Meet a former market maker. He's David Wehrle, a veteran of the third market who switched to the buyside. He's director of trading and operations at Archetype Risk Advisors Inc. (ARA). ARA, a quantitative money manager, wanted a trader with traditional sellside tape-reading skills, so it hired Wehrle in March.
Wehrle spent about 15 years as a market maker in listed stocks. He started with Trimark Securities, which later became part of Knight Capital Markets. He traded 15 large-cap listed stocks, including Hewlett-Packard, Compaq and Tyco. ARA, which has $75 million under management, was recently reorganized. The New Haven, Conn.-based firm has two managed futures funds and a fledgling quantitative equities fund that was launched in August with $11 million in seed capital. A former equities fund, which had $650 million under management, was closed in 2003. Wehrle oversees trading for all the funds. Many traditional OTC and listed market makers have complained about decimalization. It compressed spreads and took the wind out of many a P&L. Wehrle finds it ironic that decimalization contributed to the quantitative equities strategy he's now executing.
The model behind ARA's quantitative long/short equity portfolio was designed by Thomas Cover, a Stanford University professor. Cover's algorithm starts with a portfolio of longs and shorts and reassigns the weightings of the longs every day, based on their volatility and momentum. However, because the strategy involves high-frequency trading, it wasn't practical until decimalization changed the markets.
"When the strategy was backtested a few years ago, trading costs and slippage ate into the returns at too high a level to make it profitable," Wehrle says. That equation changed with lower commissions and the rise of algorithmic trading systems. The portfolio currently includes about 80 long positions and 80 short positions. These are almost all volatile small-cap stocks. At the moment Wehrle doesn't use brokers. He executes the algorithm's daily shifts in weighting allocations himself. These trades typically can take up to two hours to implement.
ARA uses the ITG Triton front-end system, which provides a range of algorithmic trading tools. If Wehrle wants to get a set of trades on the tape immediately, he sends them to a smart router. If he wants to match the one-hour volume-weighted average price, he funnels those trades into ITG's activePeg server, which can sniff out hidden liquidity.
With algorithmic tools expediting trading for three-quarters of any day's orders, Wehrle is free to concentrate on the 10 or 15 harder trades that the model sends down. Wehrle trades these low-volume, volatile names the way a market maker would. "I look at the montage and I throw out my own probing orders to see if there's liquidity that's not being shown," he says.
For example, there's one stock that Wehrle trades almost every day. It's a low-volume name, so selling 2,000 shares would have a significant market impact. Wehrle no longer puts out his own bids or offers, even if he has only a couple of hundred shares to buy or sell. He's afraid of being pennied. Instead, he watches the prints to identify the stock's direction and to see whether there's a buyer or seller not displayed. Although Wehrle trades mainly small-cap Nasdaq stocks, some of the names in his portfolio are listed on the Big Board. He welcomes the NYSE's efforts to improve Direct+. But even the feature that will allow traders to sweep the book up to a 5-cent limit is no more than a "quick fix," in his view. "Ultimately, the human capital on Wall Street will be replaced to a large degree by systems that can do the job quicker and cheaper," he says. "The exchange has just bought itself some time."
