Buy-side divided over use of benchmark algos

Have benchmark-hitting algorithms such as VWAP had their day? While more sophisticated, liquidity-seeking strategies might seem more suited to today’s unpredictable trading environment, responses to theTRADEnews.com’s March poll tell a different story.
By None

Have benchmark-hitting algorithms such as VWAP had their day? While more sophisticated, liquidity-seeking strategies might seem more suited to today’s unpredictable trading environment, responses to theTRADEnews.com’s March poll tell a different story. Asked whether they would be cutting their usage of benchmark algorithms this year compared with 2008, a majority of traders said they will use benchmark algorithms less this year, but a significant minority – 46% – said they would not.

One explanation for this split is the differing execution priorities of various types of fund. Execution strategy can come down to the resources available: if you don’t have the time to take an active approach and use advanced liquidity seeking algorithms, you can at least leave certain orders to benchmark algorithms to deliver predictable results.

“These approaches are not right or wrong – it often depends on the type of operation: how much time traders have to spend on orders, how many orders are on the pad, whether they are dealing programmes or single stocks,” says Chris Andrew, head of product for J.P. Morgan’s

Electronic Client Solutions division, EMEA.

Benchmark algorithms are also better suited to certain types of investment strategy. “There clearly is still demand for VWAP strategies,” says Paul Squires, head of trading at buy-side firm AXA Investment Managers UK. “It very much depends on the sort of fund manager you are trading for.” For example, Squires suggests that a low-touch, high-frequency quantitative firm may be happy leaving orders running in a VWAP algorithm. However, firms looking to add value on difficult trades, such as large orders, need a more sophisticated approach. “Most of our activity is for fund managers who are taking a long-term view based on fundamental research, which puts us into a very different category,” he says. “If you are trying to buy 2% of a listed company you are not going to get very far simply putting it into a VWAP algorithm.”

While current volatile conditions and the availability of more sophisticated strategies suggest that benchmark algorithms should not continue to enjoy such strong support in 2009, users are customising and tweaking them, which is helping to extend their

shelf-life.

“Traders are using VWAP in a more assertive manner,” says Scott Bradley, head of sales for J.P. Morgan’s Electronic Client Solutions division, EMEA. “We are seeing a greater use of limits on VWAP orders, both in terms of price and percentage of volume constraints.

If you add user defined constraints and take control over certain parameters such as price and overall percentage participation it can help to minimise concerns surrounding benchmark deviation, particularly during volatile market conditions.”

“We are seeing more people using algorithm features such as ‘get it done’ or ‘would if I could’ prices – essentially having their cake and eating it,” adds Andrew. “If the market doesn’t come within their strike area they are happy with a more vanilla execution, but if the market suddenly comes into their range they can be very aggressive and seek out liquidity instantly without having to monitor the order too closely.”

As traders look to combine strategies to get the best results in unfamiliar trading conditions, they are fusing benchmark algos with other trading tools. Some brokers, for example, now add order-routing and liquidity-seeking functionality to their suite of benchmark strategies to allow them to execute more effectively in fragmented markets.

Despite the emergence of newer, more complex algorithms, and the waning popularity of VWAP as a benchmark, VWAP algorithms are likely to remain a core part of traders’ toolkits for some time to come. “Regardless of how the industry comes to view the benchmark itself, there will always be a demand for simple, predictable, over-the-day execution mechanisms. VWAP algorithms do this very well,” says Andrew at J.P. Morgan. “Even for more complex tasks like program trading, they are indispensable as simple but powerful tools that allow the coordinated execution of a portfolio.”

«