RDP 2008-06: Promoting Liquidity: Why and How? 5. An Increase in Holdings of Liquid Assets

In the various assessments of the recent credit market turmoil, a frequent conclusion has been that financial institutions and supervisors did not pay enough attention to liquidity risk. FSF (2008), for example, lists a number of shortcomings in liquidity management. These include banks: not adequately planning for system-wide stress; not adequately considering the links between market liquidity, funding and credit risk; and not anticipating the need to honour committed lines of credit or the need to provide financing to clients in order to protect their own franchise value.

Essentially, these reviews are arguing that banks have held too few liquid assets, or assets of unpredictable liquidity, and have under-priced the provision of liquidity services to their customers. It is difficult to argue with this conclusion, as it now seems clear that, over recent years, proper liquidity management slipped off the radar screen for many financial institutions. A number of recent reports have pointed to the way forward here, including the more extensive use of stress tests, the development of robust contingency funding plans, and the need to allocate appropriate liquidity to all business lines (see, for example, BCBS 2008; IIF 2008; IMF 2008). Financial regulators are likely to have a role to play in achieving progress on a number of these fronts, as private institutions are unlikely to fully internalise the benefit to the system as a whole of maintaining high levels of liquidity.

This points to important questions that do not seem to have attracted the attention that they deserve: that is, to what extent financial institutions should be required to fully ‘self-insure’ against system-wide liquidity problems, and to what extent the public sector should assist when such problems emerge. In raising these questions, we want to make it clear that, in most cases, institutions should be able to deal with idiosyncratic liquidity problems, without any assistance from the public sector. Furthermore, institutions need to be able to deal with significant disruptions to asset markets and to their funding sources. But full self-insurance against generalised and widespread disruptions could come at a significant cost to both financial institutions and the economy more broadly. As a very rough illustration, suppose that such insurance required institutions to hold an extra 10 per cent of their balance sheets in highly liquid, high-quality assets, and that the expected return on these assets was 1 percentage point lower than the alternative. This type of portfolio shift would reduce the banking system's return on assets by 0.1 of a percentage point, and the return on equity by around 2 percentage points. Institutions might then be expected to increase their lending margins, which in turn might lead to a lower stock of capital in the economy and less output than might otherwise be the case.[8] In addition, if financial institutions had to fully self-insure they might not be prepared to provide as much long-term funding as is currently the case. The issue is whether some insurance by the public sector is a better way to deal with these problems than financial institutions having to deal with them alone.

The extent to which financial institutions insure against system-wide liquidity problems is a current issue in Australia, as it is in many other countries. Over recent decades, the Australian banking system's holdings of ‘liquid’ assets have fallen significantly as a share of their aggregate balance sheet. In the 1960s, around 30 per cent of the banks' total assets were held in government securities, and a further 8 per cent were held on deposit at the Reserve Bank of Australia (RBA) (although the vast bulk of these assets were held to meet regulatory requirements and so were not available for short-term liquidity purposes). Today, government securities account for just 0.5 per cent of total assets, and deposits at the RBA account for a further 0.2 per cent. This decline reflects both regulatory changes and a reduction in the supply of government securities on issue.[9]

A related feature of the Australian environment is that around 90 per cent of the Australian banking system's liquid assets are ‘inside assets’, by which we mean the liabilities of other financial institutions (Figure 3). As at May 2008, these assets accounted for around 15 per cent of the system's domestic assets, which is up from 12 per cent a year earlier. When the strains first developed in financial markets in August/September last year, the banks' demand for liquidity increased significantly and, in response, they issued securities to one another, allowing each to record an increase in their liquid assets. Of course, at the same time, the banks' short-term liabilities also increased. This heavy reliance on inside assets is unusual by international standards. In the United States, for example, banks' holdings of such assets account for around 6 per cent of their total assets, with securities issued by the US government and federal agencies accounting for a higher 14 per cent.

Figure 3: Banks' Liquid Assets

This reliance on inside assets poses some challenges for dealing with system-wide liquidity problems, particularly if those problems are associated with system-wide credit quality concerns (which has not been the case recently). There are, however, simply not enough ‘outside assets’ in Australia for banks to hold the bulk of their liquid assets in securities issued by entities other than banks. Currently, the total stock of outstanding Commonwealth Government bonds is around $55 billion, with another $70 billion of state government bonds, and $45 billion of supra-national debt. This is in comparison to the total liquid assets of the banking system of around $350 billion.

Reflecting these developments, the RBA has, over the past decade, broadened the range of assets it will accept in repurchase agreements (‘repos’) to include securities issued by financial institutions. This has substantially increased the stock of securities that the RBA will accept under repo in its market operations. In comparison to a situation in which banks hold their liquid assets in outside assets, this potentially exposes the RBA to more risk; however, this increase in risk is limited by the fact that in the normal course of operations, banks are not able to sell their own or related securities to the RBA under repo.

In the following section we discuss in further detail the various ways in which the public sector can help deal with system-wide liquidity problems, including by providing some form of liquidity insurance or other services to the private sector.


Of course, if the cost of funding was reduced for an institution that held more liquid assets, the effect would be less than outlined here. [8]

See Grenville (1991) for a discussion of these changes. [9]