"So, you plan on computerizing the decision- making for Listed stocks?" he said, with a twinkle in his eye. "OK, from now on, you're The Angel of Death." And so, for the next several years, he happily yelled "Angel of Death, how are ya..." whenever he saw me, across the trading floor at 390 Greenwich Street.

This story has its roots in 1999, when I was asked to leave the global portfolio trading group that I had started at Salomon Brothers, in order to become the "global architect" of a new trading system, for the merged firm of Salomon Smith Barney. Our focus for the next several years was on building a platform to automate much of the institutional trading process. The system traded retail orders automatically, categorized and routed inbound institutional orders to either algorithms or traders, provided tools for traders to work large orders and positions, and centralized the desk's risk management.

By late 2003 our new system was handling most inbound order flow and was used for all trading by the NASDAQ trading desk. However, traders for NYSE stocks were not able to use it effectively. Why, you ask? Well, at the time, NYSE rules still made it problematic for members to trade listed stocks away from the NYSE and liquidity was still highly concentrated on the physical floor of the exchange. There was a certain "club vibe" to the floor and despite all of the advancements in technology, "The Specialist" still had to manually approve all trades.

Despite having a good understanding of the obstacles, I knew I needed to delve deeper. So, I scheduled meetings on the floor of the NYSE to try to understand what further automation was possible. Following a demonstration of the new Salomon Smith Barney handheld computer system, I was shown around the floor by the head of our NYSE floor business, Vince. He introduced me to several Specialists, who were all quite friendly and happy to answer most of my questions. The notable exception was the Specialist in General Electric, who was basically glued to his keyboard. He had to constantly click "enter" to approve trades from the electronic "DOT" system and did not have the time to talk. (Note, GE, at the time, was the most active stock and even though users of DOT thought of it as a fully electronic system, it still required Specialist intervention to match trades). After about an hour walking around the floor, we went back to the Salmon Smith Barney booth. I briefly described our system for trading NASDAQ stocks to Vince and he asked a lot of smart questions.

When we settled into the booth, Vince and I had the following conversation:

(One of the very rare occasions in my life I was actually speechless - I just looked at him with a horrified expression…)

As it turned out, Regulation NMS was proposed, which ushered in true competition to the floor with predictable results. Our system became able to trade NYSE listed stocks, the aggregate volume in NYSE shares increased, execution costs for most investors declined, and market share for the NYSE declined precipitously. Floor based business models were severely impacted and the value of specialist firms declined dramatically. While this was somewhat offset by gains in competing exchanges and firms, the overall effect was a decline in spreads and in frictional costs to investors.

(As a postscript, I want to point out that Vince worked extremely hard to place many of the floor staff from his team in other jobs. He was ahead of the game when the inevitable headcount reductions started. His foresight ended up benefiting many of those people…)

The purpose of telling this particular story is to highlight many of the considerations that policymakers should have when evaluating the so called "Grand Bargain" proposed by the NYSE. The "Grand Bargain" proposal is nothing more than an old fashioned "horse trade". The dominant exchange is offering cost cuts to the largest banks in exchange for their support for anti- competitive regulations to help exchanges regain market share. This proposal is not an exchange between the NYSE and its large customers; in fact, it is likely to make money for both…

While the NYSE offer sounds like they will make less money on trading, it is not necessarily the case. Exchanges only make money on the difference between the rebates that they pay to one side of the trade and the fees they charge to the other. Thus, it is entirely possible that the NYSE will make as much or money per trade under the proposal. It also needs to be said that capping the fees will restrict competitors from using innovative fee structures to compete and the reduction in rebates will directly penalize natural providers of liquidity.

The other part of the "Grand Bargain" is the proposal that a universal "trade at" rule be adopted. The NYSE is heavily politically invested in this proposal as they seem to view it as a key way to regain lost market share. Such a rule would severely limit non-exchange competitors such as ATSs and broker dealer "dark pools". While there are lots of opinions about "dark pools" and ATSs, there is no escaping the fact that these competitors have driven innovation and led to lower costs for investors. Had a "trade at" regime existed for the past 10 years, the BATS exchange, whose markets started as non-exchange ECNs, would likely not have been started and innovative markets such as Bids Trading, PDQ ATS, and, of course IEX, might never have gotten off the ground. It is, therefore, not surprising that the CEO of BATS has recently penned a letter to the SEC arguing against the "trade-at" proposal. The question is, why large banks, that own the largest non-exchange venues, would be interested in this bargain?

The answer, of course, is economics.

The majority of the largest banks on Wall Street have trading strategies that, on balance, demand more liquidity than they provide. Since trades that demand liquidity are subject to access fees, a dramatic decrease in those fees will significantly improve their profit margins. This proposal would also decrease the value of operating their ATSs, as the relative difference between accessing liquidity on exchange and on their ATS would decline. While it is unclear how many banks would shut down their ATSs in this environment, it is fairly certain that the gain from lower fees would likely offset a decline in those businesses.

Considering that exchanges and large firms would benefit from the anti- competitive rule, it should be incumbent upon policy makers to ask; whom actually provides liquidity and whom absorbs the negative impact of the "Grand Bargain"?

The answer is that there are two main groups: retail investors and leading electronic market makers.

Both retail investors and market makers are beneficiaries of the current rebate. Retail investors may not directly benefit from the rebate structure, but their brokers often do. This money is typically used to either keep commission rates low or to subsidize the innovative products and services that brokers provide to their customers. It's reasonable to assume that if these payments were eliminated, investors would end up paying more for what they currently receive. Whether this would lead to less trading, is impossible to predict. The fact that it would be a wealth transfer from retail to the largest banks and exchanges, is certain.

The effect on market makers is less obvious and deserves a longer discussion. What is clear, however, is that market making in less liquid stocks will get more difficult when there is a lower rebate paid for liquidity provision. The predictable result of lower rebates will be wider and more volatile bid offer spreads in these stocks on exchanges. The "trade at" part of the proposal could compound this effect by restricting the ability of traders to leverage competitive venues. While not easily quantifiable, it is certainly likely that the result of this proposal could be lower liquidity and increased trading costs for less liquid securities.

Instead of the "Grand Bargain" being proposed, we should learn from our past experience and consider rules that are pro- competition. In the recent BATS letter, they proposed to tier access fees based on the trading characteristics of stocks. I agree with them that market structure conversations should focus on reforming the current "one size fits all" policy. For many actively traded stocks, one cent is artificially wide and .3 cents is too high a fee. In those cases, it is probably good policy to have lower fees, but also to have a lower tick size. Such rulemaking would decrease trading costs in the aggregate without providing inherent advantages to any particular market participant. Importantly, it would not make the situation for smaller and less liquid securities worse. (We have already commented on the notion that, for less active stocks, wider tick sizes could make sense, and we may wish to consider raising the fee cap in those cases to incentivize liquidity. In all cases, the guiding principle we should follow is to make rules that encourage competitive innovation instead of restricting it.

David Weisberger, Managing Director, Trading Services at Markit
Posted 5 February 2015

S&P Global provides industry-leading data, software and technology platforms and managed services to tackle some of the most difficult challenges in financial markets. We help our customers better understand complicated markets, reduce risk, operate more efficiently and comply with financial regulation.

This article was published by S&P Global Market Intelligence and not by S&P Global Ratings, which is a separately managed division of S&P Global.