"SPEEDING UP SETTLEMENT: THE NEXT FRONTIER" REMARKS BY CHAIRMAN ARTHUR LEVITT UNITED STATES SECURITIES AND EXCHANGE COMMISSION SYMPOSIUM ON RISK REDUCTION IN PAYMENTS, CLEARANCE AND SETTLEMENT SYSTEMS THE PIERRE HOTEL, NEW YORK JANUARY 26, 1996 This conference brings together many people with first hand experience in managing clearance, settlement, and payment system crises. Jerry Corrigan presided at the Federal Reserve Bank of New York during the October 1987 market break and the demise of Drexel Burnham Lambert. John Corzine was part of the team at Goldman Sachs that helped resolve the gridlock that overtook the mortgage- backed securities markets immediately following the insolvency of Drexel. I've had my own experiences with settlement questions -- which is often a euphemism for "settlement crises." As a brokerage firm executive during the 1960s and 1970s, like all my colleagues, I had to meet the multiple challenges of new technology, wildly fluctuating trading volume, and new-found competition following the "unfixing" of commission rates in 1975. We weathered the paperwork crisis, and learned first-hand how important back office systems are to the well-being of both firms and markets. As Chairman of the American Stock Exchange, I worked to apply technology so that investors could trade with confidence, especially during periods of high volume and price volatility, such as the market break of October 1987. And as Chairman of the Securities and Exchange Commission, it has been my goal to have the markets settle their trades faster, in three business days instead of five. One of the first lessons I learned as a brokerage executive was that huge numbers of people and firms are involved in making payments, clearance, and settlement systems work. They rely on a dedicated core of exchanges, clearing agencies, and hundreds of banks, brokerage firms, and companies. Their degree of interdependence can be staggering. But this also means that such systems are only as strong as their weakest link -- so any attempt to strengthen the chain must strengthen every link. Over the years, I've come to believe that payments, clearance, and settlement systems are the real backbone of our securities markets. They must be strong, resilient, and flexible. A flexible settlement system is especially important in this country because of our widespread participation in the market. With more than 51 million individual shareholders, the US has a constant and compelling need to maintain individual investor access to its markets. Our settlement system must accomodate the needs of those investors, not try to force them to conform to the system's needs. Settlement systems also play a huge role in maintaining confidence in the market -- investors want to know that their trades will be completed on time, that they won't lose their funds or securities because of insolvent firms, and that the markets will be open if and when they want to buy or sell. I arrived at the SEC in 1993 with a firm commitment to reduce risk in our markets. The primary tool available to us was to speed up settlement -- indeed, The Bachman Task Force Report, published the year before, had suggested that three-day settlement would reduce the credit and market risks to the intermediaries clearing securities by as much as 58%. This was the prelude to T+3, which required that standard trades in most securities be settled in three business days, as opposed to the previous standard of five. This was the first reduction in settlement times this century and it brings us back to the time frame that was prevalent at the time of the Korean War. Although we did this for many worthy reasons -- to reduce credit exposure in securities markets, to bolster investor confidence, to add greater discipline to the process, and so on -- we faced some opposition from almost every quarter: from brokerage firms, from banks, and of course from individual investors. Whether in town meetings with individual investors, or in conversations with industry executives, I was questioned often about the benefits to be achieved. Investors feared it would cut off their access to the securities markets, while firms thought the costs -- in lost float, or compliance costs -- would be too high. More than 1,500 individuals took the time to write to the Commission and Congress. Let me quote from just a few of the letters we received: * from a gentleman in Texas: "The new rule would seem to be an effort to squeeze the small investor out of the market." * from a couple in California: "You have seen to it to take another freedom away from the small investor." * from a woman in Kentucky: "Just whose bright idea was it to institute the 3-day settlement requirement. . . ? Does everyone in Washington spend their time thinking up ways to make the lives of little people utterly miserable? Or is there some element of collusion with the brokers in the U.S. to force us to turn our securities over to them and to set up money market accounts with them?" * and from a couple in Maryland: "We question the need to penalize the small investors for the sins of the big players." Many investors told us their postal service war stories -- items sent but never received, letters that took weeks to reach their destination -- and asserted that postal delays would inevitably translate into failed settlements. Many objected vehemently to establishing financial or asset management accounts with their brokers and viewed the change as forcing them to engage a host of lawyers, accountants, and professional managers. The problem, of course, was that the benefits to the investor -- increased liquidity, less systemic risk in their markets -- were long-term, while the inconveniences were quite immediate. If I had to do it all over again, I would invest even more effort in communication with the public about the benefits of faster, more reliable settlement -- and about how we all have to do our share to strengthen America's markets. That's a lesson for the future. Despite the criticism, the SEC rode on and implemented T+3. Let me share with you a few observations drawn from settlement statistics that, I think you'll agree, indicate that we made the right decision and that the conversion went reasonably well. Broker-dealers are settling more trades on-time among themselves in three days than they did when settlements could be completed in five days. As you know, the National Securities Clearing Corporation's continuous net settlement system is where most broker-dealers settle exchange or over-the-counter trades with other broker-dealers, including exchange specialists and OTC market makers. In May 1995, before T+3, and with average daily volume running at 726 million shares in NYSE, Amex and Nasdaq securities, NSCC "failures to deliver" were an average of 8.43% of all deliveries. In November 1995, after the T+3 conversion, with average daily volume running at 830 million shares in the same securities, NSCC "failures to deliver" declined to 7.67%. At the same time, broker-dealers appear to be settling about the same percentage of institutional trades in three days as they did in five. The percentage of trades authorized to be settled between institutional investors and their broker-dealers has not changed significantly, either -- it was 89% in November 1995 and 90% in May 1995. Obviously, more can and should be done to speed up the process by which the various agents for an institution -- the investment manager, the brokerage firm, and the securities custodian -- go about arranging for the movement of funds and securities. I have heard anecdotally, however, that T+3 has meant that more people know what their positions are on the trade date and, as a consequence, are less likely to renege on their trades. Fund and investment managers no longer have the luxury of waiting until T+1 or T+2 to book trades "as of" trade date. Similarly, it's now possible for customers executing trades to find out on trade date the net cost of their trades -- something that was not always possible before T+3. In an ideal world, we could reduce risks by shrinking settlement periods further, even as far as same-day settlement. That's the settlement cycle for commercial paper, and faster than the current cycle for trades in US Treasury bonds, bills, and notes. "T+0" would have many advantages: * Faster settlement would further reduce the risk of loss to investors and intermediaries from the insolvency of other market participants. * Same-day settlement would require market participants to "pre- position" their assets and, if the transaction involved an extension of credit, to arrange for that credit at or before the time a trade was placed. This would reduce "fails" by several more percentage points. In such an environment, people trading in the market could be extremely confident that settlement would take place. * Faster trade settlements in equity markets could also help reduce lags between those markets and the derivatives markets -- both exchange and over-the-counter -- thereby reducing the need for credit on intermarket positions, and making credit available for other purposes. We've made considerable progress bridging the time gaps through cross-margin programs -- the Options Clearing Corporation, as well as various futures markets, have taken the lead. We ought to take a fresh look at how these programs work, to see if they can be improved. * Of course, the collapse of Barings Bank and the demise of American firms like Kidder Peabody, E.F. Hutton, and Drexel Burnham Lambert should teach us that regulators and market participants must remain vigilant to concentrated credit and market risks. As market participants develop and use financial products that are difficult to price independently, they must undertake to keep control procedures current and maintain a well-trained staff. * We should not discount the freedom same-day settlement would give market participants to focus their attention on what they are doing today and what they want to do tomorrow, instead of what happened yesterday, or solving problems that arose a few days ago. It could reduce the number of disagreements over trades and make those disagreements easier to resolve. * Finally, same-day settlement offers extraordinary benefits - - and extraordinary challenges -- in an area of special concern to all of us: maintaining the pre-eminent role of our markets in the international arena. The world economy is becoming more integrated. Foreign companies continue to look to US markets for capital and liquidity. 103 foreign companies, from 26 countries, entered the United States public markets for the first time in 1995. At the end of that year, there were more than 738 foreign companies from 45 countries filing reports with the SEC. The registered public offerings of foreign companies amounted to $48.3 billion in 1995. This has increased the choices available to American investors, to their benefit. At the same time, American investors are looking abroad for diversity and growth as never before. Payment and clearing systems must keep up with this growth. If US markets are to respond to this interest in foreign stocks and compete effectively in this area, multi-currency settlement capabilities must become a reality. All aspects of trading non- U.S. stocks in U.S. markets, including clearance and settlement, must be as competitive as other secondary -- if not the primary - - markets for those stocks. This will pose a challenge to existing U.S. payment systems, which today generally do not accommodate non-U.S. currencies. It will also pose challenges to clearing organizations, their members, and market participants, to develop the systems and train their personnel to manage currency risks. If the fine effort at implementing T+3 is any sort of a guide, we should be up to the challenge. I've outlined some of the advantages of speeding up the settlement process. But faster settlement doesn't come without its costs -- and not just dollar costs, but practical difficulties that militate against immediate implementation. I'll focus on just two of them: * First, same-day settlement could limit the ability of individual investors who hold certificates to participate in securities markets. The US has the largest number of individual investors participating in any market. We must be careful not to close the door on, or somehow alienate, those investors who are our most important national asset. Many investors want to be recorded directly with the company as owners instead of holding stocks in "street name" through their broker. Many of these direct registered shareholders also want physical certificates -- as evidence of their ownership, or so they can have the flexibility to choose among brokers when they wish to sell. Our markets draw their strength from being a marketplace for all. This means we must find ways for those direct-registered, buy-and-hold investors to access the markets quickly when they want to buy and sell. This will be challenging, but it can and will be done. Broker-dealers have seen a significant increase in the number of retail accounts and equity holdings for individual investors during the last few years leading up to T+3. Many brokerage firms encouraged their customers to leave their securities in street name because that practice would be more convenient when the investor wanted to sell. To ward off the overzealous, we asked firms not to tell their customers that the SEC rule mandating T+3 required street name holdings. But even as one segment of the investing public is increasing its use of broker-dealers, a growing segment is investing directly with corporate America through direct purchase and dividend reinvestment programs. Estimates indicate that perhaps 6 million shareholders participate in dividend reinvestment programs. We face a challenge in finding ways to connect this segment of the investing public with existing clearance and settlement systems. * A second practical difficulty is that same-day settlement could be expensive for institutional investors and their agents, who must coordinate their actions. As you know, institutional investors typically don't engage the same firms to make investment decisions and to hold their securities portfolios. This means the typical institutional trade involves at least three entities: an investment manager who orders the trade, a broker-dealer who executes the trade, and a securities custodian to whom the broker-dealer delivers securities or funds. The institutional experience with T+3 suggests several links in the chain need to be stronger; for example, settlement instructions were in place on trade date for fewer than 10% of institutional trades submitted to the Depository Trust Company in November 1995 - - the most recent month available. This suggests that much more needs to be done if we are going to shorten the settlement cycle beyond T+3. Some investment managers who serve many clients wait until the end of the day to inform the broker-dealer of the particular fund and securities custodian to whom the broker-dealer must deliver funds or securities. Investment managers who function this way would have difficulty operating in a same-day settlement environment. The answer may well lie in further adaptation of technology to move information, securities, and funds quickly and safely. The brokerage industry, the Depository Trust Company, and service vendors, such as Thomson Financial Services, are exploring ways to streamline the communication and information management process. Firms are researching ways to integrate their trader support systems with their clearance and settlement systems. Firms are also building standard communication protocols and open electronic communication systems. Today, for the first time, these things and more are within the realm of possibility. Direct registration, home banking, consolidated asset management, and cybercash are more than just buzz-words for technophiles. These products are being tested and perfected in the marketplace. The challenge now is to make these and other electronic systems reliable through appropriate back-up facilities and security procedures, so that data cannot be corrupted, misused, or misappropriated. If the last 20 years are any indication, security questions will be resolved, and technology will indeed help us surmount some of the key obstacles to even faster settlement. But only if our legal and regulatory structure allows it. There's no way around it: speedier settlement will require greater integration and coordination not just among financial firms, but among their regulators as well. We will need to make a concerted effort to focus less on turf, and more on common ground. As a regulator and a participant in the President's Working Group on Financial Markets, I can vouch for the effective coordination that has developed among the Commission, the Federal Reserve Board, the Commodities Futures Trading Commission, the Treasury Department, and other interested federal bank regulators. It doesn't take a crisis to bring us together -- we and our staffs converse regularly on market and interagency concerns. The Working Group provides a basis for coordination. I've cited some of the arguments in favor of faster settlement, as well as some of the obstacles. While I'm not sure how long it will take to overcome the obstacles, let me leave you with three steps we can take to lay the foundation for even faster and safer trading in the next millennium: * First, we must finish implementing conversion to same-day funds settlement. Starting in late February 1996, the securities clearing agencies are planning to stop using checks in settling payment obligations with their members. This means trades will be final on "T+3", not the day after, and the clearing agencies and their members will have to have the collateral and credit facilities lined up to meet their expected payment obligations. It will increase the predictability of settlement, even in high volume, high volatility markets. * Second, efficient systems must be built to link direct registration and securities depositories. Individual investors should not be forced to choose between registration with an issuer as a stockholder, and easy access to securities markets through their broker of choice. If assets must be pre- positioned, then we should find a way to make the issuer's share registry one of the locations from which investors can quickly meet their delivery obligations. I call on brokers, transfer agents, and corporate secretaries and their representative organizations to make this a reality, sooner rather than later. * Finally, we must improve the technology and process flows for institutional trading and asset management. We must streamline -- to one -- the number of times data is entered before a trade is settled. We must make sure that the information is secure and promptly communicated to all parties who need to know it. And we must ascertain that the technology contains sufficient safeguards to minimize operational failures in the markets. System reliability and security will become even more important as settlement times shorten -- there will be less room for error, and for breaches of security that could shake investor confidence in our markets. The Commission has been focusing on the integrity and reliability of computer systems at the exchanges and clearing agencies for quite a few years now. We've emphasized the need to obtain regular, rigorous, and independent assessments of computer systems. To date our focus has been on the market and clearing organization system. However, firms and service providers must also address these issues, both for obvious competitive reasons as well as for the sake of market integrity. The three steps I've outlined are not a panacea -- they will not lead us to the Promised Land of perfect settlement anywhere on the globe, anytime, in any currency. But together, they constitute the next horizon along the way. They are logical and realistic measures that will bring us closer to where I think we all want to be -- where we all know we ought to be. It's said that even the longest journey begins with a single step. When it comes to settlement, I hope that this symposium will be that step, and that years from now, we'll be able to look back on these discussions and say that we made a difference. # # #