This column is a lead commentary in the VoxEU Debate "Banking reform: Do we know what has to be done?"
The past decade has witnessed rapid growth in a distinct form of financial intermediation: shadow banking. Today, in many advanced countries shadow banking rivals the traditional banking system. Since shadow banking is a recent phenomenon, its economic role is not yet well understood. This makes it hard to formulate a policy response or analyse existing proposals. Yet, regulation of shadow banking has been prominent in recent news (see FSB 2012).
Our recent paper (Claessens et al. 2012) aims to clarify the debate by focusing on the two economically most important shadow banking activities: securitisation and collateral intermediation. These functions are what we might call ‘bank-like’, as they involve risk and maturity transformation and can, like banks, be unstable. Indeed, the collapse of securitisation and the distress of dealer banks (which are central to collateral intermediation) were major factors behind the depth and the duration of the current financial crisis.
The first key shadow banking function, securitisation, is a process that repackages cash flows from loans to create assets that are perceived by market participants as almost fully safe and liquid. The repackaging occurs in steps, and takes the form of risk transfer (Figure 1). First, risky long-term loans are ‘tranched’ into safe and complementary (‘equity’ and ‘mezzanine’ respectively) tranches. Then the safe tranche is funded in short-term money markets, with additional protection provided by liquidity lines from banks. The resulting assets, such as Asset-Backed Commercial Papers (ABCPs), were regarded prior to the crisis by market participants as safe, liquid, and short-term, i.e. almost money-like, but with returns exceeding those on short-term government debt.
Prior to the crisis, the demand for these private money-like assets came from two sources. One was rapidly growing cash pools – held by corporations and the asset management complex – that faced a scarcity of safe investment opportunities (Pozsar 2011). Another was banks that used securitised assets for regulatory arbitrage (to minimise capital charges) and as collateral to attract repurchase agreement (or ‘repo’) funding (Greenlaw et al. 2008, Gorton and Metrick 2011).
Figure 1. The shadow banking securitisation process
Notes: The credit transformation special-purpose vehicles (SPVs) have matched maturity funding and issue asset-backed securities (ABS) or collateralised debt obligations (not shown). The maturity transformation SPVs are funded short term (are maturity mismatched) and issue asset-backed commercial paper (ABCP) or other structured money market instruments, such as auction rate securities (not shown). Private collateral (PC) includes ABS, corporate bonds, and equities. $1 NAV is the stable net asset value (a promise to repay at least $1 on $1 invested). R = repo; RR = reverse repo.
Effects of the crisis
The crisis showed some fundamental flaws in this process. Importantly, the perception of safety led to an ignorance of ‘tail risks’, the possibility of rare negative events, which ultimately materialised in the form of a broad decline in US house prices. As a result, claims that were initially perceived to be safe proved to be risky. When market funding dried up, banks faced unexpected exposures on their liquidity lines, on some of which they had to renege. This triggered a run on money funds that held many of the problem structured assets. The run in turn put other banks at risk, led to a breakdown of interbank markets, and caused an economy-wide freeze in credit to private borrowers, including non-financial corporations.
This form of securitisation has been largely inactive since the crisis. Once economic activity and private credit demand recover, some of this securitisation may resume. Securitisation will, however, most likely resemble itself as it was in the 1980s: subdued, with better recognition of risks, relying on more sophisticated investors.
Another key function of shadow banking is supporting collateral-based operations within the financial system. Such operations include secured funding (of bank and, especially, nonbank investors), securities lending, and hedging (including with OTC derivatives). Collateral helps deal with counterpart risks and more generally greases financial intermediation. One of the main challenges in using collateral is its scarcity. The shadow banking system deals with the scarcity through an intensive re-use of collateral, so that it can support as large as possible a volume of financial transactions. The multiplier of the volume of transactions to the volume of collateral (the ‘velocity’ of collateral) was recently about 2.5 to 3 (see Singh 2011).
A small number of dealer banks, all ‘systemically important financial institutions’, that is, banks whose failure could trigger a global financial crisis, are uniquely placed in their ability to facilitate collateral-based operations1. The dealer banks derive comparative advantages from economies of scale and network centrality effects, and (undesirably) from the perceptions of very low counterparty risks thanks to being too big to fail.
The best way to describe the re-use of collateral is to visualise it in chains (Singh and Aitken 2010). Dealer banks source collateral from parties that require funding (such as hedge funds), or from agents that want to enhance return by ‘renting out’ assets as collateral (insurers, pension funds, and sovereign wealth funds acting through custodians). Then, collateral is pledged to other parties to obtain funding or support other contracts. This starts a system of repeated re-use of collateral where a single unit can support multiple transactions (Figure 2).
Figure 2. An example of repeated use of collateral in a dynamic chain
Note: The over-the-counter (OTC) positions are in parentheses since they are off-balance-sheet items. UST = US Treasury bond; GS = Goldman Sachs; CS = Credit Suisse.
While facilitating the effective use of scarce collateral, collateral chains are associated with systemic risks and other distortions. Dealer banks are exposed to significant liquidity and credit risks (Duffie 2010), creating financial stability risks. Being systemically important financial institutions, they also benefit from cheap funding, which implicitly subsidises shadow banking. Additional implicit subsidies are obtained through the qualified financial contract status for derivatives and ‘repos’ that prioritises them in bankruptcy at expense of other creditors (Perotti 2010). A distinct part of the collateral intermediation process, the tri-party ‘repo’ market presents its own, and very significant set of systemic risks (Tarullo 2012).
The overall view
Shadow banking is a complex ecosystem. It combines multiple nonbank agents, is linked to traditional banks, and extensively uses the services of dealer banks. It is useful to visualise the shadow banking system, as in the Figure 3.
Figure 3. Financial intermediation through the shadow banking system
Notes: ABS = asset-backed securities; ABCP = asset-backed commercial paper; SPV = special-purpose vehicle. Full-size image available here.
The analysis of demand factors and regulatory weaknesses driving shadow banking helps clarify policy recommendations. Some parts of the shadow banking system are fragile and can pose systemic risks, yet commonly lack appropriate regulation. The most pressing concerns here are addressing risks in dealer banks, money market funds, and the tri-party ‘repo’ market; these are the focus of the recent Financial Stability Board proposals (FSB, 2012). Spillovers from the shadow to the traditional banking, and the possibility of banks using shadow banking for regulatory arbitrage also have to be addressed.
The crisis showed that demand-side pressures can lead to the creation of privately-provided safe assets, but these assets are unstable. Some proposals therefore suggest limiting the volume of shadow banking activities or integrating shadow banking in the main banking system. A more realistic proposal is to explicitly acknowledge the demand-side pressures by accommodating a shortage of safe and liquid assets with publicly guaranteed short-term debt.
It is also essential to consider broader macroeconomic issues surrounding shadow banking. Shadow banking is highly procyclical: secured lending and repos rely on mark-to-market prices and margins/‘haircuts’ that adjust over the financial cycle; securitisation produces claims that are inherently exposed to ‘tail risk’. Shadow banking is also hard to resolve in times of distress, since it encompasses many agents with complex contractual links. Shadow banking is also likely to have important interactions with monetary policy, both affecting interest rate transmission and being affected (e.g. in determining risk-taking) by prevailing interest rates. These issues raise specific sets of policies.
Addressing policy issues in shadow banking is a complex task. Research is yet to catch up. Some outstanding analytical issues include better differentiating economically-useful shadow banking activities from regulatory arbitrage. If we can better understand the economic value of useful activities, we can get cost-benefit insights for better regulation. Regardless, a policy response to address evident systemic risks is necessary and urgent. Such response, if effective, will probably make the shadow banking system smaller in size but still able to perform its useful economic functions in much safer ways.
Disclaimer: The views expressed here are those of the authors and do not necessarily represent those of the institutions with which they are affiliated.
Claessens, Stijn, Zoltan Pozsar, Lev Ratnovski, and Manmohan Singh (2012), “Shadow Banking: Economics and Policy”, IMF Staff Discussion Note, 12/12.
Duffie, Darrell (2010) “The Failure Mechanics of Dealer Banks”, Journal of Economic Perspectives, 24(1), 51–72.
FSB (2012) “Consultative Document: Strengthening Oversight and Regulation of Shadow Banking - An Integrated Overview of Policy Recommendations,” 18 November (Basel: Bank for International Settlements).
Gorton, Gary, and Andrew Metrick (2012), “Securitized Banking and the Run on Repo,” Journal of Financial Economics, 104 (3), 425-51.
Greenlaw, David, Jan Hatzius, Anil K Kashyap and Hyun Song Shin (2008), “Leveraged Losses: Lessons from the Mortgage Market Meltdown,” US Monetary Policy Forum Report, 2 February.
Perotti, Enrico (2010), “Systemic liquidity risk and bankruptcy exceptions”, VoxEU.org, 13 October.
Pozsar, Zoltan (2011), “Institutional Cash Pools and the Triffin Dilemma of the US Banking System”, IMF Working Paper, 11/190.
Singh, Manmohan, and James Aitken (2010), “The (Sizable) Role of Rehypothecation in the Shadow Banking System”, IMF Working Paper, 10/172.
Singh, Manmohan (2011), “Velocity of Pledged Collateral: Analysis and Implications”, IMF Working Paper, 11/256.
Tarullo, Daniel (2012), “Shadow Banking After the Financial Crisis”, Speech at the Federal Reserve Bank of San Francisco Conference on Challenges in Global Finance, 12 June.
1 The main dealer banks are Goldman Sachs, Morgan Stanley, JP Morgan, Bank of America-Merrill Lynch, and Citibank in the United States; and Barclays, BNP Paribas, Crédit Suisse, Deutsche Bank, HSBC, Royal Bank of Scotland, Société Générale, Nomura, and UBS. All are classified as SIFIs by the FSB.