Seth Goldstein

What is the value of a relationship?

It probably doesn’t come as a surprise to anybody reading this that most large investment banks have proprietary trading desks, which buy and sell different investment instruments to generate earnings for the banks. Although the banks don’t generally break out the specifics, the overall results are impressive. In its last quarter, for example, Goldman Sachs reported $4.12 billion in revenue from trading and principal investments, up 47 percent from the year-earlier quarter and up 57 percent from the fourth quarter of last year. At the same time, many of these same banks (for example, Goldman, Morgan Stanley and Bear Stearns) are managing more and more prime brokerage assets on behalf of hedge funds. For example, according to Richard Lindsey, president of Bear Stearns’ global clearing services unit, Bear Stearns’ global clearing unit has generated about 25 percent of the company’s pretax profits since 2000.

My intention is not to bore you with the nuances of these firms’ lines of revenue, but rather to point out what appears to be at the heart of the unholy alliance between the buy-side and the sell-side of Wall Street. Investment banks are disclosing the results of their proprietary trading activities, at the same time as they are busy looking to prime broker hedge funds assets . These banks have employees that are paid to trade internal assets. They also have employees whose jobs are to become as familiar as possible with the movement of of their clients’ hedge fund assets. As far as I am aware, there is no current or pending legislation establishing any sort of wall between these two conflicting activities. This is not to say that some banks aren’t being proactive in terms of separating these functions. It’s just to acknowledge the status quo.

A few weeks ago, the Financial Services Authority fined Deutsche Bank’s Morgan Grenfell unit along these lines:

The FSA found that Morgan Grenfell commenced proprietary trading in seven of the constituent securities of a client’s programme trade, prior to its award, based on limited information provided to enable the firm to quote for that business. The proprietary trading resulted in the client paying more for the programme trade than they would otherwise have done.

For the full release, click here.

As the FSA uncovered at Deutsche Bank, some of these banks are so successful at managing their own internal proprietary trading desks because they are learning from their external hedge fund clients. What is the trading cost here to investors? Maybe it’s not a $.05/share price after all, but rather closer to $.08 or $.10 per share when you factor in the costs of the banks themselves mucking up the hedge fund’s trades.

Hedge funds are some of the smartest financial minds around and they definitely are not continuing to entrust their funds to bulge bracket investment banks out of ceremony. In addition to asset management, trading technologies, executive recruiting, etc., hedge funds also assess value to getting access to shares of Initial Public Offerings. In a nutshell, funds are willing to put up with a lot of shitty execution and mediocre research in order to get the highest possible allocation of Google shares at the IPO.

And so for a brief diversion I offer you Miami Vice:

For me, everytime I think about Hot Issues and Safe Harbors, I think about Tubbs and Crockett. On one level, the metaphors of heat and waterways immediately conjures up that pulsating introductory sequence of fast boats cruising through Miami. On a deeper level, Tubbs and Crocket demonstrated multiple levels of moral conflict in order to penetrate the netherworld of sophisticated drug brokers and dealers. This is very much like the balance struck by the buy-side and sell-side. Everybody is conflicted, to a certain extent, since each is trying to profit from the other: “you hook me up with Google shares and I’ll do more trades through you;” “you help me set up my infrastructure and I’ll let you roll me out as the next hot US hedge fund manager your clients in Europe.” At stake are enormous fortunes paid out each year for the fund managers at the top of the food chain and for the brokers that help them get and stay there.

Yes there are structural tensions underlying the relationship of the investor and the trader, but these tensions are another word for edges which is what the buy side is forever looking for. Last week I was fortunate to have an audience with the former head of the equities division of a premier sell side firm. He is launching his own fund soon, and so he is pretty engaged with the notion of what the sell side can do of value for the buyside.

In less than a minute he described a scenario that for him represented maximum value. It wasn’t a single function but rather a combination of services that revolved around trading smartly in the market. Imagine a sales trader that you could call up about an idea or a theme. He had access to the most current company information, and was able to put this in the context of what everybody else on the street was expecting from the company. This benefit was further supported by anonmyous trading which nevertheless had real capital behind it. As the trader played out potential outcomes over different time periods, he was able to suggest options and other derivative strategies to take maximum value of the idea.

In the midst of these complex interconnections of learning about, buying and selling shares in a company, the relationship between the institutional investor and the broker dealer persists and even thrives. The obfuscation of interests benefits both parties. While many may correctly refer to markets as efficient, the relationship of the parties making these markets is not necessarily so. The inefficiency manifests itself in high transaction costs that affect the industry.

The next logical question is then whether high transaction costs are necessarily bad, and if so for whom. Even if the ultimate execution of a trade is done electronically for less than a penny per share, chances are that the idea generation and analysis that led up to the trade, not to mention the broader context of staying smart about a sector through attending conferences, meeting management, etc, together add pennies and even quarters of costs to each share traded.

This would seem to be “expensive” in terms of ultimate costs to investors, and yet at least in my limited dealings with hedge fund managers and analysts, I continue to hear their willingness to pay a premium for what they perceive to be a proprietary advantage. In the end, their competition is less other hedge fund managers and more the growing efficiency and commodification of equity markets. In a friction-free environment, there is no reason why everybody should not invest in index funds anf etfs. What distinguishes successful fund managers is their ability to outperform the averages and take advantage of unique sources of information and insight to produce alpha. The fully loaded commission and the performance fee of the hedge fund manager are related costs. I wonder if with increasing downward pressure on transaction costs (from the regulators, from the media, from shareholder services, etc) that there will be that much more scrutiny there will be of the economics of hedge fund management.

One can argue that hedge funds are unregulated and therefore should be free to charge investors whatever the market will bear, and that funds can pay out commissions based on whatever they perceive to add the most value. The relationship between buy-side and sell-side continues to pad the margins that benefit both sides of the trade.

What will more transparency into this relationship do?

Will funds continue to build out more functions that have been traditionally the role of the sell-side, such as electronic trading platforms, independent research and their own internal analyst corps?

Or will funds establish deeper links with investment banks at levels that are increasingly difficult to monitor.

Or perhaps both?

How is research distributed?

Is it embedded in the trade solicitation?

Or is the idea distinct from how it may get traded?

Why is this important to begin with?

I would argue that the essential value of wall street lies in the location of where the buy side accepts external ideas as a part of its investment process. To a great extent, the buy side has all of the tools at its disposal to do its job, without aid from the traditional wall street brokerage house. The trader at the fund has 6 screens and can monitor pools of liquidity and place trades seamlessly. The virtues of anonymity, transparency, liquidity can all be addressed by technology just as easily as they can by a human sales trader. Some large complicated trades may still require human ingenuity to break up and place properly, but these too will become easier to manage through technology than individuals. Even flow, the near-sublime quality of a human trading desk has become simulated through syndicated web sites, and message boards .

So from a trading perspective it becomes more and more dubious why to pay anything more than the bare minimum ($.004/share maybe for now). To the extent that humans will be required, their skills are such that they will be easily exported to emerging countires at a fraction o the cost.

This may take years, not months, but it is inevitable. Financial markets have an endless appetite for technology solutions and business models that help automate their operations.

If trading becomes a commodity, then one has to look at the other products and services of the sell-side to justify their economic position. Besides trading, it would seem that the remaining values are access to hot issues (IPOs), access to management, conferences, and research.

In terms of hot issues, it seems certain that the allocation of IPOs will come under increasing restrictions by the SEC and the exchanges themselves. Granted, we are entering another bounty of successul public oferings and nasdaq recently eased restrictions on brokers’ allocation abilities; nevertheless, Spitzer has demonstrated clearly than any artificial price manipulation benefitting one institution over another (investment banking, market timing, specialist firms, and so on) will come with major penalties for the firms and their principals

Access to management is, after Regulation Fair Disclosure, more of a concierge service than anything else. It’s not that different from brokers procuring sports tickets for their best clients, except in this case it’s CEOsand CFOs of public companies instead of star 3rd basemen. There is no longer any suggestion that the management content is different based on who is introducing the company to the fund. Simply that management’s time is valuable and there are only so many “in person” mtgs they can afford, even if the content is exactly the same as all other meetings.

The buy-side continues to ascribe value to physical contact, over and above any other form of information. Since companies are legally barred from selective disclosure, and understanding how much funds currently value face to face contact with management, we are talking about the most expensive facial expressions this side of jim carrey.

Conferences are somewhat similar, insofar as they tend to be simply opportunities to have multiple “management 1 on 1s” in a single location. The lineup of speakers is interesting, and the collection of friendly buy-side analysts is convenient, but the value again rests in the tics and gestures.

Which leaves us with value of research.

Research as representing the stream of ideas codified in some form of communcation that comes from a sell side firm directly to the buy-side. Of course there are many kinds of research, from the old school of financial models and analyst opinions to the new school of forensic accounting, credit analysis, and access to networks of experts.

Technology threatens to commodify the value of research as well. Smart analysts that attend management meetings, pore through 10K’s and produce intelligent analysis are now facing considerable pricing pressure from syndicated management calls, easy access to public disclosure and cheap analytical labor in remote places such as India.

The most successful research firms are those that aren’t simply running ahead of the inevitable surge of commodifying technologies, but rather those that figure technology fundamentally into their research process. For example, many expert networks such as Gerson Lehrman’s effectively use the Internet to recruit, rank and provide access to experts across the world. The buy-side can to a certain extent bid on this expertise as if they were using Ebay. This is an example of a research product that uses technology to scale insights. To the extent to which it is able to tie the scaling of insights to the generation of trading commissions, then such a firm successfully distributes research in a manner consistent with how the buy side will look to pay for research in the future.

The critical factor that I am struggling to address is why the buy side needs research from the sell side any more than it needs to spend $.05/share to execute trades through the bulge bracket? The economics of a hedge fund are so rich that such funds are able to attract the best and brightest analytical minds to work for them directly. In the past, the sell side was able to subsidize analyst costs through investment banking revenue and so were able to offer folks like Jack Grubman $20M per year. Those days are over, while today a key analyst at a large hedge fund can make more than $1M per year and a fund manager can make 20+x that in a good year.

The solution for the sell side is to come up with products and services that take advantage of its primary focus on research and minimize any consideration of other features such as trading, banking, etc. In the end, the fund is in business to make money trading stocks (or bonds, or commodities, or whatever), so that the fund’s internal research will always be a means towards an end but never an end in and of itself. The competition of the sell side is not other sell side firms, but rather the internal research departments of the funds themselves. Every day, the successful investment research firm needs to demonstrate value (in terms of number of experts in its network, or amount of proprietary data, or variety of analytical tools) that grows proportionate to attention and focus. Since the sell-side can’t win the war of having the smartest standalone analysts (who will naturally gravitate towards the buy-side for better economics), it can instead emphasize the scale it has achieved in certain areas that a fund will never on its own accomplish (such as building out market research panels, negotiating complex license agreements, cleansing raw data, etc.).

Again, I am talking my position since the company I work for Majestic Research is explicitly trying to take advantage of these changes. What continues to confuse and fascinate me, however, is how the development of these new sell side investment research systems will drive new forms of payment from the buy-side outside of traditional bundled commissions.

The name of this blog is transparent bundles, and by that I am suggesting a means of combining trading and research in a manner than maintains the efficiency and velocity of commission-based compensation while avoiding the shadows and whispers of traditional trading relationships. The MFS rant about soft dollars, Fidelity’s scrutiny of bundled commissions, and Spitzer’s unyielding assaults on market timing, investment banking conflicts and whatever he comes after next, are all combining to force the need for a “transparently bundled” solution to the forefront of the debate.

Complicating matters even further is the fact that funds are looking to cut down on their trading relationships as opposed to adding new brokers.

Need to do some more thinking on this one and report back when I achieve additional clarity.