The VWAP, the trading industry performance yardstick, has many critics. And rightfully so – it is used not because it is the best measuring rod, but simply because it is the easiest way to gauge a trader's performance.
Since the mid-1980s, VWAP, or volume-weighted average price, has been the most common benchmark to evaluate a trader's or broker's executions.
VWAP is calculated by dividing the total dollars traded in a stock by the total shares transacted over the same time horizon.
Gathering the transaction information is easy enough. The data on individual trades, the number of shares executed and the stock price are available from the consolidated tape. Many trading systems calculate VWAP as it evolves over the course of the day.
The comparisons provided by the VWAP serve at least one basic purpose: Keeping an eye on the trader's judgements, skills and market savvy. The VWAP shows a positive measure if a buyside trader's execution price is lower than the VWAP. The measure is mediocre – if not downright negative – if his execution price is higher than the VWAP.
Still, in recent years, daily VWAP's star has dimmed a little. That's because there are now hundreds of benchmarks that purport to measure what traders are trying to accomplish. Not to be outdone, there are several versions of VWAP, which vary based on the type of data used.
One version, for example, might exclude block trade data. Another might record the time the trade entered the market.
There are other measures, besides VWAP. For example, some are not based on average trade data through the day. They use time stamps, closing price, and decision price.
As the benchmark battle rages, best execution – and the importance of benchmarking performance – has gotten a boost from the bear market. If a portfolio manager becomes more efficient at running money, if he can save 100 basis points by executing orders more effectively, that's significant money saved in current market conditions.
Regulators have also been issuing guidelines and watching best execution practices. The final result could be increasing knowledge about what exactly best execution is for traders using different styles, and on large, multi-day, multi-portfolio orders.
But traders are not likely to dump the VWAP. "We all look at VWAP," said Madison Gulley, director of global trading at Franklin Templeton Investments, which has $260 billion in assets. "It happens to be the most prevalent and measurable indicator that exists."
But should VWAP be the sole measure of performance for how a trader handles a stock? Absolutely no, according to many pros.
Indeed, the industry view on VWAP is changing. Trading firms are demanding even more sophisticated approaches. Abel/Noser Corp., an agency broker, became one of the pioneers in trade cost analysis for plan sponsors, based on the VWAP.
"It's not the end-all measure," said Peter Weiler, senior vice president in Abel/Noser's investment management group. "It should be used in tandem with other measures."
VWAP was once criticized for comparing a trader's execution price to a snapshot of the day's trading. If a trader got an order on the desk at 1 p.m, for example, it wasn't fair, say experts, to compare his execution price to the average price across the day's volume – all of which wasn't available to him.
Abel/Noser, for example, now looks at VWAP over a daily period, order-length period, or investment performance period. But it also looks at strike-price measures and has a benchmark test based on volume considerations for large orders.
Effects of Delays
Wayne Wagner, chairman of Plexus Group, the Los Angeles-based consulting and investment analysis firm, has never placed much stock in VWAP. One reason the use of VWAP is at times misleading, Wagner contends, is that it "misses the effects of delays and the opportunity costs of not being able to trade."
Added Wagner: "If it's a trade that can be done instantly, it's not a bad measure. But if it involves a half-day's volume, which is typical of institutional trading, then you must consider the tactics the trader is using to secure that liquidity because that trade might be alive for two, three or five days."
Instead of relying on market data, Plexus compares a client's trade data to its universe of money managers. A value manager can expect to have low trading costs. A growth or momentum manager, who's buying on the news, faces more substantial costs.
"If you're thinking about this as a head trader or chief investment officer, you want to look at people like yourself and see how well you're doing relative to your competition," Wagner said.
Timothy Blastek, head of equity trading at Provident Investment Council, a Pasadena, Calif.-based institutional money manager, says his firm doesn't use VWAP as a benchmark. He says his firm doesn't want to settle for an average performance. Instead, Provident, which has about $5 billion in assets under management, uses Plexus Group's cost measurement service.
Plexus analyzes Provident's trade data on a quarterly basis and measures it against its own benchmark. In the past, Plexus has told the money manager that if traders had accelerated the rate at which they made sales on an order, they would have been better served.
Basis Points
Plexus also said that going more slowly cost the firm some basis points in performance. Plexus also offered color on the portfolio manager's decision point. "They've come back many times and chastised portfolio managers, saying that had they got orders in earlier in the day, giving the trading desk the opportunity to trade for a longer period of time, there would have been a benefit," Blastek said. "That changed the way portfolio managers got their orders to the desk."
To be sure, there are many ways to look at data and evaluate trades, but not everyone is prepared to discount VWAP, even though it's a measure of average performance. "It's a reasonable measure if you have to trade 20 percent of a stock's daily volume and you're going to trade it all day long," said Franklin Templeton's Gulley. "How you spread it out and the decisions you make matter."
For a small order, say 1,000-5,000 shares of Microsoft, many pros say it's better to get it executed quickly and efficiently, and forget about comparing performance to the VWAP.
For very large orders, which involve buying five or six days worth of volume, VWAP isn't adequate. In those cases, Gulley says, the solution is a benchmark relative to volume. "Volume is a liquidity measure," he said. "It's about what you are trying to accomplish versus what the market can handle."
Franklin Templeton has run its data through a number of systems, including Plexus Group, Elkins/McSherry Co. and ITG Inc. It now has an in-house system that looks at VWAP and a value-added measure that takes volume into account on large orders that must be executed over multiple days.
"So if the benchmark says a trader should be able to trade 20 percent of the stock's daily volume at VWAP, the question is: How much volume did the trader buy and, based on that, what happened the following day to that stock?" Gulley asked.
"If the trader bought 40 percent of the order and the stock went up the next day, that was positive," he added. "If he bought only five percent of the order and it went up one or two points, that's true opportunity cost that's lost."
VWAP is a useful benchmark for a "logical participation strategy," according to Ananth Madhavan, a managing director of research at ITG. That's a strategy in which a trader tries to take advantage of all the liquidity present in the market over the course of the trading day.
"If I don't have a big order and am not likely to move the market through my own trading, and I don't have information that's short-lived," Madhavan added, "then it's a reasonable strategy to try to achieve VWAP."
Cost Forecast
Another way to benchmark costs is to use a pre-trade model. The post-trade performance analysis – using VWAP, the snap price or the closing price – can then be handicapped relative to the pre-trade model. That forecasts what the trade should cost at the time the order is received, says Madhavan.
Meanwhile, the attention on best execution is increasing around the world, fuelled by changes in market structure and bear market conditions. Weiler at Abel/Noser notes how decimalization has expanded the number of price points in a stock but has reduced market depth. That's made it more important – and difficult – for traders to carefully work orders.
Non-U.S. data is still frequently not clean enough to analyze as closely as U.S. data, but the direction in which trade analysis and measurement is heading leaves no doubt. The objective, after all, is indisputable: figuring out whether a trader made the right decision on a stock.
