Balance sheet liquidity management — the task of maintaining sufficient cash to fund ongoing operations — is a key determinant of bank survival in times of stress. During the banking recession of the late 1980s and the early 1990s, many large and small banks failed or were forced into mergers because they ran out of funds to support their balance sheets. Despite the past decade’s financial and risk management innovations, many banks are no more prepared than 10 years ago. We have several reasons for concern.

• Banking industry liquidity has evaporated.

Strong economic growth has rapidly expanded bank and thrift loan portfolios. Together, banks and thrifts grew their loan portfolios at a compound annual rate of 13% from 1989 to 1994, and 19% from 1994 to 1999. Note that this growth is beyond new loans funded by securitizations. During this time, traditional sources of stable funding in the form of consumer and commercial business deposits have only grown by abut 4% a year.

The banking industry has responded to this growth imbalance by funding new loans with wholesale funds and tapping into the liquidity embedded in their balance sheets. Among the most popular sources of liquidity are FHLB advances, money market assets, and investments available for repurchase agreements and for sale. Unused lines in the interbank market are an oft-identified secondary source.

These funding changes have resulted in bank liquidity deterioration. According to a traditional measure, the loan to deposit ratio, liquidity is at its lowest level ever.

A more advanced measure of excess liquidity is “contingency liquidity,” which measures the available sources of funds should normal access to funding markets be disrupted. Not only has contingency liquidity been shrinking, it has become negative. The commercial bank contingency liquidity ratio has decreased from positive 8% overall to negative 2% between 1992 and the third quarter of 2000. This represents a downward shift of about $500 billion in available liquidity.

• Bankers are surprisingly unaware.

In a recent survey we conducted, 67% of financial and risk managers indicated that their institutions did not view liquidity as an issue. Many of these institutions would have difficulty in stressed situations. For example, one large institution did not review liquidity regularly and had developed excessive dependence on one funding market — on many days, this institution was the largest single buyer of volatile funds in that market. Other similar weaknesses: no centralized control of collateral, limited coordination between overnight and term funding desks, and no tracking of pending large cash outflows.

• Liquidity-risk metrics are underdeveloped.

Few institutions measure liquidity with the same determination and sophistication as they apply to their interest rate risk position. The best determine the types and frequency of reporting based on the complexity of their balance sheet. They tend to have daily metrics of their volatile funds position, monthly metrics of balance sheet trends, annual reviews of liquidity strategy and strategic position, and focus on forward views of liquidity.

More than 95% of institutions tracked some measure of liquidity. However, many of these monthly and daily reports measured only one or two dimensions of liquidity (for instance, daily funding needs, liquid asset ratio, loan to deposit ratio). Few, if any, looked at projected needs in either normal or stressed situations.

Unfortunately, some of the largest institutions are the ones most likely to experience problems. They have stretched their liquidity resources the most and become reliant on the “hottest” sources of funds. We examined contingency liquidity ratios for 20 of the top 25 commercial banks and found that half of these banks had a sharply declining trend going back to 1995. Seven banks dropped into double-digit negative ratios in 1999 and only four posted positive figures in all of the last five years.

• Triggers for liquidity crises abound.

As liquidity has drained from the system, financial markets have become less supportive in a crisis. Two related events in 1998 — debt default by Russia and the collapse of Long Term Capital Management — caused market participants to rapidly decrease interbank funds availability and securities purchases. The market’s appetite for new asset-backed securities decreased sharply during the third quarter of 1998, with issuance down by nearly 17%, or roughly $10 billion from the previous quarter.

In a crisis, markets treat potential problem as real problems and cut off substitute liquidity resources. In some cases this occurred even with contractual funding agreements in place. This phenomenon repeats itself in every liquidity-stressed situation — at some point institutions prefer breaching obligations to putting principal at risk.

• Adequate management approaches are rare.

Many of the largest institutions have inappropriate measurement and reporting of liquidity positions, weak contingency liquidity plans, no real stress test measurements, and uncoordinated day-to-day management of important liquidity resources.

How can institutions prepare better? Distinctive liquidity management is built on a comprehensive approach involving: clear guideposts for risk tolerance, forward-looking metrics of liquidity in many dimensions, regularly testing liquidity in situations where problems are most likely to occur, and management plans that increase liquidity preparedness.

Should we enter recession, institutions with appropriate liquidity risk measurement and management will weather the storm, but others may not. As one of our team said after managing a near-death incident of bank illiquidity, “We did not know we had a liquidity crisis approaching, even the day before it hit us.”

Mr. Turner is a vice president and Mr. Sjaastad a consultant for First Manhattan Consulting Group, New York.

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.