What lessons can we draw from September 11 to help us safeguard the financial system against possible future disasters?
THE EVENTS of September 11 posed unprecedented challenges to the operation of the dollar payment system in New York. In its broadest definition, the payment system is a network of banks, central banks, securities firms, service providers, and industry-owned utilities that ensures both the smooth transfer of funds for business and financial transactions and the timely settlement of securities transactions. At its core are the two principal wholesale payment systems, Fedwire and CHIPS (the Clearing House Inter-Bank Payments System), their bank participants, the government securities settlement systems operated by the U.S. Federal Reserve System and two principal clearing banks, and the clearing and settlement organizations for corporate and other securities. For its liquidity and risk management, as well as its transaction flow, the payment system depends on well-functioning financial markets and an extensive communications network.
The World Trade Center, one of the targets of the September 11 attacks, housed several wholesale brokers and large trading operations that play essential roles in the financial markets. Some firms suffered devastating losses of employees. In addition, the communications network and the primary operations of many payment services providers near the World Trade Center were disrupted as fire, debris, and water destroyed or damaged much of the power, telecommunications, and transportation infrastructure serving lower Manhattan’s financial district. Workers in the vicinity of the World Trade Center were forced to evacuate their buildings immediately after the attacks occurred. Others followed as the day progressed, as city officials limited access to lower Manhattan to rescue, emergency, and security personnel.
Given the scale of death and destruction, the dollar payment system performed well in the wake of the attack. Fedwire and CHIPS operated throughout September 11 and the days that followed, in part because significant parts of their operations were not located in downtown New York. Crucial elements of the clearing and settlement system for U.S. government securities, especially the Bank of New York, labored under severe difficulties for some time, but trading in government securities was able to resume on September 13. Redemptions and rollovers of commercial paper were temporarily disrupted, but equities trades made in the days just before September 11 were fully processed and settled. By September 17, when the major U.S. stock exchanges reopened, payments and securities settlements were able to accommodate the largest volume of trading that had ever occurred in a single day in New York Stock Exchange history.
The ability of the largely electronic U.S. wholesale payment system to withstand the direct and indirect damage of September 11 is especially noteworthy in light of the system’s complex structure. How was the payment system able to weather and recover from such a massive attack? September 11 and earlier disasters like fires, storms, and power losses have revealed that there are three crucial elements in preventing gridlock in the financial system.
Contingency planning and built-in redundancy in operational capacity have long been important elements of business strategy at financial institutions. Two key goals are the ability to promptly resume business activity and the safeguarding of transactions and financial information.
While traditional contingency planning seeks to minimize the time between a disruption of business operations and their resumption, avoiding even a momentary interruption of service is seen as crucial in major payment and settlement systems. Providers of payment services locate contingency sites on electrical and communications grids that are separate from the primary processing site. They record transaction data simultaneously at a second site, and the availability of real-time data strengthens their ability to shift processing flows from one site to another on short notice. Key risk-management and accounting systems that complement the systems that process transactions are designed to be available in a contingency event.
Contingency planning advanced substantially in the runup to the century date change (Y2K). Y2K planning incorporated the involvement of senior management and boards of directors and asked business managers to consider how they would operate their businesses (and not just their back offices) in the event of a Y2K-related disruption. In a letter to financial institutions in March 2000, federal bank regulators noted that detailed contingency plans developed for Y2K specified the minimum level of output and services necessary for each major business process. They also noted that simulated operational failures and scenario building helped reduce the time needed to respond to operational problems and improved decision making and internal communication.
External communication was a special focus of Y2K planning and included maintaining lists of contact numbers for financial institutions, regulators, and key infrastructure providers.
But September 11 far exceeded the scenarios considered in traditional contingency planning. Certainly, the preparation of backup sites, the testing of contingency plans, and simulations of disasters enabled many firms to grapple with the extraordinary circumstances of the terrorist attacks. Nonetheless, the September 11 disaster required a substantial reordering of the priorities of traditional contingency plans. The first priority became the safety of people, both those directly affected by the World Trade Center’s destruction and New Yorkers more generally. Access to lower Manhattan was severely restricted to facilitate the rescue and recovery effort and to prevent injury. Traffic was curtailed in the city’s tunnels and on its roads and bridges.
As a result, a large number of institutions lost access to their premises. They activated their contingency plans and needed to transport contingency staff to backup sites. Normal business operations were broken off suddenly, creating reconciliation problems, especially in the government securities market. Businesses had little opportunity to make the preparations they would normally make during another type of disaster, such as a storm or a flood, and their employees found it difficult to travel to contingency locations.
The loss of access to premises highlighted the value of having contingency sites far enough away from the area of destruction so that it was possible for financial institutions to continue to do business in safe surroundings. “Hot” sites that continuously replicated transactions at the primary site proved especially valuable. Travel restrictions, however, created a difficult hurdle to staffing contingency sites. A few institutions found that having managers who were knowledgeable and able to assume control of the business away from their primary sites in lower Manhattan—in the United States or in an overseas financial center—was enormously helpful.
“Given the scale of death and destruction, the dollar payment system performed well in the wake of the attack.”
The extensive and largely unprecedented damage to communications channels further frustrated efforts to sustain payment activity. The communications problems in the first few days after September 11 reflected not only the loss of a major central switching station and a portion of the cellular phone network in lower Manhattan but also a surge in the volume of calls. Moreover, the rapid accretion and unsystematic routing of telephone lines over decades of service, as well as the shared use of lines by competitive telephone companies, meant that many telephone customers who believed they had redundant telephone service found they did not. Even some firms distant from lower Manhattan discovered their service was disrupted in the wake of September 11. While the cellular network was largely restored with temporary transmission towers, restoring conventional telephone service was complicated by the sheer size of the task and the complexity of the telephone system.
Once staff was in place at contingency sites, the reestablishment of voice and data communications linkages was necessary to keep the payment system operational. The large-scale relocation of firms and the disruption of communications on September 11 made the process of simply locating counterparts at other financial institutions difficult. In the early stages, industry associations facilitated the distribution of updated contact information.
Connectivity was another major issue. The simultaneous activation of multiple individual contingency plans meant that, for the first time, backup sites needed to connect to other backup sites. However, traditional contingency plans envisioned the incapacitation of only one or a few firms, so communications testing had been limited largely to those between a firm’s primary and contingency sites and between a firm’s contingency sites and payment services providers. On September 11, the untested connections between one firm’s backup site and another’s presented a new set of problems to be worked through.
The payment system also is a complex liquidity-management system. Trillions of dollars in transactions are processed each day, yet the base of liquidity used to facilitate these payments is a small fraction of that volume. In large part, major liquidity needs are economized through the timing of electronic payments—the receipt of funds from one source finances the payment of funds to another. Temporary imbalances in the flow of payments can be accommodated in some systems, such as Fedwire, through intraday credit.
An operational blockage in the payment system disturbs this finely calibrated balance. Some market participants end up with excess liquid funds, while others experience shortages. A blockage can trap liquidity in one corner of the payment system, and the disruption can quickly spill across the entire financial system.
Large imbalances, especially in government securities settlement, developed on September 11. Fedwire transaction values and volumes reflected the extent of disruption to processing in the U.S. payment system. Transaction values on Fedwire fell 24 percent below their “normal” level (measured by the August average) on September 11, but then recovered in the next day. The volume of Fedwire transactions initially dropped more sharply below August average levels, some 40 percent on September 11, and recovered gradually over the balance of the week to near-normal levels. The more pronounced volume drop reflected both disruption and the actions market participants took to consolidate payments and reduce operational burdens.
Liquidity imbalances persisted throughout the week beginning September 11 as a result of connectivity and other problems. Normally, the management of liquidity risk by individual financial institutions allows for the possibility of unexpected shortages of funds. In the days after September 11, many financial institutions and their customers drew down liquidity reserves or called on correspondent banks to help complete payments and settlements.
But the imbalances proved to be extraordinarily large, and the ability to work around operational difficulties or to tap alternative sources of liquidity was highly constrained. The strong financial condition of U.S. banks and securities firms made it clear that the imbalances reflected only a liquidity problem, not credit strains. To meet liquidity needs, the Federal Reserve injected massive amounts into the U.S. financial system through discount window loans and open market operations. For example, on September 12, the Federal Reserve had $46 billion outstanding in discount window loans and arranged $38 billion in open market operations. By September 14, discount window borrowing had dropped to low levels, but open market operations arranged that day reached their peak of $81 billion.
The Federal Reserve also made other adjustments to facilitate liquidity, such as suspending charges for intraday overdrafts and penalty fees for overnight overdrafts and relaxing rules on the volume of securities it would lend from its portfolio to the market. To cope with shortages of dollar liquidity outside the United States that could not be met through the correspondent banking network, the Federal Reserve entered into temporary swap arrangements with the European Central Bank (ECB) and the Bank of England, and temporarily augmented its existing arrangement with the Bank of Canada. The ECB drew on its swap line in the three days following September 11, with the amount outstanding peaking at $19.5 billion on September 13.
As market conditions began to return to normal in the following week, the Federal Reserve was able to sharply reduce the amount of liquidity it temporarily provided. As processing backlogs were worked down, the value of Fedwire transactions rose sharply on September 13 and 14 and stayed high in the week of September 17. By September 21, both discount window and outstanding repurchase agreements had returned to near-normal levels.
Collective problem solving
Every crisis produces unexpected problems. The ability to solve them is strengthened by the ongoing dialogues among financial institutions, among regulators, and between the two groups. In the U.S. markets, industry associations, such as the Bond Market Association, and committees, such as the Foreign Exchange Committee and the Payments Risk Committee, both sponsored by the Federal Reserve Bank of New York, are key forums for dialogue. These industry groups traditionally have played an important role in identifying and addressing issues related to market practice and supporting infrastructure. The contacts and interaction developed in normal times have proved to be invaluable in times of crisis.
In the days after September 11, meetings and industry association efforts were crucial in the sharing of information and in assuring open communication lines. Industry officials decided when to reopen the financial markets after meeting with government officials to discuss the options. Industry associations organized conference calls to provide market-wide status reports and identify emerging problems. For example, the Bond Market Association sought to ease strains in the settlement of government securities by recommending that firms temporarily extend the normal settlement period from one day to five days, a recommendation that was widely adopted by market participants. A longer settlement period provided more time to work through operational problems, even though it meant a temporary doubling up of processing loads when the market reverted to one-day settlement.
A new level of planning
Today’s management of payment risk reflects, in substantial part, all that we have learned from experience. The lessons learned from September 11 will no doubt take contingency and liquidity planning to a new level, to ensure continuity of business and financial operations in the event of a crisis of similar severity. Within a week or so of the destruction of the World Trade Center, efforts began in the public and private sectors to examine how financial institutions and the payment system responded to the crisis, with a special focus on the specific problems confronted on September 11. These efforts involve such issues as the placement of contingency sites and key personnel, the methods available for reducing vulnerability to massive disruption to communications systems, and the necessary degree of redundancy to ensure that payment systems continue to operate at an adequate level even in a severe crisis.
The events of September 11 have made us all more aware of the vulnerabilities in modern payment systems. In the short run, reducing vulnerability invariably involves costs. A key question for financial institutions and the financial industry in the near term will be how best to balance costs and risks in reducing vulnerabilities.
The rapid advances in technology and risk-management practices offer hope that the balance of costs and risks can improve in the medium term. The global payment system was on the verge of substantial change even before September 11, although much of that change is still in the realm of ideas and has yet to be realized. This suggests that the time is opportune to consider how the lessons of September 11 can be integrated with proposals for more seamless processing, easier transfer or use of collateral and securities across systems, and the adoption of sound practices in payments and securities settlement globally. With the lessons of September 11 in mind, the private and public sectors can work together to create a payment system that is stronger and more resilient.
Christine M. Cumming is the Executive Vice President and Director of Research of the Federal Reserve Bank of New York. The views expressed in this article are the author’s and do not represent the views of the Federal Reserve Bank of New York or the Federal Reserve System.
TWO BOOKS FROM THE WORLD BANK ON KNOWLEDGE FOR DEVELOPMENT
China and the Knowledge Economy: Seizing the 21st Century analyzes and makes policy recommendations for China to build its knowledge-based economy by updating the economic and institutional regime, upgrading education and learning, and building information infrastructure. Also recommends diffusing new technologies throughout the economy, improving the research and development system, and exploiting global knowledge. This book was prepared by the World Bank in collaboration with the Chinese government and leading academics.
World Bank Institute Development Study. 2001. Stock no. A15005 (ISBN 0-8213-5005-6). $25
Korea and the Knowledge-based Economy: Making the Transition provides a framework for analyzing a range of policy options in education, information infrastructure, and innovation systems. The study stresses the need for better coordination among the government, private sector, and civil society to enhance competitiveness and advance economic and social development. Based on a joint study by the World Bank and the Organisation for Economic Co-operation and Development (OECD), this book breaks new ground in its attempt to develop a comprehensive set of national policy responses to the knowledge revolution.
World Bank Institute Development Study. 2000. Stock no. A14881 (ISBN 0-8213-4881-7). $30.
World Bank Publications
Visit our website at www.worldbank.org/publications
To order, contact: World Bank Publications
P.O. Box 960, Herndon, VA 20172-0960, USA
Telephone: 703-661-1580 or 800-645-7247 Fax: 703-661-1501