By HAL S. SCOTT
Originally published by the Washington Times on January 21, 2016
Contrary to the views of Michael Lewis and other critics, America’s equity markets are not “rigged.” U.S. investors are actually much better off in today’s high-speed automated marketplace than they were in the old, largely floor-based markets when the NYSE and NASDAQ operated as virtual monopolies.
Let’s take a look at the facts. The most important characteristic of strong equity markets is that they provide a fair marketplace for investors to buy and sell stock at the lowest transaction cost possible.
For retail investors, a widely used measure of transaction cost is the bid-ask spread, which is the difference between the best bid to buy a stock and best offer to sell it. Since 2007, the average spread has dropped in half to about 1.5 cents.
Because spreads are so low, the only meaningful cost to a retail investor for buying and selling most stocks is the commission charged by their broker. These commissions have dropped dramatically in recent years; the average commission charged by the three major retail broker-dealers is now approximately $10 per trade — many multiples lower than in the old markets.
Measuring the cost of executing an institutional order, like for a large mutual fund or hedge fund, is somewhat more complicated, because institutional investors must minimize the “price impact” of their large orders by executing them in large blocks off exchanges or by breaking up their large orders into many smaller orders.
Vanguard estimates that the shift from the old market structure to today’s automated market structure has reduced trading costs for institutional investors by 35-60 percent. They quantify the impact of reduced transaction costs on long-term investors, finding that $10,000 invested in a mutual fund over 30 years would now yield a long-term investor $132,000 instead of $100,000. In other words, strong equity markets make a big difference for Mom and Pop.
Now, this is not to say the current market structure is perfect. As Michael Lewis and others have noted, exchanges sell costly access to high-speed market data and execution services to those that are able and willing to pay for it. However, they fail to note that close to 90 percent of investors access these high-speed services. Retail investors access these services through brokers that are willing to pay these costs so their customers get better prices.
Nonetheless, exchanges retain monopoly-pricing power over these services, because those seeking to access the markets at the highest speeds must pay for them. Although the Securities and Exchange Commission regulates these practices and imposes explicit limitations on their cost, the SEC has allowed exchanges to regularly increase the fees that they charge for these services. The SEC should consider ways to eliminate these monopolies. For example, the SEC could eliminate its prohibition on competition over the provision of consolidated market data.
Michael Lewis and others believe that the trading venue IEX, which just applied with the SEC for approval to become an exchange, should be granted hero status because it would level the playing field for investors. More specifically, IEX would not seek to profit from selling high-speed services and would slow down certain high-speed traders. But these critics fail to point out that IEX would, as an exchange, get monopoly prices for their data from the consolidated market data feeds.
IEX’s exchange application poses several other concerns. First, the 350-microsecond delay that IEX would apply to all traders does not change the fact that the traders located closest to IEX would still be able to trade on IEX faster than others. Additionally, the speed bump would not apply to IEX’s affiliated broker-dealer. IEX, and the customers of its broker-dealer, would therefore have a distinct speed advantage to other exchanges and broker-dealers routing orders for investors.
Second, IEX’s 350-microsecond speed bump directly contravenes SEC rules that prohibit exchanges from imposing intentional delays on orders. The reason for this prohibition is that any sanctioned delay could result in an investor missing a better price at another exchange. This is bad for investors.
The SEC should not address the speed bump issue by acting on an exchange application. This issue is a matter of general concern and should be dealt with in a formal SEC rule. If it accepts this speed bump without a rule then the SEC is raising the question — what’s next? Other exchanges with much longer delays or completely different methods of trading?
The SEC should propose a rule that would clarify what types of exchange execution delays and/or processes are acceptable and why. Otherwise they risk throwing a wrench into an automated system that is providing investors with lower transaction costs than ever before for some uncertain benefit.
Hal Scott is professor of international financial systems at Harvard Law School and director of the Committee on Capital Markets Regulation.