Subcommittee on Securities and Investment

Hearing on "The Implimentation and Future of Decimalized Markets."

Chairman Enzi, Senator Dodd and Members of the Senate Banking Securities and Investment Subcommittee, thank you for this opportunity to testify today about the impact of trading in decimals. My name is Kenny Pasternak. I am Chairman, Chief Executive Officer, President and co-founder of Knight Trading Group, the world's largest wholesale market maker for shares of both Nasdaq and non-Nasdaq equities. We are also a growing options market maker and an asset manager.

At Knight, we like to think of ourselves as providing the processing power behind the explosive growth in securities trading via the Internet. We founded Knight with the purpose of empowering self-directed individual investors - give them the same speed, low cost and dependability in their securities transactions long enjoyed by large institutional investors. We are committed to offering the same efficient, dependable service, whether people wish to trade a single share of stock or a thousand shares-so our interest in the issues under discussion here today is acute.

The Subcommittee has asked me to assess how decimal pricing is affecting the trading environment; whether the anticipated benefits of decimalization are being realized; and whether the transition to decimalized trading has brought about a need for change in securities regulations.

To answer the first of these questions, U.S. stock markets recently adopted decimalized trading in the belief that it would narrow spreads and lead to less costly order execution.

There is no doubt that spreads have fallen. According to preliminary data, quoted spreads for Nasdaq stocks have fallen by an average 51% since the advent of decimalized trading. The decline has been greatest for the most active, lower-priced securities, which saw a 71% decline in quoted spread. Effective spreads-a better measure of investor cost-fell by an average of 46%.

It is important to note that the introduction of decimalized pricing is not in itself responsible for the narrowing of spreads. Rather, spreads have narrowed as a result of the decision by major market centers to reduce their Minimum Price Variation, or MPV. In 1997, when the major market centers dropped the MPV from one-eighth of a dollar (12.5 cents) to one-sixteenth of a dollar (6.25 cents), the narrowing of spreads reduced trading costs for many investors, but increased costs for others. With the MPV now at a penny, the number of investors for whom costs have increased has been magnified. In other words, a higher percentage of orders are being processed less efficiently.

Today's one-penny MPV has reduced price discovery, diminished liquidity and increased the level of trading activity required to execute an entire order. No market participant has an incentive to quote in size, as others can easily co-opt that information and trade ahead by as little as one penny. While such problems were once experienced only by institutional investors, they now affect increasing numbers of individual investors as well.

Knight Trading Group supports decimalization. We believe that by making stock prices easier to understand, decimalization encourages market participation and, therefore, benefits everyone. My purpose today is to focus this Subcommittee's attention on the adverse effects of the dramatic reduction in MPV adopted by the major market centers at the time when they implemented decimal pricing.

The one-penny MPV is, in many circumstances, too small. In my view, the most important issue for this Subcommittee's consideration is how to encourage the markets to arrive at the correct MPV for a given set of circumstances. The thrust of my testimony today will be to explain why this issue is vital to the depth, liquidity and overall strength of our markets.

The major market centers' recent adoption of the one-penny MPV has had enormous unintended consequences. Preliminary data suggests that, even for retail-sized orders, trading outside the spread has increased dramatically. The price discovery process has been impeded because displayed depth on Nasdaq has also declined-to approximately one-third of what it was before the transition to trading in decimals.

Meanwhile, we see that more and more investors feel compelled to break down their buy and sell orders into smaller lots in order to achieve more control over their executions. Trades generally have increased in number but declined substantially in size.

In addition, the one-penny MPV has occasioned a significant decline in the average number of market participants at the inside quote. In other words, those who provide enhanced liquidity to the markets are now committing less capital for the execution of individual investor trades than they did before the one-penny MPV.

Those who proactively provide enhanced liquidity constitute a large and highly competitive industry that is absolutely essential to the efficient operation of the markets. They often provide liquidity automatically in amounts larger than the National Best Bid and Offer (NBBO). This automatic execution feature expands as the MPV increases and shrinks as the MPV decreases.

At present, because of the one-penny MPV, many retail investors are paying more-as the nature of executions changes from fast and complete to slow and fragmented. For example, during the month of January, Knight Securities alone provided liquidity above the NBBO size level in excess of 179 million shares. Without our enhanced liquidity, we estimate conservatively that investors would have paid nearly $4 million more in execution costs.

With the advent of the one-penny MPV, however, we and our competitors are providing enhanced liquidity in many fewer instances. Whereas we previously provided instantaneous automatic execution at the NBBO price for any order of up to 2,000 shares, we now reserve automatic execution for much smaller orders. And in many instances, we do not offer automatic execution no matter how small the order.

This decline in the availability of enhanced liquidity for Nasdaq stocks, coupled with the lower average quoted size, translates into slower and more fragmented executions-and ultimately higher costs for investors.

In the listed markets, what we are seeing is a pronounced increase in the number of ITS "trade-throughs." As you know, ITS-the Intermarket Trading System-is the linkage between the major exchanges and other trading centers. When an ITS participant like a New York Stock Exchange specialist sees another trading center offering a better price for more than a hundred shares, the specialist is obliged to either match that price or else forward the order on the investor's behalf. When the specialist fails to make good on that obligation, the omission is known as a "trade-through." It means that the investor may actually receive a price dis-improvement.

Our subsidiary, Knight Capital Markets (or KCM), which trades New York and American Stock Exchange listed equities on the Nasdaq InterMarket™, has seen a marked increase in the number and frequency of exchange member trade-throughs. During the first six trading days in February 2001, when decimal trading was introduced in all listed stocks, KCM experienced an average of more than 2,500 trade-throughs per day by NYSE members-a more than four-fold increase over the pre-decimalized period.

Before investors can realize the full benefits of decimalization, including cheaper order execution, Congress, the SEC and the major market centers must address the negative consequences of narrowing spreads. Some of the blame can be traced to outdated rules and regulations that were written for an era of fractional pricing and substantially larger MPVs.

One area that needs review is the long-held notion that our NBBO standard provides the most accurate barometer as to whether an investor has received best execution. The concept of best execution remains a cornerstone of our markets, implying that broker/dealers have a duty to seek the most advantageous terms for their customers' transactions.

For many years, the NBBO has been regarded as the ultimate measure of best execution. The NBBO, the consolidated stream of transaction reports and quotations mandated by Congress in 1975, has been available to the public on a real-time basis. Retail investors have come to expect automatic execution for orders up to a thousand shares at the NBBO price or better.

However, automatic executions and guaranteed liquidity are decreasing as liquidity providers become less pro-active. It is time to acknowledge that the NBBO has become less indicative of market liquidity. With the advent of the one-penny MPV, the NBBO has lost its value for all but the smallest orders. It gives no indication as to the price at which many orders can be executed in their entirety and at what speed.

Indeed, the NBBO prices can actually mislead the public with respect to the quality of order executions available among various market centers trading the same issues. The quality of executions afforded investor market orders by various market centers can vary widely in relation to the consolidated NBBO at the time of order receipt. The price at which an order was executed may not reflect the greater liquidity that might have been available at a competing market center.

In short, the NBBO should no longer be regarded as sacrosanct.

In many instances, price should not be the primary consideration in determining best execution. In the past, a broker/dealer could argue persuasively that it had discharged its best execution obligation merely by providing price improvement. Today, however, in an environment of narrowed spreads and diminished liquidity, providing price improvement may be less important than executing an order promptly and in its entirety. Broker/dealers should be obligated to consider such criteria as order size in seeking the destination that will provide the best execution.

Other impediments to fairer, more efficient markets

While the Subcommittee and the SEC are reviewing the rules governing the National Market System, we would encourage them to consider a number of other impediments to fairer, more efficient markets. These include:

• The issue of connectivity, as manifested through market-center linkages.
• The proper means of encouraging market participants to quote and perform price discovery in larger sizes.
• The Short Sale Rule and other outmoded regulations that actually impede the efficient handling of orders. Short-selling by market professionals increases market liquidity, helping to offset temporary imbalances in supply and demand. It also reduces the risk that the price paid by investors will be inflated by temporary supply impediments.
• And finally, the need for a new Trade-Through Disclosure Rule for equities, similar to the rule recently adopted by the SEC for options trading. Such a rule would require broker/dealer disclosure whenever a customer's order is executed at a price that is inferior to a published quote on another exchange. The rule would not supplant the broker/dealer's duty to provide best execution, but it would provide investors with the means to monitor broker/dealer performance, while encouraging broker/dealers to develop effective means of accessing better quotes published by other markets.

Let me conclude by citing the words of recently retired SEC Chairman Levitt:

That same "simple and salient truth" is at the heart of the issues we are discussing today: Whether our domain is Wall Street or Main Street, we are obliged to do what best serves the interests of all investors-large and small.

The implementation of trading in decimals has made the markets more accessible to investors, but it has also led to a number of profound changes that diminish market liquidity, as well as the speed and efficiency with which investor orders are being executed.

Only by recognizing and adapting to the changes occurring in the U.S. markets can we continue to protect the investor, strengthen market integrity and maintain the position of leadership that our markets enjoy globally. Therefore, we hope that Congress and the Commission will review the Securities Exchange Act of 1934-especially its rules regarding the National Market System-then take the necessary steps to create markets that are fairer and more efficient for the largest possible number of trading participants.

As this Subcommittee ponders the significant public-policy issues surrounding decimal pricing, my colleagues and I at Knight would welcome any opportunity to contribute further to the discussion.

Thank you.