Of Sunlight, Shadow Banking and Horizontal Intermediation

Huge thanks to Gillian Tett for pointing out in Friday's FT the report on "Shadow Banking" by staff of the Federal Reserve Bank of New York. It even comes with a handy map, which she points out:

"...is a reminder of how clueless most investors, regulators and rating agencies were before 2007 about finance. After all, during the credit boom, there was plenty of research being conducted into the financial world; but I never saw anything remotely comparable to this road map."

The NY Fed's report defines "traditional banks" as depository institutions that are insulated by public sector insurance from sudden 'runs' on deposits. Whereas "shadow banks" - finance companies, credit hedge funds, broker-dealers and an alphabet soup of intermediaries - operate without any such public sector insurance. I'll come to what shadow banks actually do shortly. But the fundamental problem, as the report makes clear, is that the shadow banking system has indeed gained access to the public purse during the financial crisis, via bail-outs of some of their subsidiaries (bank subsidiaries, in the case of 'financial holding companies' or retail/commercial banking groups; and industrial loan companies and federal savings banks, in the case of diversified broker-dealers or investment banks):

"The liquidity facilities of the... government agencies' guarantee schemes were a direct response to the liquidity and capital shortfalls of shadow banks and, effectively, provided either a backstop to credit intermediation by the shadow banking system or to traditional banks for the exposure to shadow banks."

In other words, public sector guarantees are necessary to stop 'runs' on the shadow banking system in the same way they avoid 'runs' on deposit-taking banks.

In these circumstances, you could be forgiven for thinking it crazy for the shadow banking system to go unregulated. But who are the shadow banks? What do they really do? Does each function really need to be regulated? If so, why? How? This has been the focus of the European regulatory agenda (including a focus on unregulated derivatives) during the past few years, and the harsher elements have met fierce resistance with allegations that the realities, wider implications and scope for unintended consequences are not well understood by legislators and regulators.

So the publication of the detail in NY Fed's report is very much worthwhile.

The NY Fed concludes that 'some' but not all segments of the shadow banking system are of "limited economic value". It says that "equally large segments of it have been driven by gains in specialisation" and would be more aptly described as a "parallel banking system." Nevertheless, the report concludes that "private sector balance sheets will always fail at internalizing systemic risk. The official sector will always have to step in to help."

So how does the shadow banking system work? Well, instead of making loans to hold onto them, like traditional banks, the shadow banking system involves the making of loans for sale through a series of intermediaries - "shadow banks" - each of which specialises in one step in "a vertically integrated, long, intermediation chain":

"These steps essentially amount to the “vertical slicing” of traditional banks’ credit intermediation process and include (1) loan origination, (2) loan warehousing, (3) [Asset-backed Security or ABS] issuance, (4) ABS warehousing, (5) ABS [Collateralised Debt Obligation, or CDO] issuance, (6) ABS “intermediation” and (7) wholesale funding... Typically, the poorer an underlying loan pool’s quality at the beginning of the chain (for example a pool of subprime mortgages originated in California in 2006), the longer the credit intermediation chain that would be required to “polish” the quality of the underlying loans to the standards of money market mutual funds and similar funds."

In reality, these chains could involve many more steps and "CDOs squared", depending on how many times the loans had to be "polished". The various exhibits in the report illustrate the eye-watering complexity very well.

Having studied the various processes, the NY Fed believes that "regulation by function is a more potent style of regulation than regulation by institutional charter." Figuring out which functions contain the root cause of our current financial woes is therefore necessary.

The NY Fed believes the current financial crisis grew out of mispricing of ABS CDOs, (steps 5/6 above), which caused problems up and down the chain:

"...the growth of ABS CDOs not only masked but also created an underlying pricing problem in the primary ABS market (Adelson and Jacob, (2007)). In particular, in the early days of securitization, the junior tranches of home equity deals were purchased by real money investors. However, these investors were pushed aside by the aggressive buying of ABS CDOs, which resided on the trading books of large broker-dealers. The mispricing of the junior ABS tranches permitted issuers to distribute loan pools with increasingly worse underwriting. ABS CDOs suffered from the same underlying problem as the underlying ABS, which required the creation of CDO-squared products."

In essence, poor quality debt was re-packaged again and again in order to remove the risk, but the risk was misunderstood and the resulting instruments were mispriced each time.

Notwithstanding these pricing problems, the NY Fed's regulatory vision is that vertical credit intermediation can reduce the costs of screening and monitoring borrowers in the traditional banking model, and facilitates investor diversification, by transforming credit quality, maturity dates and adding liquidity. And the grading of securities by a "credible rating agency" can reduce information asymmetries between borrowers and savers.

Yet this assertion meets a series of fundamental challenges, not all of which are explicit in the report:

1. The separation of lender and borrower, and fragmentation of the original loan note makes it harder to adjust loans when borrowers get into trouble (explored in Confessions of a Subprime Lender and evident from the 'fraudclosure' and 'forced repurchase' problems in the US).
2. The process of transforming 'maturity' (changing the date when loans or debt instruments are due to expire) creates balance sheet risk for the intermediary.
3. It is unclear whether ratings, accounting and audit functions really do remove information asymmetry between borrowers and lenders. Do we have "credible" ratings agencies, when only three dominate the market and they are paid by the issuers of the securities they grade? Similar problems exist in the accounting and audit markets - hence the calls for reforms in these areas.
4. There are huge challenges to undertaking adequate due diligence on large volumes of underlying original loans.
5. Pressure to reduce the amount of capital required to operate this vertical chain of intermediaries results in a game of regulatory, tax, capital and ratings arbitrage that spans the globe and creates endlessly complex corporate structures.
6. Various factors lead to underestimation of the capital required for the private and implicit public sector guarantees required to support it. This is further complicated by the fact that "...the performance of highly-rated structured securities... in a major liquidity crisis... become highly correlated as all investors and funded institutions are forced to sell high quality assets in order to generate liquidity."
7. The knowledge that the market can ultimately 'put' problem securities on the taxpayer (whether this is explicit, implicit, direct or indirect) creates a moral hazard that seems to increase in line with the demand for the securities until the system irretrievably melts down.

These fundamental challenges and the length of time it is taking to confront them underscore the need to find alternatives to the vertical model for credit intermediation. The situation is all the more urgent, given that the huge taxpayer bail-outs of the past few years have only reduced the liabilities in the shadow banking system from $20 trillion in March 2008 to about $16 trillion in Q1 2010, when liabilities in the traditional banking system were about £11 trillion.

One alternative is horizontal credit intermediation, which is a feature of the new peer-to-peer funding platforms - like Zopa, Lending Club and Funding Circle - that each borrower's loan amount is provided via many tiny loans from many different lenders at inception.

That approach does deal with the fundamental challenges outlined above, mainly because the margin between lending and borrowing rates are too slim to support more than one intermediary. The one-to-one legal relationship between borrower and lender/loan owner is maintained for the life of the loan via the same loan origination and servicing platform (with a back-up available), allowing for ready enforcement. The intermediary has no balance sheet risk, and has no temptation to engage in regulatory, tax or other arbitrage. Loan maturities do not need to be altered to achieve diversification across different loans, loan terms and borrowers. The basis of the original underwriting decision remains transparent and available as the basis for assessing the performance of the loan against its grade, as well as for pricing the loan on any resale or refinance, making due diligence easy. To the extent that credit risk were to concentrate on certain borrowers or types of borrowers, those risks would remain visible throughout the life of the loan, rather than rendered opaque through fragmentation, re-packaging and re-grading. The scope for moral hazard is contained by the transparency around loan performance.

This also reflects a trend towards the democratisation of markets that impact consumers.

But I would say that, wouldn't I?