one can barley make it through a full day of tweeter skimming without encountering at least one anti fragmentation rant. And yet the truth of the matter is fragmentation is not a bad thing, provided it is done properlly. Lets consider US equity market history to illustrate the point.
The NYSE was started under a buttonwood tree on Wall Street. Within months regional exchanges were popping up all over the country, trading the very same stocks. Marconni had yet to invent the radio, so regional exchanges were needed to service the needs of local investors. As technology allowed for orders to be routed to a central market, and market data to be diseminated around the country, the national market gained power and prominence. This promence lead, as is so often the case to a sense of entitlement, lazyness and disappointiing service and innovation. As the national market lost its way, regional markets would gain share and force the NYSE to up its game. Competition worked to improve the markets. When the NYSE stopped servicing issuers in the manner they deserved, NASDAQ stepped up and fragmented the listing market. When Nasdaq and NYSE didn’t service traders ECNs lead by Instinet came to market. And when the specialists took advantage of large orders the buyside approached regional broker Jeffries to create the first dark crossing network (POSIT). All examples of fragmentation leading to better markets.
In the late 80’s the optiions market was dominated by the CBOE. But they too became comfortable and needed competition from the Philly and Amex to light a renewed fire. The options markets of the early 90’s created some of the earliest problems with multi quotes and trade throughs. Routers of the day routed options to a static exchange – i.e. all Micron options for a given broker might be routed to the CBOE – and the tables were updated infrequently. The driver behind where to route was typically size commitments made by the specialists. The CBOEspecalist on Micron would call up large option trading firms like MASH or Pershing and commit to 50, 100 or 200 contracts at the touch if they got first look at all orders. the market makers at other venues would then compete by offering more lucrative commitments. Note the commitment was to the benifit of the end clients, not the dealer agents. All good.
So why has fragmentation gotten such a bad rap these days? A few key reasons. First, ever since Bill Karsh had Direct Edge set up a second venue fragmentation has most often not represented a net new entity, but rather another venue within an existing family. Such venues don’t compete for existing flow. Instead they try to segment the flow – retail to venue A and institutional to venue B for example – and thus create an opportunity for intermediaries to stich the flow back together with volume and revenue enhancing trades.
Further the exchange operators of yesteryear used real innovation and service to compete with the incumbants. Sadly today’s exchange executives typically have less grey matter than the back wall of Kurt Cobain’s garage, and consider innovation to be whatever the HFTs from Kansas City and Chicago tell them to do, without fully grasping the actually impact of their follies. This results in markets competing not for natural flow, but for intermediation. Markets that compete for naturals tend to do well by naturals, markets that compete for intermediaries do well for the intermediaries.
The problem today is not too many markets, but too many markets trying to create opportunites for intermediation. Competition is good, provided it is aimed at the sustainable costumers. Competition aimed at the unsustainable middlemen makes for unsustainable exhcange models and frustrated investors. But try explaining that to an exchange official who has never traded a size order in his life, or actually read any of the dozen of academic papers he routinely quotes to justify his lazy behaviour.
But i might be biased, i actually trade for a living and do my own homework.