Over the past several years, the equity trading markets have advanced so quickly that it might be difficult for many traders to remember the days before advanced electronic tools, millisecond response times and, in particular, the growing number of equity dark pools.

Not only has the number of dark pools grown -- from six to more than 30 in the last five years -- but the share volumes executed on these systems also have grown. As firms identify new and more efficient ways to access liquidity, many are considering what role dark pools will play in the future, what it could mean for their firms and the way they trade, and which trading tools to incorporate into their practices.

For many players, a core component of adapting to this new trading environment is the adoption of liquidity mapping algorithms -- predictive statistical techniques to help detect where liquidity may exist in the marketplace, even in the dark markets.

While the sheer number of trading venues and the connectivity effort involved in accessing all of them can be daunting, until recently the issue of fragmentation has been more illusion than reality. According to recent Fidelity analysis, liquidity remained heavily concentrated in the largest exchanges as recently as 2005, with the New York Stock Exchange and NASDAQ collectively representing approximately 70% of market-wide equity volumes.

Over the past several months, however, we have seen a noticeable shift in liquidity patterns that may signify a dramatic change in the trading landscape. After reviewing the data from September 2008 through April 2009, we see that the percentage of trading conducted with these non-traditional liquidity pools has grown steadily from 15% to 24%. CrossStream, our own dark pool, also saw similar growth during the same period. Conversely, the market share of the single largest trading venue -- NASDAQ -- fell from 30% to 20% over the same time period.

It would be an oversimplification to suggest that liquidity migrated en masse from NASDAQ to the dark pools, but as the graph illustrates, the two statistics are highly and negatively correlated.

Of course the trading conditions of the past six months have been highly unusual, with record volumes and volatility, massive de-leveraging by large market participants, and industry-wide asset re-allocation. While it is difficult to isolate the impact of these cyclical issues from longer-term secular trends in the industry, we believe three fundamental shifts may be driving this remarkable reversal in trading behavior:

1. Broker/dealers open access to proprietary dark pools. Broker/dealers continue to evolve their dark pool business models by extending access to their proprietary liquidity to other firms, and improving their technology and product functionality.

Initially, dark pools were structured as stand-alone venues, capitalizing on access to the liquidity or preferred counterparties. Eventually, most of these private pools have opened their doors to brokers through mutual access agreements and enhanced dark aggregation technology. Most recently, brokers have begun incorporating dark books into their traditional algorithmic, routing, and direct market access products.

2. High-frequency traders move aggressively into dark pools. Historically, electronic communication networks (ECNs), with their speed, scalability, and favorable economics, have been the primary home for high-frequency liquidity providers.

However, more recently, the industry has seen this same group begin moving into the largest dark pools, attracted by increasingly robust technology platforms and the diverse liquidity that the venues offer.

The exchanges are responding to this shift, as evidenced by the flurry of pricing changes implemented from the major venues. But if history is any guide, episodes of rapid market share erosion (e.g., SuperMontage in 2003-04 and NYSE in 2007-08) can be very difficult to reverse.

3. Advent of sophisticated liquidity mapping techniques. As market players refine and, in some cases, overhaul their routing technology for sourcing liquidity, they are realizing that historically their liquidity sourcing algorithms have relied heavily on displayed quotes. Unfortunately, quote-centric routers are not ideal in the today's trading landscape as dark pools do not display quotes.

Liquidity mapping, on the other hand, incorporates predictive techniques to help traders detect and respond to liquidity in real time by analyzing market data and execution activity, even in the dark markets. These new trading techniques are a significant improvement over rigid, quote-based routing methods that may quickly lose their effectiveness as the percentage of off-exchange volume grows.

The convergence of these factors means that the industry's recent market share reversal may be more than a passing blip. With high-frequency players and brokers simultaneously gravitating to the non-displayed markets, both sides of the trade are increasingly likely to find one another in a dark pool. And as these participants refine their liquidity mapping technology, they will become ever more adept at executing trades efficiently.

We may be witnessing the early stages of a sustainable and fundamental shift in the trading landscape, as dark pools move from the periphery and take center stage. This could have a profound impact on traders and investors of all stripes -- especially as they evaluate their electronic trading techniques and determine where they source liquidity and how they achieve best execution.

Jeff Brown is senior vice president at Fidelity Capital Markets, the institutional trading division of Fidelity Investments', providing trading, products and services to a wide array of clients and Fidelity businesses.