Ebook Imperfect Competition in the Interbank Market for Liquidity as a Rationale for Central Banking
The liquidity squeeze during the ongoing sub-prime crisis of 2007-08 has been likened by some observers, including the IMF, to the financial sector turmoil of the Depression era. A nagging problem faced by central banks during the early part of this crisis was the difficulty in getting open-market operations, discount window and securities lending to channel liquidity to the most needy parts of the financial system. Some of the lending facilities such as the discount window were not availed by players, and others when availed merely resulted in hoarding of liquidity by banks and other institutions.
In the UK, for example, banks’ liquidity buffers have experienced an almost permanent upward shift of 30% in August 2007 (relative to their pre August levels) and the result has been a rise in borrowing costs between banks and an almost complete drying up of liquidity exchange in money markets beyond the very short maturities. In response, central banks around the world, most notably the US Fed, have undertaken significant changes to their lender-of-last-resort facilities, in particular, by extending maturities of discount window and open-market operations, extending eligible collateral to include investment-grade debt securities, and making such adjustments for lending to primary dealers as well.
The episode begs several important questions: Why have the interbank markets, which in normal times act as lubricant to financial flows amongst banks, dried up so suddenly? Why have the traditional forms of central bank’s lender of-last-resort facilities failed to allocate liquidity to places needing it most? Indeed, going forward, how should central banks provide these facilities for them to be effective during crises? Do limits to the precision of supervisory information about banks compromise central banks’ effectiveness in overcoming the failure of money markets? Our paper attempts to answer some of these questions based on a specific market failure stemming from the exercise of market power in liquidity transfers between banks.
We propose that during crises, efficient liquidity transfers may not occur between surplus and needy banks. We attribute this inefficiency to the market power of surplus banks in the market for interbank liquidity transfers and the strategic gains they derive from buying assets from needy banks at fire sale prices and, more generally, gaining market share at their expense. We determine conditions under which a central bank can mitigate this inefficiency by standing ready to lend to needy banks. We report historical episodes in support of this rationale for central banking and discuss implications for recent debates on the supervisory and lender-of-last-resort functions.
We consider liquidity transfers through two markets: the interbank lending market and the asset sales market. Our model has three main ingredients. First, we assume that some assets are bank-specific, i.e., they are worth more under current than under alternative ownership. For instance, alternative owners may lack the current owner’s expertise. For this reason, asset sales may be less efficient than borrowing. Second, we assume frictions in the interbank lending market, which we model as arising from a moral hazard problem. Specifically, we assume that banks can monitor their assets to improve their performance, and that monitoring is costly. A bank borrowing on the interbank market must retain a large enough claim on its own assets to have incentives to monitor them. This friction limits banks’ borrowing capacity, leading to inefficient asset sales. Third, we assume that during crises, liquidity is concentrated with a few banks, giving them market power.
In this context, we show that the market power of surplus banks can lead to more asset sales, and importantly, more inefficient asset sales by banks in need of liquidity. The intuition is as follows. Banks with market power in the interbank lending market will charge higher interest rates to banks seeking liquidity. As a result, the latter will retain smaller claims on their assets, which diminishes their incentives to monitor them. If monitoring is low enough, it becomes preferable to sell some assets to a liquid bank. These asset sales also generate liquidity that can provide relief for the selling bank’s other assets. The higher the liquid banks’ market power, the higher the interest rate and the greater the illiquid banks’ incentives to sell assets.
Surplus banks’ ability to exploit market power is limited by the outside option provided by the market for liquidity outside the banking sector. Therefore, the problem and the implied inefficiencies are more acute the weaker is the outside market, a scenario that would arise, for instance, in liquidation of information-sensitive and bank-specific loans made to small borrowers.
Our analysis also implies a rationale for the existence of a central bank. A central bank that is credible in providing liquidity to needy banks curbs the market power of surplus banks in the interbank lending market and thus improves the efficiency of liquidity transfers. In particular, the central bank can play a “virtual and virtuous” role: In our model, it never actually lends to needy banks, but merely improves their outside options when bargaining for liquidity with surplus banks. We show however that for such an improvement requires the central bank to be ready to extend loss-making loans and/or be better than outside markets at extending loans to needy banks. The latter situation is more likely if the central bank has also a supervisory role, allowing it to improve its ability to monitor its loans to needy banks.
We also study the possibility for banks to insure against liquidity shocks. This possibility reduces the inefficiency in interbank liquidity transfers. However, as long as banks can only get partial liquidity insurance, surplus banks’ ex post market power still increases or creates inefficiencies in the allocation of liquidity. Indeed, if banks that are likely to have excess liquidity and market power ex post are also the best liquidity insurers ex ante, their market power ex post will reduce the scope for liquidity insurance ex ante. Other reasons for liquidity insurance to be only partial include the impossibility to enter binding long-term contracts, the fragility of implicit contracts in crises situations, or the possibility of aggregate liquidity shortage combined with liquid banks’ cost of capital being non-verifiable.
To summarize, our model illustrates that the public provision of liquidity (in fact, its mere credibility) can improve its private provision even in times of aggregate liquidity surplus. This lender-of-last-resort rationale for the existence of a central bank complements the traditional one pertaining to times of aggregate liquidity shortages and contagious failures (e.g., Holmstrom and Tirole (1998), Diamond and Rajan (2005), Gorton and Huang (2006)). Our analysis also clarifies why central banks should assume both roles of supervisor and lender-of-last-resort.
Our paper is related to the literature on the failure of interbank markets that justifies the lender-of-last-resort role of central banks. Goodfriend and King (1988) argue that with efficient interbank markets, central banks should not lend to individual banks but instead provide sufficient liquidity via open market operations, which the interbank market would then allocate efficiently among banks. Others however, argue that interbank markets may fail to allocate liquidity efficiently due to frictions such as asymmetric information about banks’ assets (Flannery (1996), Freixas and Jorge (2007)), banks’ free-riding on each other’s liquidity (Bhattacharya and Gale (1987)), or on the central bank’s liquidity (Repullo (2005)). Instead, our paper focuses on the (additional) frictions brought about by market power.
Donaldson (1992) is, to our knowledge, the only paper with a similar focus. Using Dunn and Spatt (1984)’s strategic pricing model, it shows that even if aggregate liquidity is in surplus, if some banks have a significant proportion of the excess cash so that the other cash rich banks’ resources are not enough to satisfy the total liquidity demand, banks can exploit this captive demand and charge higher than competitive rates.
While some theory papers study the reallocation of funds (e.g. Holmstrom and Tirole (1998)) and others that of assets (e.g. Shleifer and Vishny (1992), Gorton and Huang (2002)), ours studies both and illustrates the trade-offs involved. We believe banks’ dual role as each other’s financiers and business competitors to be an important and specific aspect of their relationships.
Our paper is also related to the literature on predation, i.e., when rivals take actions that weakens a firm’s access to finance (Bolton and Scharfstein (1990), Cestone (2000)). Here, how-ever, the dual relationship between banks means that the predator is also the prey’s financier.
The paper proceeds as follows. Section 2 reports historical evidence. Section 3 presents the model and Section 4 its analysis. Section 5 discusses the banks’ outside option. Section 6 presents the rationale for central banking. Section 7 liquidity insurance and its limits. Section 8 discusses broader policy implications and failures in liquidity transfers among non-bank financial institutions. Section 9 concludes. Proofs are in the Appendix.
Posted in :
