Washington, 13 September 2010

When I started in business school many years ago, the very first case that my class worked on was about a company called TANSTAAFL Sound – some kind of loudspeaker company, as I recall. For those of you not to have been among my classmates, the name is an acronym for There Ain’t No Such Thing As A Free Lunch.

The current angst surrounding the May 6 “Flash Crash” has brought back certain of the memories that accompanied that first day of class – giving rise to a critical question that needs to frame the arguments about proposed market structure changes: If the genie that is the current market structure is to be put back in the bottle, who is to pay for the changes?

Let me first say that I have no absolute knowledge of what the proximate cause of the May 6 event was. The real cause, as has so eloquently been explained by Steve Wunsch*, is the Securities and Exchange Commission and the actions that it has taken over the last decade-plus. Most significant of those is the 2005 enactment of Reg NMS

Reg NMS precipitated a chain of events that has created a market structure completely foreign to anyone who was a trading practitioner before 1997 and the enactment of another set of rules. It was then that the Order Handling and Display rules were promulgated in response to the findings of trading improprieties on the part of NASDAQ market makers. Those rules, while certainly well-intentioned, (as are all rules, I’m willing to stipulate) continued the process that really began way back in 1975 with the abolishment of fixed commissions – one rule that has been continually shrinking the profit opportunities for liquidity suppliers.

Fast forward to 2005 and Reg NMS and that, combined with decimalization from a few years earlier, furthered the process so that, depending on where one sat, it was either a disaster or the creation of a spectacular opportunity. To the New York Stock Exchange, it was arguably as disaster – contributing directly to the ill health of the exchange’s stock trading business as its share of trading in NYSE-listed stocks has plunged to less than 30% from 80% plus in the subsequent five years. To the Knights, GETCOs and DirectEdges as well as the internalization efforts of the big brokerages that picked up much of the lost market share, it was a boon.

But the new market structure that emerged was not envisioned and intentionally created by the combination of regulatory efforts. It has, in fact, evolved on an ad hoc basis. The result: the market today is the same type of complex system that a skeptic could, upon a bit of detached reflection, compare to NASA’s Challenger or Apollo 13, Deepwater Horizon or the Northeastern power grid. No prizes for guessing where I’m going.

These are all very complex systems that have demonstrated the capacity to fail spectacularly in ways that no one was a priori able to foresee. Perhaps the real wonder is that what we experienced on May 6 didn’t happen sooner. With some perspective, it’s now apparent that stop-loss orders and the like would fuel the downdraft and that so-called stub quotes would get hit in ways that no one could have anticipated. Going forward, I suspect that those once-innocuous practices will never be used again in the same way.

So what of TANSFAAFL? It has been the apparent design of the SEC to democratize the market and provide a level playing field for all market participants, but there IS no free lunch. Costs borne by investors to have access to liquidity are the revenues enjoyed by liquidity providers. Stated otherwise, brokerage revenues are what investors pay to get their trades done. Since 1975, those revenues have been continually squeezed – no doubt to the benefit of investors both small and large. But have we gone too far – to the point where the market can not or does not support its intended functions?

On one important count, I think that the answer is clearly yes. The current structures have evolved to where no one seems able to make sufficient profits to support the capital formation functions that enabled the IPO market to grow steadily through the end of the 1990s and into 2000. Where the revenues generated from NASDAQ market-making activities could be combined with investment banking revenues, that capital formation and provision function seemed to work. No longer.

Quite simply, with decimalization and the impact of the Elliot Spitzer Global Research Settlement, profitability has been squeezed out of the majority of that business. While certainly there are other significant factors also contributing to the lack of IPOs, fixing the market structure-related one will require some kind of different market structure that will support dramatically widened spreads and thus providing some semblance of the market support that new companies enjoyed prior to 2000.

Turning to the secondary markets as a whole, we seem to have made the implicit tradeoff between lower trading costs and market stability. For the small investor who requires liquidity in mostly very small amounts, accessing sufficient liquidity is rarely a problem – absent events such as those of May 6.

The institutional investor faces a different set of issues. The large buyside desk used to deal with broker trading desks over which they had some influence – after all, they did control the commission flow. That relationship has been supplanted to the point where the buyside desk is now mostly interacting purely with technology where market makers and specialists have been replaced by the computers of high-frequency trading operations. On balance, they’ve adapted and gotten used to seeking liquidity via algorithms and slicing orders to counter the efforts of stat-arb trading operations. They’ve also adapted to the volatility that the diminished price formation capabilities the current structure fosters by using VWAP algos so as not to be side-swiped by intra-day volatility. And while, on balance, the cost of acquiring liquidity certainly has trended down to historically low levels, it has come at the expense of greater volatility and less market stability. An elevated VIX is generally the evidence of that but then we have a May 6.

Back to TANSTAAFL. Changing the structure to provide greater stability will cost money (think liquidity provision). At both the recent SEC discussions regarding the flash crash and elsewhere, it has been suggested that those providing liquidity universally be burdened with a greater obligation to quote. Considering the razor-thin margins that characterize most of today’s trading, most any reform will increase costs.

High-frequency traders, including the internalization pools of the big broker/dealers can trade at their option and when it’s profitable. Anything that requires capital (with the associated capital costs) or changes the cost dynamics for existing market participants (such as a “Trade-at” rule) will necessarily carry implications with respect to trading costs, available liquidity and inherent or potential levels of market volatility.

The lesson to be derived here is that actions create consequences – diminishing market volatility would likely only be accomplished by increasing certain types of trading costs – costs that are not likely to be equally borne by all classes of investors. “Fixing” a market attribute that any particular interested party finds objectionable will have implications. Great care must be the order of the day in any new regulation. Remember, the fixes of five years ago have left us with the current, presumably unacceptable, status quo.

And by the way, I don’t recall how the TANSTAAFL Sound case was resolved either.

Ends --

Kenneth J Burke, Capital markets industry veteran

www.tabbforum.com