What's missing in today's high-speed markets

There's a great deal of commentary circulating through the securities industry about what should be done to improve modern markets. In particular, we're talking about the stock market, although many of these thoughts could translate to other electronic markets. There's hand-wringing over all kinds of issues: spoofing, pennying, layering, gaming, co-locating and theoretical front-running; and conflicts over rebates and payment for order flow, lack of transparency, fragmentation and complexity. Though the benefits of electronic markets are numerous for investors big and small, will the increasing obsession with speed be the system's undoing?

Kevin Winter | Getty Images

As the stock market has become nearly exclusively electronic, participants and observers have become both captivated and driven by speed. For the last 15 years, the ability to trade faster than the next guy has consistently produced new profits. During my tenure at Getco in the early 2000s, each time we improved the speed of our systems, we made more money—and the markets we traded also became more efficient. As we've seen since, however, an unintended consequence of faster and faster markets is that they degenerate into a two-dimensional paradigm. The market has become either "hit-or-take" as algorithms scurry around to find posted orders and attempt to access liquidity across dark and lit markets simultaneously. In this hit-or-take frenzy, a very important component of the market has been pushed aside: broader market information.

Read More'Flash Boys' author Michael Lewis still thinks the market is 'rigged'

In the current paradigm where speed is king, specifics regarding depth of book are overwhelmed by the dynamics of the speed race. Experienced market professionals know that inevitably, however, the speed obsession will correct itself as a function of competitive advantage. We may have already reached the point where incremental dollars invested in speed do not yield a sufficient return, and participants will spend their resources elsewhere — namely, with a renewed focus on information-based execution.

In the days of floor trading, when a stock broker physically carried an order out to the specialist post, he would enter the crowd and ask, "What is the market in GM?" The specialist and the crowd would respond with their prices and size for GM, and based on this information, the broker would use his experience to arrange the best execution of his client's order. Crucially, when he asked "What is the market?" the broker did not reveal his customer's intention, only the name of the stock in question. In a speed-obsessed market there is little opportunity to independently collect and evaluate this information. The current market is a bit like playing "Whack-a-Mole": You don't know when or where orders will pop up, but you keep whacking away until the trade gets done. The ability to assess the overall state of the market, so that the initiating trader can evaluate best outcomes given that broader information, has become nearly impossible in the current electronic environment.

Read More'Flash Boys' claims a 'big lie': Bart Chilton

How, then, can a trader compile this much-needed data? To realize this improved model we need to forgo our speed addiction, which has created a two-dimensional market, and introduce a pause in the trade process. It sounds like heresy in a speed-obsessed world, but a brief pause of 10-20 milliseconds is actually enough time to solicit liquidity from the market, and to reintroduce the concept of "What's the market?" to trading protocol. Similar to the floor broker's request for a market in GM, when soliciting liquidity a trader need only to reveal the symbol to be traded. Competitive market makers and natural liquidity providers respond with their market quotes, and the initiating order can be completed at the best available price. By soliciting liquidity instead of taking only what appears to be available, the initiating order can actually reverse the flow of the Whack-a-Mole game.

Meanwhile, almost all the issues cited above are diminished or largely eliminated. In an auction model created by a pause in execution, there is no opportunity for spoofing, layering, pennying and gaming. Co-location and front-running opportunities disappear because price becomes more important to execution than speed, and by slightly lengthening the pause, orders from different venues can be aggregated, which allows for liquidity consolidation. Other hot-button cost issues such as rebates and payment for order flow can be priced into the auction, so benefits and fees can be fairly disseminated to all parties.

Read MoreDom Chu goes to high-frequency trading school

As technology has improved and our markets have become faster and incredibly efficient, there is a delicious irony in thinking that real improvements in execution could come from a concept central to the floor trading model from years ago: answering the most basic of questions — "What's the market?"

Commentary by D Keith Ross, Jr., the chairman and CEO of PDQ Enterprises, which operates an alternative trading system. Previously, Ross was the CEO of high-speed trading firm Getco from 2002 to 2005.