“Your article was pretty accurate, but it had one thing wrong,” Rich Winter, who runs sales for BGC’s new US Treasury platform, told me at the Fixed Income Leaders Summit in Boston last week. Rich and I spoke on two occasions for 30 minutes about BGC’s plans for New eSpeed. Rich also gave a presentation to the buy-side that was off-the-record. Which part of the article wasn’t accurate? “It’s not an all-to-all platform.” BGC has already acknowledged that the sell-side, buy-side and high-frequency trading (HFT) firms can be customers of New eSpeed. Who’s left? “Retail investors.” While US Equities have significant retail participation, the $500 billion/day US Treasury market is dominated by institutional investors and has limited retail demand. Why is the “all-to-all” label so sensitive for BGC? Dealers hate when you go after their buy-side clients. Investment banks spend a fortune on servicing customers. They offer customers free research and trade ideas, they invite them out to lavish dinners and swing for luxury boxes at sporting events, they put on conferences where former Presidents speak, and they used to even let them partake in a little dwarf tossing. All this “free stuff” by dealers isn’t charity though; it’s to entice customers to send order flow their way. Dealers understandably get upset when a new platform springs up and offers firms, who don’t pay a customer service tax, direct access to their clients. Because banks are in the business of extending credit, they have tremendous power in determining which platforms succeed and which platforms fail in asset classes where equal access isn’t regulated and clearing isn’t centralized (e.g., US Equities and Futures). Platforms that can’t get a clearing deal at a reasonable rate are dead on arrival. BGC and sister firm Cantor Fitzgerald don’t have that problem though because they run a large US Treasury clearing franchise. However, banks do have other points of leverage: 1) They do a lot of business with BGC in other asset classes and 2) the US Treasury market structure requires dealer participation. Banks are primarily liquidity consumers (crossing the bid/offer spread) in the interdealer market and banks have dozens of other options to choose from. A new platform may never achieve escape velocity if it fails to attract enough interest from dealers. When banks cooled to the “all-to-all” vision of New eSpeed, BGC reportedly retooled their offering by splitting it into a central limit order book (CLOB) and a direct stream. Like BrokerTec, the CLOB will allow dealers and HFT to trade anonymously with each other. Participants can both provide and take liquidity in the CLOB. In the direct stream, there is a clear delineation between liquidity consumers and liquidity providers. Providers can’t trade with each other and compete to service buy-side customers. The sticking point for banks participating on the platform is that HFT firms can be liquidity providers to the buy-side. Rich made a case to me that HFT firms won’t be able to meet the minimum size requirements, but in my experience, there aren’t any banks that can provide more liquidity in a stream (not a request-for-quote) than Jump, Citadel, or KCG. BGC is betting that their CLOB will attract enough interest with its “science” and smaller tick sizes (sixteenths, 1/512th of a point, in the long end and 1/1024th of a point in the short end) that dealers will have no choice but the participate. I don’t mean to pick on BGC or Rich here. The battle over who owns access to client flow will always be a point of major contention in the market structure. In US Equities, an agency system arose that allowed for banks to get paid commissions for executing orders on behalf of their clients. Fixed Income has been slow to develop an agency model and likely won’t unless the liquidity situation becomes dire (e.g., corporate bonds) or regulation forces it through an expansion of rules around equal access and best execution. FENICS UST, and many other venues, are walking a delicate tightrope of appeasing their primary customer base while attempting to move the market structure forward.