the US mortgage market – with multiple individual loans packaged into credit securities, sold to end investors, who held direct ownership stakes in the loans, rather than claims against intermediating banks.
So there is really was nothing new about non-bank lending – and indeed it has accounted for many years, for about 70% of all lending in the US, and 30% in Europe.
But over the last decade, and at an accelerating pace after the 2008 crisis, we have seen new forms of non-bank direct lending: including
- Growth of private debt funds, pooling the investments of institutional investors and high net worth individuals, and directly lending to the mid-corporate market
- Platform lenders concentrating on the small business market, and offering to their investors pooled portfolios of multiple small business loans of different combinations of risk and return
- And other platform lenders performing the same function in the market for personal unsecured loans
…and with multiple and continually evolving precise business models, involving for instance an increasing role for institutions within the investor base of the lending platforms.
And one consequence has been to extend non-bank lending to segments previously almost entirely dominated by banks, – in particular the SME segment, where securitisation has played only a trivial role in either Europe or the US, but where direct lending platforms are significant and growing
3.Benefits of direct lending – a new system of credit analysis ?
So what benefits could result from this greater role for direct lending?
Some rather excited commentary suggests that we’re seeing a revolution which entails new approaches to credit underwriting and supply.
But I want to suggest that the reality is more prosaic than the hype sometimes suggests, but that for that very reason direct lending is likely to become a stable , significant and useful part of our total credit supply system
…because while peer-to-peer lending platforms are not deploying a radically new approach to credit analysis, they might in some cases be able to do it better than banks, and they can certainly in many cases do it at least as well while providing better customer service.
The idea that peer-to-peer lending is radically different rests on the assumption that it really is peer-to-peer lending – that individual investors are driving their own selection of individual loans or portfolio of loans, and that the internet has made possible an information exchange rich enough to enable such decentralised credit analysis.
But that is not the dominant reality of the emerging industry.
For while the lending platforms are not maturity-transforming banks, the core function they perform is sophisticated and centralised credit analysis which enables them to present to the end investors a choice between portfolios of different expected risk and return
… and in some cases only a small percentage of investors now directly select specific portfolios of loans, and even those investors rely crucially on the platform’s centralised credit assessment of the relative riskiness of different loans.
And that credit assessment is not radically different in approach to that already deployed by the incumbent banks…
- for many years, banks have done credit underwriting for personal unsecured loans on the basis of automated credit scoring techniques – and the platforms do the same
- and for many years banks have been evolving a hybrid approach to small business credit underwriting, with a large and growing role for automated credit scoring but some remaining role for direct human contact – and lending platforms have tended to use the same hybrid approach;2016.10.11 – Lord Turner speech to LentIt Europe 2016
- while above the small business segment, when we move from loans of say £10,000 or £50k to loans of £5 million or £20 million, good credit analysis cannot be wholly automated, either by banks or by lending platforms, but has to depend not only on detailed quantitative analysis but on judgements about the entrepreneurial and managerial competence of the borrowers , and assessments of the quality of business plans, which often have to be informed by extensive face-to-face contact and customer premise visits.
This hierarchy of different credit approaches has been in place for many years, and I am not convinced that the growth of direct non-bank lending, or the use of information technology by challenger banks will radically change it.
But while the new lending platforms are not doing something fundamentally new in credit underwriting , both they and the challenger banks, might be able, if managed well, to do well-established forms of credit analysis as well or better than the incumbent banks, and they can certainly aspire to provide better customer service. That reflects two major competitive advantages:
- much tighter customer segment focus
- and the ability to get technology right first time, rather than struggle with existing legacy systems.
Yes, banks have been doing automated credit underwriting for unsecured personal loans for years, contradicting any idea that “Fintech” approaches to credit underwriting are entirely new, but platform lenders with a razor-sharp focus on applying the best big data analytical approaches to a specific chosen customer segment, might still be able to do it better.
And, yes, established banks have been doing a hybrid form of credit underwriting for small businesses for years, but it’s also true that large established banks have often struggled to achieve sustained focus on the small business sector, often switching its organisational home backwards and forwards between the retail bank and the corporate bank, but with small businesses a bit of a Cinderella in either home. A platform lender, or a challenger bank, totally focused on this customer segment, may be able to do credit writing underwriting at least as well as a major bank, while providing better and faster customer service.
And for both platform lenders and challenger banks, the freedom to build IT systems from scratch , gives an enormous potential service and cost advantage.
So while I have my doubts about how fundamentally new direct lending approaches are, and more generally about whether “fintech” is truly revolutionary, new competitors may still do very well simply by executing existing approaches better than can the incumbent players.
- Benefits of direct lending – a more stable system?
As a result, I am certain that many direct lending platforms, or indeed at the larger loan size level – private debt funds, will, along with challenger banks, play an increasing and profitable role within overall credit supply system.
But will more direct lending also make the financial system more stable? There is a prima facie case that it might,
- since a system of direct lending will not be subject to the risk of funding runs , which is inherent in any system of bank maturity transformation,
- and because the capacity of lending platforms to help funnel investor money from investors to borrowers is does not dependent on the adequacy of the equity cushion.
With banks we face the danger that when the economy turns down, and credit losses rise, those real economy effects can be magnified by a credit crunch in which banks which have incurred losses constrain further lending in order to conserve and rebuild capital.
The banking system indeed has an inherent tendency to impart pro-cyclical impetus to both economic upswings and down springs, and while regulators have recognised this, and introduced counter cyclical capital offers to offset this effect, those buffers are still untested and may prove to be only partially effective.
So it seems that – as some people put it – direct lending could add a “spare tyre” to the credit supply system, making credit crunches less likely.
And I believe there is some merit in this argument.
But I also want to introduce a caveat to this optimistic story, and one with implications for how macro prudential regulators should monitor the evolution of the direct lending market.
For we have been here before – with optimistic stories before the 2008 crisis about how non-bank lending through credit securitisation was going to make the financial system safer, with the spare tyre of non-bank credit supply abolishing the credit cycle.
Thus the IMF in its Global Financial Stability Report of April 2006, noted with approval a growing recognition “that the dispersal of credit risk by banks to a broader and more diverse group of investors… has helped to make the banking and overall financial system more resilient. The improved resilience” the IMF went on to suggest “may be seen in fewer bank failures and more consistent credit provision”.
Similarly Bill Dudley and Glenn Hubbard argued in a paper that non-bank credit supply was bound to be less volatile, and that as a result “ ‘credit crunches’ of the sort that periodically shut off the supply of loans to homebuyers… are a thing of the past”.
But in fact, this securitised form of non-bank lending, blew up spectacularly , producing first an over exuberant explosion of credit supply, and then a credit crunch quite as bad as any traditional banking system has ever managed to produce.
So what went wrong? The answer lies in the combination of at least three factors:
- first, the product complexity introduced by credit tranching. For the credit securities created were not simple pools of loans with all investors receiving pooled combination of risk and return, but tranched into layers of seniority, from super senior to senior , mezzanine and equity – in a way bound to undermine transparency, and make it difficult for investors truly to understand what they were investing in. All tranching processes create a significant systemic risk – and that indeed is one reason why banks are systemically risky , because in essence a bank is a risk tranching system rather like a CDO . But with banks we recognise that risk and we impose capital requirements to contain it;
- second, long and complex distribution chains in which the same securities passed through multiple different links, and which introduced maturity transformation at several points in the chain, so that in effect the multiple steps together became the equivalent of a maturity transforming banking system, but outside the remit of capital and liquidity regulation;
- third the fact that securities with long underlying maturity could be traded in somewhat liquid markets, in a process bound to produce risks as well as benefits . Like bank maturity transformation, market liquidity usefully makes it possible for short term investors to finance long-term real economy investments: but market liquidity also reduces incentives for good credit analysis, and generates the risk that swings of market sentiment, potentially exacerbated by mark to market accounting, may produce swings in credit supply.
As a result the securitised credit model of the pre-crisis years, far from reducing risks arising from the credit supply system, took the inherent risk of any bank-based credit system and put it on steroids.
So what follows for today’s non-bank lending system – seen from the point of view prudential regulator (or in this case a past one) concerned about future possible systemic risk?
The answer is keep it simple, and keep it transparent.
For what will worry prudential regulators is if, under the influence of competitive pressure and the continual process of financial innovation, the simple transparent models of today morph into more complex and opaque systems
- if loan portfolios are turned into tranched credit securities and sold along long distribution chains;
- and if people find ways to introduce leverage and maturity transformation into those chains – giving investors promises of immediate liquidity combined with apparent, though illusory, capital certainty.
Provided this does not occur, my judgement is that the direct lending industry will grow and play a useful role, alongside tightly-regulated banks, in our overall credit supply system.
How large that role should be, I don’t know, but that will be determined by competition not by regulation.