Shadow Banking: Why Modern Money Markets are Less Stable than 19th c. Money Markets but Shouldn’t Be Stabilized by a 'Dealer of Last Resort'

Abstract

An important policy question that is currently being discussed by central bankers and academics is whether the “shadow” banking system should have a permanent backstop from the central bank akin to the extraordinary support that was provided by the Federal Reserve to the shadow banking system in 2008. I answer this question by (i) using a macroeconomic analysis of banking (ii) to explain the role played by the first lender of last resort in 19th c. Britain and (iii) applying this analysis to the modern shadow banking system. I conclude that because of its heavy reliance on collateralization, the shadow banking is a poor substitute for the traditional banking system and does not merit the support of a “dealer” of last resort.

My macroeconomic framework explaining the role played by banks in economic growth builds on Cavalcanti and Wallace (1999) and was developed in Sissoko (2006) and Sissoko (2007). This framework explains that in an environment with liquidity constraints banks are special because their history of default is public, so unsecured bank borrowing is incentive compatible when bankers profit from their special characteristic by underwriting the debt of the non-banks in the economy. This debt circulates as bank liabilities and can resolve the liquidity constraints faced by the economy entirely, facilitating economic growth (and generating fees for bankers). Because being liquidity constrained is a significant penalty, a simple trigger strategy played by banks against defaulting non-banks can support an environment where borrowers limit their own debt, credit is generally available, and the economy grows to its maximum potential. The distinctions between this model and the most important competing model of banks as issuers of “information insensitive” assets in Gorton and Ordonez (2014), are first that safe assets take the form of unsecured debt, and second, that they are actually safe, because borrowers themselves do not want to borrow more than they can pay. Crises take place in my environment when there is a loss of confidence in the credit system.

To explain the role of the lender of last resort, I turn to the actual historical environment on which my model of banking is based, 18th-19th c. Britain. In Britain the privately-issued bills that circulated in the money market were so safe that collateralizing them was viewed as clearly superfluous. Several institutions made these privately-issued bills safe: they were guaranteed by at least two parties, the issuer, the bank-acceptor, and anyone who chose to sell the debt, bank owners faced capital calls for bad debts the bank had guaranteed, the bills were short-term, and the credit system itself was protected by the existence of a lender of last resort. The lender of last resort made the provision of banking services incentive compatible for bankers who effectively faced unlimited liability by providing high-powered money in liquidity crises and thereby averting the threat of a transition to a no-credit equilibrium. The lender of last resort played a second very important role: in order to maintain the “safe” quality of the money supply the central bank withdrew support from any bank that it believed to be overissuing debt. The words “last resort” themselves refer to this management of moral hazard by the central bank.

Shadow banking developed in an environment where the second duty of the lender of last resort had been forgotten, and solvency for large banks was effectively redefined to incorporate public sector support. Shadow banking has been described as market-based short-term finance of long-term lending, but in fact refers to a bank-guaranteed system of finance based on commercial paper and repurchase agreements, that provides little funding for private sector assets and significant funding for investment banks. In other words, to the degree that shadow banking has disintermediated commercial banks, it has done so by reintermediating investment banks – using repurchase agreements, which are a form of funding that is even more unstable than deposits, due to their reliance on collateral that is remargined daily. Furthermore, the growth of the collateralized money market that shadow banking represents is probably destabilizing the incentive structure that is the product of centuries of institutional evolution and that undergirds the traditional unsecured money markets upon which the past 250 years of economic growth have been founded.

Although it is often claimed that the purpose of a “dealer of last resort” is to support liquidity on asset markets, in practice, the policy is designed to give the largest dealer banks access to central bank credit and to support them though a crisis – as we saw in 2008. I explain that this policy only protects asset markets from fire sales of assets belonging to the select group that has access to central bank lending, not from fire sales in general, because dealer banks do not extend credit in the same way that commercial banks do. Finally, commercial banks traditionally bear risk for the economy, whereas dealer banks traditionally avoid bearing risk over time themselves but instead facilitate the allocation of that risk to others. As a result, dealer banks should not receive support similar to that of commercial banks, because they do not play the same role in the economy that commercial banks do.