Location: Cornerstone Research dinner, Skinner’s Hall, London
I am very grateful to Cornerstone Research for giving me the opportunity to speak to this expert group.
My observations are mostly based on Kraus and Aquilina’s Occasional Paper 18 (OP18), ‘Market-Based Finance: its contributions and emerging issues’ and our related article ‘The rise of market based finance’.
We wrote OP18 because, as a competition authority, we were concerned that much of the public debate about shadow banking focussed on regulatory arbitrage. We believed that there might also be an important story to tell about shadow banking as a manifestation of an appropriate and positive process of competition based on new ideas and technological change.
What do we mean by ‘shadow banking’?
There is a narrow definition: credit intermediation carried out by non-banks. This is usually thought to have four main characteristics: maturity transformation, liquidity transformation, leverage and credit risk transfer. The problem with this definition is that it focuses on the institution that performs the activity – as if the activity were identical to what banks do -but omits what is truly distinctive.
Think about it a moment. Banks are large incumbents in credit markets. If different institutions offered credit in the same way as banks and at roughly the same prices and quality, why would anyone bother to spend their time on switching to the unknown? Such entry would be unlikely to be successful. But shadow banks have grown enormously. So there must be something economically distinctive about them. I argue that it is the way in which prices and/or quality are improved.
Specifically, the critical point about shadow banking is the approach to credit intermediation. That is, it is credit intermediation carried out on and priced on global markets for money and risk.
Does it make more sense to talk about ‘market-based finance’ (MBF)?
I think it follows from my answer to question 1 that it does make more sense to talk about MBF. This captures the essence of the activity. Also, the name is not value-laden. As I will explore below, MBF is not just about banking in the shadows with a view to benefitting from compensation schemes and bail-outs that are entirely paid for by others. And, I might add, doing so without even having to inject into the business the socially optimal level of capital – the amount that is considered to force banks to internalise externalities or to reduce the probability of crisis to a socially acceptable level.
Another advantage of the MBF label is that it gives an intuition about why MBF may be more efficient than traditional banking. The latter depends upon bilateral relationships. The potential advantage of these is that lenders over time build up substantial private information about their customers. But it is still in practice largely a fixed relationship in which the matching of supply and demand may be highly imperfect. Indeed switching may be difficult, leading to inefficiently high prices. In MBF, on the other hand, the market works as a matching device bringing together different suppliers of credit and capital and a range of would-be recipients. In principle, this should enable both sides to get a better fit, in terms of matching risk preferences and supplier characteristics.
It may be thought that banks’ expertise in credit screening trumps this market mechanism. On the other hand, for example, new fund management companies like Motif that exploit their own expertise plus the wisdom of crowds appear so far to be out-performing traditional active managers. Even more to the point, there is a long history of lending by banks that was so imprudent that bank crises ensued. Either banks did not screen credit well or they chose to ignore the results of their own screening. There is a lot of literature on this, including an interesting approach, based on the Bank of England’s archives, in Bank Underground, Christmas Special, 2016.
What is the story on regulatory arbitrage and competition?
I do not argue that no shadow banks undertook regulatory arbitrage. There almost certainly was some cynical activity. But this does not mean that healthy competition did not produce other examples of MBF. Nor does it mean that it cannot be reasonable for credit intermediation to occur without societal intervention in terms of capital, liquidity and so on.
On the first point, we describe in OP18, following Mehrling et al, how one can utilise existing markets (global money markets, asset management, derivatives and capital funding) to produce new, market-based means of credit intermediation. This is innovative use of existing markets.
A complementary development is exploitation of new technologies to bring lenders and borrowers together. This is the credit intermediation version of a story familiar in many markets: the use of on-line platforms. Virtual currencies and distributed ledgers are also part of this story. To view MBF as a whole, we contended that it may be considered to include certain activities of a very wide range of entities, for example the off-balance-sheet lending of global banks. For brevity, I list the institutions only:
- global banks
- money market mutual funds
- finance companies for credit cards, mortgages, etc
- central clearing counterparties
- special purpose vehicles
- business development companies
- asset backed commercial paper conduits
- pension funds
- insurance companies
- credit hedge funds
- asset managers
Analysis by Riasi using the Porter Diamond suggests that MBF is not due to chance, Government action and factor conditions (raw materials, capital, human resources, etc) but is due to:
- firm strategy, structure and rivalry
- demand conditions
- related and supporting industries (distributors, researchers, finance, tech, etc)
We might quibble about the diamond having six aspects, and what the role of tougher regulation was: higher bank capital requirements do suppress lending in the real economy for a period, as our research in the Financial Services Authority (FSA) showed. And the resultant unsatisfied demand must create opportunities for other providers of credit. But the main story arising from the above analysis using the Porter Diamond is clear.
On the second point in this section - that sometimes it is reasonable for credit intermediation to occur without societal intervention in terms of capital, liquidity and other requirements – the critical issue is whether the activity can or cannot impose material externalities, whether directly or indirectly. One obvious case of indirect externalities is where market based financiers are funded by banks to such an extent that their failure could undermine banks sufficiently to cause a bank run.
Sometimes, however, MBF operates in a way similar to the (hypothetical) model of 100% capital-backed lending by banks that was much discussed after the financial crisis of 2008: any losses are born fully by ‘equity’ investors so, arguably, there should be no threat to stability or to payments systems.
What are the benefits of MBF?
From macro perspective, the benefits may be very large. MBF has been estimated as being in the region of £14 trillion in the UK. To be clear, this number is the total liabilities funding relevant activities of the MBF entities in my list above. Also, to be sure, the amount included with respect to CCPs is zero. £14 trillion is a lot of finance and should mostly translate into beneficial economic activity. To give a sense of scale, the amount is approximately eight times UK GDP and greater than China’s GDP. In some large, developing economies, up to 35% of total business finance is supplied by MBF.
The FSA research already mentioned, which was undertaken using the National Institute’s (NIESR’s) macro model, implies that if there had been no MBF the prolonged recession after 2008 would have been even deeper. This is a specific case of the so-called ‘spare tyre’ argument. In other words, when traditional banks are unwilling or unable to lend, MBF may be available, reducing the impact of economic shocks.
From the macro to the micro, and an example of MBF making a market work well. The Economist reported the case of Hall and Woodhouse. This is the brewer of (excellent!) Blandford Ales and a very successful and long-established company. After the financial crisis, Hall and Woodhouse’s bankers sought to increase the price of their credit and shorten its duration. H&W itself had done nothing to deserve this, so sought alternative funding. At a time when most banks were reluctant to take on new customers on favourable terms, it eventually obtained funding for an appropriate term and an acceptable price in the shadow banking market, where M&G acted as broker between H&W and underlying providers of funds.
The H&W story is a nice example of MBF acting as a competitive constraint on the banking market, which could otherwise be oligopolistic.
Another pro-competitive aspect of MBF is that it can solve the problem of scaling up. This is because it can draw on a wide pool of investors, thereby providing access to larger markets for products that are successful in small niches.
What costs and risks arise from MBF?
Issues to consider here are risks to stability and direct risks to consumers. On stability and MBF, there is a considerable literature. The European Central Bank (ECB) sees MBF as providing important services to the real sector, including market risk sharing and liquidity risk sharing, and bridging information gaps. But it is concerned about leverage and funding of MBF entities by banks: large outflows from mutual funds might cause systemic risk given maturity and liquidity mismatches. The ECB does not have data to quantify this risk and one should consider the true likelihood of extremely large systematic outflows. Even if the past is not necessarily a guide to what will happen in funds!
The effects of credit shocks on one point in a network may depend upon a number of layers of connected exposures and eventuate in failure at distant points in the network.
As an aside, Europe-wide or even national level network analysis is extremely challenging. The effects of credit shocks on one point in a network may depend upon a number of layers of connected exposures and eventuate in failure at distant points in the network.
Luck and Schempp argue that the systemic risk in MBF depends upon its size relative to the trading markets for its assets but interestingly, in their model, see some reduction in risk if regulated banks operate ‘shadow banks’.
Vives argues that big changes in financial regulation post-crisis have caused major distortions in competition. He accepts the idea of a trade-off between competition in banking and stability (to be clear, much literature does not) but says that this trade-off in fact depends upon the trade-off between competition policies and prudential regulation. Thus these need to be co-ordinated. Otherwise, more competition, without the attention to market failures needed to improve social welfare, may produce greater instability, in turn leading to greater incentives to take risks.
Another aside. If we think of regulation as pertaining only to particular, given markets and as a reaction to perceived imperfections or failures in these observed markets, we are indeed in the place just described. Stability and other benefits will depend upon correcting the failures. On the other hand, if we conceive the role of regulation to include market design such that markets as a whole efficiently do what they are meant to do for the real sector, we can try different approaches. For example, it may be that having lots of separate retail debt product markets or even separate retail debt product and asset product markets imposes inefficiency.
More to the point, for present purposes, a more competitive set of capital and risk markets could allocate risk and capital more efficiently. This could in principle lead to greater sustainable growth and consequently stability. Arguably, this ‘macroconduct’ idea is an interpretation of the FCA’s overarching objective: to make markets work well.
The risks to consumers from MBF typically arise from hidden risk. There are many cases, and not just in MBF, in which complex products in wholesale markets form the basis for retail offerings that neither their buyers nor, quite possibly, those advising the buyers truly understand in terms of financial mathematics.
What might happen next?
The first point here is due modesty. There are fundamental uncertainties. Of more than one kind. What will happen when QE is unwound? Will personal debt reach some indisputably unsustainable level and collapse? Will there be a major political or other shock, perhaps given changes in the world’s statesmen? Any of these could have implications for MBF, and for much else besides.
There are also narrower but important points of uncertainty. Markets with long histories have somewhat predictable behaviour, though people who have tried ‘timing’ strategies on established equity exchanges – that is, have exited the market as it starts to turn down with a view to re-entering before it rises materially - have generally found the opportunity cost far higher than expected! But in new markets such as MBF there is fundamental uncertainty. It is hard to know what will take hold and prosper, and what will fail. Consider Betamax. And the recent example of 3DTV.
It does seem safe to say that we will see new products that exploit new technologies such as online platforms and other matching devices, information pooling sites, other big data opportunities and distributed ledgers. To speculate slightly, given that we are talking fundamentally about credit, it may be interesting to see what business models credit rating agencies and credit reference agencies develop from their huge databases. In the era of MBF, do they really need banks?
Another important point is that even if authorities believe they have a reasonable intuition about how a market such as MBF may develop and some negative consequences that might ensue, there is still reason for caution in any response. As mentioned already, there is a significant and relevant debate about the relationship between competition and stability.
There is also a significant debate about ‘the risk of risk models’ in the sense used by Danielsson and James; that is the ability of authorities to work out precisely which firms are systemically risky and which are not. There are also problems with determining how the decisions that regulators seek to change are taken under the many forms of uncertainty present in financial services, to the extent that the EPSRC has just provided very substantial grants to Tuckett and Smith (UCL and the LSE) to investigate this question.
Moreover, there are problems determining what the effects of regulatory remedies will be, as the recent exchanges between the CMA, Ofgem and Professor Martin Cave show: it is not even clear when one should use remedies that cut across competition like price controls and product standards and when one should try to empower consumers and competition with information disclosures… partly because consumer decision-making itself is hard to understand, despite the FCA’s very thorough work on behaviourally informed remedies.
Again, Roger Miles argues that supply-side behavioural remedies, meaning remedies designed to promote compliance, are likely to be more effective than attempts to influence the demand side.
A lesson arising from these points is, perhaps, that one should facilitate and welcome market solutions to market problems wherever they can occur, and consider very carefully whether they really can be bettered. There is always somebody trying to make money by solving somebody else’s problem, to mutual advantage.
As we imply in OP18, market based finance has much to offer. We agree with the analysis of Schliephake and Martinez-Miera that the ability to screen and monitor borrowers is not unique to banks so that MBF can efficiently lower the costs of credit intermediation, avoid externalities (because losses fall on investor) and make competition more intense. Overall, if MBF is monitored carefully, and any significant risks clearly arising are addressed, it can generate significant welfare gains for society.