The news last month that Optimark has failed in its bid to change the dynamics of US equities trading stands as a timely reminder of the difficulties e-commerce start-ups face.
Just a year ago it had deals with Nasdaq and the Pacific Exchange, had received a total of $236 million in funding from investors such as Merrill Lynch, Goldman Sachs, CSFB and PaineWebber, and from leading venture capitalists, including $100 million from Softbank. At one point it was valued close to $1 billion and was considering an IPO.
Now it’s had to suspend all its US equities trading, which was eating up a large portion of its operating capital without bringing in a great deal of revenue – just $3 million in 1999, with 75% of that coming from just four users. CEO Bill Riese has stepped down and the company is seeking to re-cast itself as a consultancy. Failure in the e-commerce sphere is not unique to Optimark. In the credit markets two trading platforms, Trading Edge and LIMITrader, have suffered by being single-market platforms which launched in the year when their product, high-yield debt, nose-dived. In early summer LIMITrader’s founder and CEO, Ed McGuinn, a former head of debt capital markets at Lehman Brothers, resigned.
Another who had to quit is Ronald DeKoven, founder and CEO of ereorg, a platform for trading bank debt and distressed debt.
Optimark, though, was the oldest of the bunch trying to break the mould. It was set up four years ago by Bill Lupien, a former CEO of Instinet, who spent much of the early 1990s secretly developing its software.
The idea was simple and laudable: it wanted to improve US equities trading, which is hugely inefficient both on the New York Stock Exchange and on Nasdaq. The market, claimed Optimark rightly, lacks transparency and anonymity, and that impacts upon trading costs. A 1998 report by research company Plexus showed the costs of trading in the US markets. Only 15% of the overall cost to the investor is the commission, it found. The rest is made up of: announcement costs, because of the immediate impact on the stock price (7%); delays by having to slice up the deal into chunks to avoid the market finding out who’s doing what (38%); and missed trades, which Plexus defines as those not completed within four days (50%).
Optimark’s plan to change this rested on its optimization technology. Say an investor wants to sell 200,000 shares at $25 a share. The traditional way is to call your broker, or brokers, who will then break it up into chunks and take hours or days executing your trade with all the costs outlined above. With Optimark, the investor could stipulate a broader set of parameters, such as that he would be prepared to sell his shares at a lower price, sell some of them at various lower prices, and at what point he was not interested in selling any. Execution would be carried out anonymously and would tap all liquidity providers at once.
Here were the two immediate problems. First, and the biggest single complaint from potential users across the industry, it was more complex than any other system on the Street, and required time to learn, and time to use. That did not appeal to traders.
Second, it relied on too many market participants to work with it, participants who had little to gain and much to lose. Optimark had no guarantee of liquidity provision from its investment-banking investors, so faced a chicken-and-egg crisis of building liquidity.
Its system was too advanced for Nasdaq to keep pace with, so although officially sanctioned to trade 250 stocks, it could only manage 10.
It faced competition from ECNs which began their march to capturing 35% of Nasdaq flows about the same time as Optimark went public with its plans.
Compare that to the rapid progress of Knight Securities. Knight did not set out to change any models, it simply thought it could offer a better, faster service, both to institutions as well as to the burgeoning retail investor class at a time when most market makers still pinned their hopes on maintaining fat spreads.
Over 50% of its flow comes from retail, and the vast majority of that comes from a group of 27 on-line e-brokers. Knight made the link worthwhile by offering them partnerships and profit-sharing deals, as well as paying for order flow. The model is at least open and simple, unlike the opaque and convoluted relationships and fee structures investment banks have with their major institutional accounts.
As for Optimark, it is now planning to direct its technological and intellectual capital – for that is about all it has left – to being a consultant and technology developer for electronic marketplaces more broadly. The lessons for other start-ups, existing and would-be, are pretty clear: be flexible, keep it simple, don’t attempt too much and don’t expect anything from anyone else if you offer nothing in return. Choosing the right market helps too, as LIMITrader and Trading Edge are finding out. The idea that if you build it, they will come, is something best left to the movies: great ideas and technology are no guarantee of success.
Optimark’s demise is just the start of the shakeout.
