Way down at the bottom there might have been a mention of the Red Sox. I'm not sure, you better check. ;-)
Hey, some of the puzzle pieces are bigger than 2-3 lines. Your loss. Interesting how decimalization has affected the institutional trader. Check out this next article.
Global Investment Technology® June 23, 2003 Strategic Intelligence for the Securities & Investment Industry Volume 12, Number 19
Liquidnet Wagers on a Trading Model Aimed At Blunting Sell-Side’s Block Trading Edge
Electronic trading forums have undergone myriad changes in the last two to three years, with ECNs carving out their own place against Nasdaq, and some morphing into exchanges themselves. Also now firmly on the landscape are trading venues specializing in matching and execution of block trades of large number of shares, which institutional investment firms need for high volume transactions. Most notable among those venues are Harborside+ (Spotlight Global Investment Technology, April 28, 2003) and Liquidnet, which has its own competing model for block trading. Global Investment Technology spoke with Seth Merrin, Chief Executive Officer, Liquidnet, about his view of the block trading landscape and what the challenges are in that business.
GIT: How has block trading developed and evolved in recent years? How has Liquidnet’s business model for block trading continued to evolve?
SM: When the Internet began to develop into an effective business tool, it affected so many different industries in such a fundamental way. The effects can certainly be seen on the retail side of the investment business, with online retail trading models such as E-trade, Ameritrade and TD Waterhouse becoming dominant players. But until Liquidnet, the Internet did not change or have any effect on the institutional side of the business, which continued to be very manual. That manual process has been a big part of the problem, with too many people involved and too much information being disseminated. Through the introduction of Liquidnet and our Internet-based technology, institutional trading can now realize the same benefits and efficiencies as retail trading.
However, there are still many parts of this business on the institutional side that have not moved forward with Internet technology that still need to do so. Market structure itself has really broken down for the institutions. The issue is as simple as Economics 101 where decimalization and a contraction of market depth (by about 70 percent since decimalization was introduced) have caused a disparity in basic supply and demand. Institutions have pulled back on displaying their intentions so there is not an accurate representation of institutional supply and demand. Instead, institutional demand becomes matched up with retail supply — a huge disparity that only snowballs. As less institutional supply and demand is represented on the floor, even more is withheld.
Institutions are being forced to compensate for this disparity by taking an institutional- size order and slicing it up into so many little pieces it looks like a retail order to the exchanges. Many of these block orders are being sliced into 300-share pieces to be executed, because that’s the total appetite that exchanges can handle these days. Ultimately, it takes thousands of executions to get the orders done, and the longer it takes to execute an order, the more it costs. ECNs and direct access providers help the institutions achieve this illusion. But, that is simply using technology to fix a bad problem with an equally bad solution.
GIT: How does shredding of blocks affect your business?
SM: It makes a system such as ours even more critical to trading desks. There’s been a huge fall-off in institutional order flow to begin with. Average assets under management are down about 40 percent in the last couple years. So there’s less order flow to go around. That’s the biggest thing affecting everyone’s business. Ultimately, when a trader sees a match in our system, he knows he can complete an entire order with just a few clicks. He can then focus on those orders that aren’t matched in Liquidnet, and break them up if he has to in order to get it executed on the exchanges.
GIT: Is the drop-off in institutional order flow the only challenge? Are there other challenges?
SM: We’re introducing a completely different trading model. It takes traders time to understand how to use it effectively, and to understand how Liquidnet actually allows them to execute in a wholesale trading market instead of a retail-oriented environment. That’s where transaction measurement services come into play, allowing traders to discover how inefficient and costly it is to trade institutional-size orders in the retail marketplace. A different strategy should be employed in a wholesale-only environment. Liquidity itself is very fleeting. When it’s there, it’s more important for them to take advantage of the liquidity presented, because if they see a match in Liquidnet, that means Liquidnet has created — for that specific order — the perfect marketplace where equal supply and demand come together at prevailing market prices. They have to measure what the cost would be if they refused to take advantage of that match and instead chose to go out into the retail marketplace and execute that order without a natural on the other side.
GIT: How would you compare and contrast Liquidnet’s methods with those of your competitors, such as Harborside? What are the strengths and weaknesses of yours and of theirs in comparison?
SM: A lot of people like to compare us with other electronic [forums]. It’s not really a true comparison. Harborside is more closely aligned with the traditional trading desk than with Liquidnet. Our models are completely different. That said, our biggest competition is with the traditional block trading desks at the large firms. Chances are that any trade done in Liquidnet was first destined for Goldman Sachs, Morgan Stanley or Merrill Lynch.
The traditional brokers use a shotgun approach to find liquidity. They have incentive to find the other side, because then they get both sides of the commissions. But the way they do it is actually very damaging to the customer. They will pick up the phone, call other buy-side desks, advertise on Autex, and send indications of interest. The problem with that is, according to the NYSE, it only works about 25 percent of the time. The other 75 percent of the time, this approach only informs the rest of the marketplace that there’s a large buyer or seller out there in the market. That sets everyone moving who will start buying ahead of the institution, knowing that institution has a large buy order. Ultimately, they will sell their stock back to that institution ... at a higher price. Our model finds the natural contra party electronically without making phone calls, advertising or anything like that. When there’s a match, there are only two people in the whole world who know there’s something to get done on Liquidnet. The more information is kept out of the marketplace, the less the stock will move. The ability to match size with size creates that equilibrium and creates a much more efficient marketplace in which to transact business. That’s why Liquidnet is the only firm that is able to show price improvement on 9 out of every 10 trades that occur in our system. The NYSE tells people they price improve. The problem is that NYSE has an average execution size of 665 shares. Price improvement on 665 shares is meaningless to institutions. When we price improve, we do it on nearly 50,000 shares on average. That’s meaningful.
GIT: What are the challenges in finding liquidity to fill the large orders without affecting the market or users’ trading positions?
SM: The challenge is to get as many buy-side firms using our system to the fullest capacity as we possibly can. Ultimately, all order flow originates from the buy-side. Certainly, the large investment firms are doing a tremendous amount of proprietary trading. The more firms on our system and the more that those firms participate in Liquidnet, the more money we’ll be able to save them, the more matches we can provide, and the better the marketplace will be for them to execute. There is more than $100 billion a year flowing out of their fund returns into the pockets of intermediaries that really have no business getting that money: specialist firms making money from their own proprietary trading, floor traders, fast-money hedge funds, and proprietary trading desks at the brokerage firms. The more the order flow is concentrated in a particular place, the more efficient that marketplace is going to be. For almost two years, we’ve provided matches and a more efficient marketplace for about a quarter of all the orders that our institutions have. If we can get that number up to 50 percent, we could really save our clients half of this $100 billion-ayear problem. Think about what their returns will look like.
GIT: Are there other consequences of the declining trading volumes for the industry, for trading forums and platforms?
SM: Yes, it’s huge. Brokerage firms have football field-sized rooms of traders sitting around with very little to do. There are two problems with that. One problem these firms face is the huge infrastructure cost that cannot continue much longer. If this market continues the way it’s going through the end of this year, there will be massive layoffs before bonus time comes around. The other problem they have is the more people they let go, the more relationships they lose, and it’s still very much a relationship business. It’s very difficult for them to make those cuts, because they understand that each relationship they lose, they lose that business. There’s still massive overcapacity based on the volumes that we’re seeing and have been seeing for the last couple years. Firms have been very reluctant to downsize and have been shaving costs incrementally. Some of the larger firms are shifting more of their money into proprietary trading to make their numbers. They will find that the customer business is not a profitable business for them based on their infrastructure costs. They’re going to start relying less and less on the customer end and more and more on their own proprietary trading.
GIT: What’s Liquidnet’s progress in Europe now?
SM: The European markets have seen a slow-down, similar to the US. The good news is that because of the Myners Report, there’s an overwhelming desire and need in Europe to solve the market structure problem. So our traction is very good. We have far fewer clients in London, but the clients we do have are serious users of the system and really try to get everything done that they can. So we’re about halfway to break-even in Europe, which is pretty good, since we launched there only at the very end of 2002. We believe we’ll be at break even or profitable shortly.
Market Data 
Markets 
AI
