Are bigger trades better?

Chopping up large trades into many smaller ones is common practice among institutional investors seeking to minimize market impact, but the cost can be high.

And recognition of those costs might be driving investors back to block trading.

"There are times when these parceling techniques have a large effect on the (overall) trade, and you could have been better off trading in a block," said John Wightkin, co-founder of Quantitative Services Group LLC, Naperville, Ill., which provides transaction cost analysis services to institutional investors.

"By going to a parceling technique, you're introducing a lot more risk from a timing standpoint, and that can be six to eight times as large as the liquidity charges," he said. "So if you can go in and improve your parceling strategy to take advantage of more liquidity or find a block, you will perhaps have a larger (market) impact, but you could realize substantial savings on the timing of the trade."

James Bryson, president of New York-based trade analysis firm Elkins/McSherry LLC, added that by parceling trades — known as slicing and dicing — institutions could be paying a high opportunity cost as well.

"If you begin buying a stock at 90 and you're working the order over a day, two days or a week, and you finish at 105, your average cost might be 102," he explained. "By slicing up orders you lose opportunity" of completing the order at a better price — buying at a lower price or selling at a higher price.

The prime reason for the increase in smaller orders is decimalization (when stock prices went from being quoted in eighths of a dollar to pennies), according to traders. George Bodine, director of trading at General Motors Asset Management, New York, said the quicker pace of trading and the increase in electronic trading have also driven order size down.

"The ability for the broker to see the order flow and try to build a picture — there's a lot of people buying drug stocks or selling paper stocks — is not there," Mr. Bodine said. "That lack of information lessens your willingness to commit on larger blocks."

While smaller orders will have little market impact, those orders — and that impact — build up as the entire position is traded, driving a stock lower if the position is being sold or higher if it is being purchased.

But that's not necessarily the case with large-capitalization stocks like Microsoft Corp. or General Electric Co., according to Wayne Wagner, chairman of Plexus Group, Los Angeles. He said on those very liquid names, traders can do 10 or more 500-share orders a minute.

"The advantage of that, most buyside traders would tell you, is that they don't have to give up any information" about what they're trying to do, Mr. Wagner said. Giving up information, known as leakage, has become a bigger problem for portfolio managers and their traders since decimalization began in 2001. That's another reason trade size has declined.

"Leakage is a real issue when you have intermediaries — full-service broker/dealers with proprietary trading desks and specialists on the floor of the New York Stock Exchange," said Ted Oberhaus, director of equity trading at Lord, Abbett & Co. LLC, Jersey City, N.J.

Fear that broker/dealers or specialists will use a buy-side trader's information, e.g. the full order size, has led many traders to show only a small part of a larger order and to use algorithmic or computer-based trading models to complete trades anonymously.

Mr. Wightkin said the overall impact of breaking up a large trade in a highly liquid stock can be significant. "It could have a larger cumulative impact because of the larger footprint you're making when you break an order into 600 or 700 executions."

But in both very liquid stocks and thinly traded stocks, the fear of information leakage is one of the main reasons order size has declined.

According to a recent report by The Tabb Group, a Boston financial services consultant, the average trade size has dropped to about 500 shares per trade from more than 1,400 shares in 1997.

John Wheeler is head trader at American Century Investments in Kansas City, Mo., which has $87.4 billion under management. When a portfolio manager decides to buy or sell a position and initiates the process, Mr. Wheeler said, the manager and everyone else along the line — from buyside trader to broker — typically puts out smaller and smaller pieces to avoid giving anyone else in the market an indication of their true intent.

"It's going to take a long time to really change human behavior, and that's what you're talking about when you start looking at the way portfolio managers enter tickets to their desk to how the desk sends tickets to broker and how the order goes to the exchange," Mr. Wheeler said.

He said he sensed a huge amount of pent-up demand for automatic execution with deep markets in which large blocks can be traded. He added that the Securities and Exchange Commission could help drive the market in that direction with its Regulation NMS, which is out for public comment until June 30. (As proposed by the SEC in late February, Reg NMS would give institutional investors greater leeway in directing trades to electronic venues. Pensions & Investments, March 8.)

"If we could execute bigger blocks, we would," Mr. Wheeler said.

Gregory J. Rogers, a principal and head trader at Aronson + Johnson + Ortiz LP, Philadelphia, agreed that trading large blocks is preferable to slicing and dicing. The problem, however, is often one of supply or demand.

He said trading has become more work-intensive with smaller orders. His firm, which manages $15 billion, will sometimes offload large orders to brokers and let the brokers figure out how to get the order done, Mr. Rogers said.

Mr. Wightkin at QSG acknowledged that while slicing and dicing might cost more, sometimes an institution has no choice.

Taking advantage of liquidity in the marketplace is a niche that several companies have attempted to capture in order to provide institutional investors with ways to trade big blocks of stock both anonymously and without market impact.

One of the leading firms in this group is New York-based Liquidnet Inc., where the size of the average trade is about 44,000 shares.

"What we decided was, if you are able to combine all the institutions on one system, you've solved the problem of having large demand with no supply," said Alfred Eskandar, director of marketing. "If there's demand, a match (on the Liquidnet system) only appears when there's supply."

Liquidnet has 251 institutional members that collectively manage $6 trillion in assets, officials there said.

Other firms in that space include Harborside+ and Pipeline Trading Systems, both in New York.

"The fact is that for a few of these (firms) to be successful means there's probably more demand in the market" for block-trading facilities, said Michael Plunkett, president, North America, for Instinet LLC, the institutional trading unit of electronic trading firm Instinet Group Inc., New York.

Instinet and Harborside+ recently announced separate plans to expand their respective block trading services, giving institutional investors greater opportunity to reduce market impact when buying or selling large amounts of stock.