Pisani: Stock Trading Regulations Need Improvement

A number of stocks have been halted this summer due to erroneous trades

Beefing up market surveillance and creating a better audit trail will increase the trust of all investors

The U.S. Securities and Exchange Commission seal hangs on the facade of its building in Washington, DC.
Getty Images
The U.S. Securities and Exchange Commission seal hangs on the facade of its building in Washington, DC.

The SEC has its plate full with financial regulatory reform regulations, but the market structure it has created around Reg NMS needs some serious tweaking.

Reg NMS stands for Regulation National Market System. It is a set of rules enacted in 2005 to modernize and strengthen the regulation of US stock markets.

First, the SEC needs to have an honest debate on "how fast is fast enough?" We have already decided that under some circumstances we want to slow trading down: we have circuit breakers in place.

Do we want to go further? Do we really want an arms race to see if we can make trades in under a microsecond (a millionth of a second)? Should we adopt some limit on how fast trades can be made, a minimum quote duration of, say, one second?

If the answer is no, then we need more robust hardware and software that can handle the rush of orders; we can't have systems that lag behind in reporting trades, even for a microsecond.

Some other suggestions:

  • Extend stock circuit breakers to cover the entire equity universe. Some have advocated "limit up, limit down" rules that futures exchanges have used for years, so if a stock drops 10 percent, trading is restricted for a certain period, say five minutes.

The difference between this and a trading halt is that with a "limit up, limit down" rule you can't make an offer 10 percent below the price, but you can make an offer above that price. So trading is not really halted. Then add windows, so you can't move more than 10 percent in five minutes, then after five minutes you can trade another 10 percent.

  • Don't allow "clearly erroneous trades" to print. Once a trade prints, it becomes a nightmare making everyone whole. FINRA and the exchanges are considering clearer rules for breaking trades, which would require that trades only be broken once the stock has been halted and if it trades at a certain percent away from the price at which it was halted.

Fine-tuning the circuit breakers would also help. A number of stocks have been halted this summer due to clearly erroneous trades. NASDAQ has already proposed that if a single print falls outside the current trading range (the National Best Bid and Offer, or NBBO) they will not halt trading until three trades occur outside the NBBO. This is a sensible idea, providing it is adopted uniformly by all exchanges.

  • More data on who is trading, what, and when. The SEC has been greatly hampered in its investigation of the flash crash by the humiliating reality that it had no readily available access to trade data. They have had to go begging to multiple market sources with requests for information.

Beefing up market surveillance and creating a better audit trail will increase the trust of all investors. The SEC has proposed a Consolidated Audit Trail System to track information on trading orders and is reviewing comments recently submitted by the public.

More information on what types of strategies are employed by high-frequency traders is also appropriate. HFTs do not need to divulge their algorithms, but we cannot guard against the small percentage of firms that may be using abusive trading strategies if we have no idea what anyone is doing.

  • More information on the effect of dark pools and broker-dealer internalization and a closer look at "sub-pennying." About 25 percent of stock volume was executed in these "dark" markets in 2009, according to the SEC, and is likely higher now.

These venues do not display their bids and offers to the outside world (the "lit market").

It's not clear if removing that much liquidity from the rest of the market disadvantages everyone else. NASDAQ's chief economist, Frank Hatheway, has conducted a study which found that when internalization levels approach 40 percent of a stock's volume, price discovery is impaired.

Dollars and Cents: Those Pennies Really Add Up

There's another issue: sub-pennying.

Brokerage houses cannot accept orders in increments below a penny. However, dark pools and market makers who internalize can quote in sub-penny increments in order to provide "price improvement."

So, for


example, if IBM were trading at $128.00, and you put in an order to buy 100 shares at that price, a market maker might execute your order at, say, $127.9999.

Why does a market maker do this? To be first in line to get executed.

Who gets hurt here? It looks OK, because there is price improvement. But this practice seems designed to discourage other traders from providing bids and offers. Who cares? Less liquidity means wider spreads.

The SEC needs to look at whether all market participants should be held to one level of price improvement: a penny, or some increment.

  • Eliminate stub quotes and raise the requirements for what it means to be a market maker. Market makers should not be able to fulfill their obligations by putting in a phony bid of $0.01.

If you are a market maker and are required to provide liquidity, you should provide some! The SEC has discussed requiring market makers to post bids or offers that are not more than 10 percent away from the last price, but without any size requirements this won't mean too much.

  • Promote deep order books. Exchanges for years have opposed the idea of a Consolidated Limit Order Book (CLOB), which would essentially dump all bids and offers into a single "pool." It is time to rexamine this idea. There is usually plenty of liquidity, but crazy order handling rules sometimes allows absurd trades to print, far away from the best bid and offer.

The best bid and offer should always be available, regardless of what exchange you are on. If the industry is opposed to a CLOB \(they are\), there should at least be access to a deeper level of liquidity.

  • Raise the barriers to entry for exchanges. In theory, there is no reason you could not have 100 different exchanges, but in reality when you have 100 different pricing mechanisms, it would be tough for the market participants to figure out the most efficient way to route orders, let alone the difficulties regulating all these exchanges.

Finally, here's one suggestion that the SEC and the investing public should consider immediately:

  • Educate the public to stop using market orders under most circumstances! People should think, what's the lowest price I really would sell this stock at? Nobody would sell a $40 stock for a penny, but with a market order that is what could happen—and vice versa! With a buy order, you could buy a stock up to infinity!

Many traders have proposed that everyone should use limit orders. This is a step in the right direction, but in a panic you can lose your shirt with a limit order too.

Here's an example: You own IBM at $128, and you have a limit order to sell at, say, $100 at the market. The problem with a market order is that in a panic it could get executed at $80 or worse. It's unlikely you wanted to sell at that price, particularly if it bounces right back to $128 an hour later.

One solution: A rule that the order would only be executed if the price was, say, within 20 percent up or down from the current price. So in the case above, the order would not be executed below $80 (20 percent below $100), even if you had a standing order. It would get kicked back to you, to ask if you really want to do this.

(Look for our "Man Vs. Machine" special reports Tuesday-Thursday, September 14-16, on "Squawk on the Street", 9am-11am ET and "Closing Bell," 3pm-5pm ET on CNBC.)