The Agency Revolution

Is the independent agency brokerage business model a viable one for the 21^st century or are these brokerages doomed to become offerings from larger financial firms?

Take the announcement of Liquidnet's planned acquisition of Miletus Trading, for example. The deal bodes well for both parties. For Liquidnet, the deal will remove impediments to potential orders, or indications of interest, as more institutions look for simultaneous representation in block-crossing networks and algorithms. Miletus will gain entry to Liquidnet's impeccable customer list while removing the stigma of being owned by large hedge fund Thales Fund Management.

This transaction will have consequences within the equity-execution industry, but perhaps none is more important than the benefit it will provide for institutional investors, as it will introduce greater competition within the institutional-friendly agency electronic execution space.

The agency brokerage space is getting interesting. Investment Technology Group (ITG), which dominates this space, has been joined by two newly merged entities—BNY ConvergEx, nee Bank of New York and Eze Castle Software; and Nomura-Instinet.

Expect to see the competition for the few remaining independent agency brokerages erupt into several more announced deals. Let the M&A handicapping begin.

To follow the model—whether it's prop or agency trading—firms will need to offer a full algorithmic trading package, a high-end transaction cost analysis (TCA) suite of tools, a presence on the buy side's desktop, and a block-crossing capability in order to attract order flow from competing firms. All of these offerings also need to be global in nature.

The Miletus purchase begs the question: What is Liquidnet's broader ambition?

With crossing, and now agency trading, Liquidnet might seek to become the next all-encompassing agency financial services firm, a la ITG. To head down this road, Liquidnet might acquire an order or execution management platform (OMS or EMS) or a direct market access platform to make up for its absence in the desktop real estate space.

Liquidnet might also recognize that the Miletus acquisition could now threaten its broker liquidity provider relationships for H2O and seek to build up its small order flow presence.

Liquidnet CEO Seth Merrin's resume is loaded with experience as an innovator in the OMS segment. It wouldn't be surprising to learn that there is a larger plan in place and that another shoe will drop shortly. But will it be a slipper or a boot? Smart money is on the latter.

Additionally, there will be repercussions in the U.S. algorithmic trading market share rankings. Conventional logic says that the algorithmic trading leaders, such as Credit Suisse, Goldman Sachs and ITG, will suffer from the dilutive effect of a major new competitor. Conversely, their market positions are relatively stable. The losers here will be the secondary and tertiary providers who receive algorithmic orders from the top 100 or so institutional accounts.

The competition is now too stiff, and the differentiation between simple trading algorithms is too great for buy-side traders to continue to justify using these lower-end providers. They will quickly become marginalized and will compete with one another for flow from broker-dealers and smaller hedge fund clients.

Liquidnet's purchase clearly puts it in the agency model, competing directly against the bulge-bracket adherents to the proprietary trading model. A rarely discussed perspective on this agency model expansion is whether or not it can compete against the prop trading model.

Some say the technology revolution has merely been an avenue for the bulge brackets to rapidly digitize their customers' information into new data points for consumption by their proprietary trading businesses. Obviously, the bulge-brackets now control the majority of institutional order flow. Can the agency model make inroads here?

For better or worse, the days of an independent, stand-alone alternative trading system (ATS) are gone.

Lately, as every broker and exchange seems to be introducing their ATS du jour, fears of diminished, transient and harder-to-locate liquidity abound. Some brokers have attempted to aggregate as much of this block liquidity as possible, but there are significant barriers to true aggregation.

Indeed, many of the ATSes prefer only the direct-to-customer route, with their internal liquidity being preserved for their best clients. Additionally, in the world that existed prior to the ATS explosion, real crossing rates rarely exceeded 5 or 6 percent. Can further disaggregation help those rates?

The prospect is doubtful at best. To be sure, many scholars have suggested that improving the existing crossing rates is mathematically impossible given the variables involved. What we know for sure is that most institutional customers are still searching for block liquidity without being subjected to the perils of information leakage. There is a concern that these expanding business models are really just a way of saying, "This is as good as it gets."

Will the growing list of agency model devotees take significant market share from the prop-trading modelers? Barring a scandal or a new method of proving best execution, it won't happen. Their market share gains will be at the expense of the second- and third tier providers.

Given the size of the investments already made, and the liquid assets available to the stakeholders, it's almost a sure thing that additional acquisitions of synergistic niche providers will occur rapidly. Finally, the unfolding competition in the equity-trading space is likely to produce new and innovative product offerings that will continue to stimulate the industry's entrepreneurial instincts.