I am Robert B. Fagenson, Vice Chairman of Van der Moolen Specialists USA, LLC, a New York Stock Exchange ("NYSE") specialist firm. I appear before you today in my capacity as a specialist and Vice Chairman of The Specialist Association of the NYSE.
On behalf of the Association, I want to express my thanks for this opportunity to testify before the Subcommittee on the effects that "decimalization" -- that is, changing the typical trading increment for securities from sixteenths to pennies -- has had on our securities markets.
In April of 1997 I had the pleasure of testifying before the House Subcommittee on Finance and Hazardous Materials concerning H.R. 1053, "The Common Cents Stock Pricing Act of 1997." That bill, which was never passed, would have required the Securities and Exchange Commission ("SEC") to adopt a rule that would replace trading in fractions with trading in decimals. The bill also would have left it to the SEC to decide the minimum number of cents per share that should characterize bid and offer quotations for stocks and last sale reports for those securities in our markets. The primary purpose of The Common Cents Act and the move to decimal trading was to reduce the "spread" between bid and offer prices for securities. The spread was widely regarded as excessive and as a cost visited on public investors that was not subject to the moderating force of competition. Creating more pricing points in each dollar of stock traded (potentially one hundred compared to the sixteen that characterized traditional trading fractions in our markets), according to theory, would enhance price competition among traders and thereby reduce spreads and lower the costs that they represented. In short, fixed minimum fractional spreads were said to punish investors and reward traders. A beneficial side-effect of eliminating fractional trading increments was supposed to be elimination of "payment for order flow," a practice whereby a market maker or market pays money -- a cent or more per share -- to a broker to, in effect, "buy" that broker's flow of customer orders.(1)
Imagining then that the worst that could happen as a result of a change to decimal trading would be a reduction in the trading increment to as little as a penny, we predicted in 1997 the following:
1. We believed that decimalization and any consequent narrowing of spreads between bid and offer prices would reduce the profitability of payment for order flow and internalization of retail orders in listed stocks. We think that this has occurred, but not to such a degree as to eliminate either payment for order flow or internalization -- both of which we regard as bad practices.
2. We thought that decimalization would tend to reduce the ability of regional stock exchange specialists that trade NYSE stocks to do so. We also think that this is happening.
3. We said that decimalization, especially if the trading increment was reduced to a penny, would incent on- and off-floor market professionals to trade for their own accounts ahead of customers by stepping in front of their orders by a single penny. This, too, has happened. "Stepping ahead" has acquired a bad name in today's markets even though it improves the price for the other side of the trade -- something that those who characterize the practice seem to forget, focusing only on the disappointed would-be buyer or seller and not at all on the "improved" other side of the trade. We observe that "stepping ahead" of (or "pennying") customers almost always seems to involve trading by market professionals, including institutional traders, rather than specialists and market makers, at least in the exchange markets. Market professionals acknowledge that they engage in this practice, even though they dislike it, because they believe that their responsibilities force them to do so in a penny-increment trading environment.
4. We pointed out that decimalization would reduce transparency in our markets by cluttering bid-offer quotation displays with short-lived bids and offers for small amounts of shares that would flicker in and out of existence too rapidly to permit public customers to take advantage of apparent price improvements. Further, we said that the overall effect would be to make the actual prices of stocks more difficult to determine by increasing the number of trades displayed on the consolidated tape at very small price changes. This prediction also has come true -- and to an even greater extent than we anticipated in 1997.
5. Finally, we predicted that, if spreads were significantly reduced, the sizes displayed to the markets in connection with prevailing bid and offer prices also would be reduced. This, too, has occurred, and to a much greater degree than we foresaw in 1997.
Collectively, we think that the most severe and adverse of the effects that have been engendered by decimalization have been these:
1. The significance of the pricing information available to public investors and market professionals -- namely, last sale reports and displayed bid and offer quotations -- has been blurred. This is because, with one hundred pricing points are available for each dollar of stock traded instead of the sixteen that existed when we traded in fractions, stock prices now change much more rapidly, and because the amount of stock available to be bought or sold at any single pricing point is much smaller than was the case when fewer pricing points were available.
2. There has been a loss of predictable and visible liquidity at the prices of displayed bids and offers (in the form of quotes with size). This has happened because interest in buying or selling is spread out among more of the possible pricing points than was the case when there were fewer of them, and because, since only the highest bid and lowest offer in each market is widely disseminated, the degree of liquidity, or the lack of any liquidity, at prices away from the best bid and offer prices is hidden.
3. There now is a trend toward trading in increments as small as one one-hundredth of a cent or $.0001. So far, this phenomenon has been confined to trading in over-the-counter ("OTC") stocks, where visible liquidity is now significantly lower than it was in a fractional trading environment. The adverse effects of such trading have been partially obscured by the fact that the NASD's mechanisms for disseminating consolidated last sale and bid and ask information in OTC stocks are not yet capable of showing prices in increments smaller than a penny. Display of actual four-decimal place prices (or even smaller increments), we believe, would sharply amplify these negative effects.
In our view, the foregoing factors indicate that trading stocks in penny increments may have confused and disadvantaged public investors rather than helped them. Spreads may have been reduced, but investors cannot put those "saved" spread increments in their pockets. Instead they find that, in a penny-trading environment, even modestly sized buy and sell orders are sometimes broken into three or four separate parts before they can be filled -- and that those parts are filled at different prices. In addition to getting different prices, this usually also involves increased clearing costs for the customer. This did not occur when we traded in fractions. Before the change to decimals, orders of modest size did not move market prices away from the levels that prevailed before those orders were entered in the market. These costs, unlike the supposed savings inherent in "reduced spreads," are readily understood by all.
How did we get to where we are from a well-meaning effort to make stock trading more understandable and less expensive for the public? Most investors, we would think, like everyone else, really would rather do away with pennies altogether, rather than be forced to carry and use them -- though we all do. No one at all, however, carries mills or tenths of mills in their pockets or can be expected to think rapidly in terms of such units. Can it be said that a price of 10¼ is harder to understand than a price of $10.2639? We think not.
The Association believes that adjustments can be and are being made to overcome the problems that have attended the movement from sixteen pricing points for each dollar in fractional trading to one hundred such points in a penny trading environment. For example, investors are learning that the smaller sizes associated with each such pricing point displayed as part of a quotation can be and often are quickly exhausted, forcing prices to the next higher or lower level. As a result, investors are finding that it makes sense to place a limit or "cap" on the prices they are willing to pay or accept when they buy or sell and to confer discretion on their brokers to work within the cap to fill their orders. This trading process, however, and the diminished liquidity available at any particular pricing point, may well result in average prices for investors above the lowest offer or below the highest bid displayed at the time their orders were entered in the market.
Under these circumstances, are investors paying more to buy stock or getting less when they sell stock than was the case before the introduction of decimal trading? We have no data to support a conclusion in this regard. At the same time, because sizes associated with displayed bids and offers seem to us to be so much smaller than they were in a fractional trading environment, it is entirely possible that it has become more expensive, on the whole, for ordinary investors to transact than formerly was the case.
Our final concern -- reduction of the trading increment below one penny -- is our most serious one. Reduction of the trading increment below a penny, in our view, could damage the integrity of the markets' pricing function and undermine public confidence in the fairness of our markets. This could occur because trading in tiny, sub-penny increments will even more completely obscure the true state of the market as it is seen by individual and institutional investors alike, as liquidity at particular price points is buried beneath very small size amounts associated with momentarily higher bids or lower offers. In turn, this effect can be expected to increase investor uncertainty as to the price at which a buy or sell order of any substantial size can be executed and to increase buyers' and sellers' anxiety that their intentions will become known to the market before their orders can be filled. Lastly, because of the spray of prices resulting from sales in increments smaller than a penny, confidence that any particular price is "the" price at the moment will be sapped.
Our securities markets are acknowledged by all to be the most powerful engine for the raising of capital ever conceived. Risking the basic pricing and trading mechanisms of that engine and public confidence in them by allowing the uncontrolled splintering of the prices at which stocks trade and in which bids and offers are made is worse than foolish: it is dangerous.
For the foregoing reasons, we urge the Subcommittee to consider legislation that would not only empower the SEC, but require that agency, to determine appropriate trading increments for different categories of stocks (e.g., actively or thinly traded, highly capitalized or less robust, and so on) and to adopt an appropriate rule that would compel the markets and broker-dealers to adhere to those increments in the trading of securities. In 1997, we thought that it would not be necessary to ask for government assistance to deal with the consequences of decimalization. Experience with decimals to date, however, has persuaded us that we were wrong -- and that such assistance is badly needed now. We are unsure whether the SEC has authority today under the Securities Exchange Act of 1934 to adopt particular trading increments. We are confident, however, that, with the help of the securities industry and academicians, the SEC will be able to develop a rational limit on what now is a dangerously uncontrolled process of endlessly splintering the trading increment.
I would be pleased to respond to questions.
1. 1 The SEC considered banning payment for order flow in 1993. Ultimately, however, and wrongly, in our view, the SEC decided to permit it to continue so long as the practice was disclosed to
brokerage customers whose orders were being sold. See SEC Release Nos. 34-33026 (October 6,
1993) and -34902 (October 27, 1994).
Home | Menu | Links | Info | Chairman's Page