Rising trading costs and less-than-optimal execution are causing problems for buy-side financial institutions already battling the effects of the global financial crisis. And armed with a better understanding of trade losses through the increasingly widespread use of transaction cost analysis (TCA), asset managers are taking dealers to task on trade execution - and, in some cases, even looking to trade themselves.
Trading costs have risen sharply in the past 12 months (see table), which is in part down to surging volatility. The Chicago Board Options Exchange's Vix index, which measures equity volatility, hit a high of 81.65 points on November 20, versus a close of 24.88 points on the same day in 2007, closing at 46.11 points on January 16. But asset managers are suspicious that dealers are also charging higher spreads in a bid to bolster flagging profits as revenues from other parts of their businesses have fallen.
Increased costs for buy-side institutions present a major challenge. Should execution prove less than optimal, many smart trade ideas fail to realise predicted theoretical returns. In essence, trading costs are having a significant negative impact on alpha generation in volatile markets. "Traditionally, the debate on alpha generation has focused predominantly on the quality of investment ideas, and brokers (broker-dealers) were remunerated for their ideas, not necessarily their execution," says Michael Corcoran, head of sales at specialised agency brokerage and technology company Investment Technology Group (ITG) in Sydney. "With the advent of TCA, this debate has shifted. It is still the case that investment ideas are at the root of the bulk of added value, but it is clear that a sub-optimal execution of these ideas can lead to a significant decay of alpha."
TCA is the analysis and measurement of the costs incurred in transactions, with the main aim of controlling such costs, achieving best execution and bringing greater efficiency to trading operations.
In particular, market impact and a lack of liquidity lie behind much of the recent spike in costs. Avoiding the impact of such conditions is typically the aim of TCA, but the surge in volatility and a widening of spreads in the past year has made doing so a lot more difficult. "On average, our expected costs per trade are around 15-20bp higher than they were last year, making the whole idea of trying to deal in size, finding liquidity and avoiding market impact a lot more important, but also a lot tougher to achieve," says Richard Coulstock, head of equity dealing at Prudential Asset Management in Singapore. "Having a desk that has the tools in place to source liquidity and knows where to find it is extremely important."
Prudential AM laid the groundwork for a more efficient execution policy around two years ago, a step Coulstock says is proving increasingly valuable in the current economic climate. The asset manager approached ITG to run a TCA back in 2006. The data demonstrated that execution performance was generally positive when benchmarked to volume-weighted average price (Vwap), a measure of the daily average price. But when benchmarked to implementation shortfall (IS) - the difference between the decision price and the final execution price - the costs were very high.
"Our analysis showed that the longer it took us to execute a trade, the more expensive that trade became in terms of market impact," says Coulstock. "It showed us that when market risk is high, it can be worth ignoring volume restrictions and paying over Vwap to complete a deal quickly, if it means you can avoid four or five days of price changes." Hence, Prudential AM switched to using IS as its standard benchmark and removed its volume restrictions to take advantage of liquidity whenever volume and price appeared at suitable levels.
TCA helps institutions conduct trading strategies that more fully consider the implications of market impact, along with the opportunity cost of remaining passive and allowing time to work against a trade idea. "In some cases - such as when a stock is not particularly volatile and opportunity cost is limited - it may be useful to trade gradually over the course of the day to reduce your market cost," says Coulstock. "However, if a name is more volatile, the opportunity costs inherent in such a strategy could far outweigh the benefits from ensuring a low market impact. In those cases, it may be better to take a greater share of what is available and accept some market impact, to reduce execution time and the risk of the market moving away from you."
Since volatility has increased dramatically across most asset classes in the past year, it is likely that a low-impact trading approach now has a higher opportunity cost than before the third quarter of 2007. "This is a key change from the one-instruction-fits-all-stocks practice we have seen in the past," says Coulstock. "Applying one methodology - such as a Vwap instruction - to a portfolio may well be relevant for some of those trades, but not all. That's really the key factor."
Schroders Investment Management also uses IS as its primary benchmark. "Vwap is just a measure of mediocrity," says Steve Wood, global head of trading at Schroders IM in London. "IS gives us the ability to measure the costs of the trade from the moment it hits the desk through to completion, allowing us to evaluate the cost structure across the history of that trade from trader timing to market impact. In the short term, it allows us to evaluate how we react to changes in market circumstances, while in the longer run, it can form the basis for a change in trading strategies."
While most firms now use some form of TCA, the proportion of firms applying Vwap as primary benchmark is still relatively high, but this is changing. The proportion of asset managers using Vwap as primary benchmark in Asia dropped from 67% in 2007 to 39% last year, and from 55% to 36% of the larger institutions, according to a December report, Asian Equity Investors - Asian Equities for 2008, from Connecticut-based Greenwich Associates. The report, based on data collected between mid-July and mid-September, says implementation shortfall is now used by 45% of larger institutions as their primary benchmark and by 39% of institutions overall.
Armed with a new awareness of trade costs and execution, and making increasing use of sophisticated trading technologies such as algorithms and dark pools, some buy-side institutions are moving to 'in-source' activities for which they have traditionally required a brokerage firm. "The improvements in technology have provided the buy side with a whole host of different tools with which they can control the execution process and differentiate themselves from their peers in a way that was not possible when everybody used the same broker channel," says David Klinger, managing director at equity trading platform Liquidnet in Hong Kong.
The growth in electronic trading has played a significant part in this process. Electronic trading now accounts for 15% of Asian equity trading today (7% algorithms, 7% direct market access and 1% dark pools and crossing networks), says the Greenwich Associates report, a figure that is expected to nearly double to 28% in three years, as cost-saving benefits and new technologies attract more players. Last year, these electronic routes accounted for 11% of trading.
As a result, the tables are turning in favour of the buy side. "When I entered the buy-side market 15 years ago, the (sell-side institutions) were the gods of the business, while buy-side traders were really just execution clerks, passing orders to the brokers who decided how to carry out the trades," says Prudential AM's Coulstock. "With all the TCA tools, liquidity avenues and transparent processes we have available today, the balance of power has shifted. If anything, it is the position of the sell-side trader that is becoming more questionable. With the technology currently available to the buy side, the traditional sales trading role is less important to us on a daily basis."
Schroders IM has also implemented a strategy that relies less on the traditional broker-dealer model. Over the past five years, the UK asset manager placed its strategic trading emphasis on electronic trade execution, whether through algorithmic trading, dark pools, crossing networks or direct market access. "It's about ensuring quality of execution," says Wood. "Rather than outsourcing that responsibility to brokers, we prefer to carry it ourselves."
The broker-dealer channel has consequently become just one of a range of liquidity sources Schroders IM will consider. "The art of the buy-side trader today is to identify best execution from a multitude of venues and to execute trades within those areas themselves," says Wood. "This approach has helped us keep our transaction costs down." But he declines to provide firm figures on the savings.
The buy side's focus on execution is also resulting in an unbundling of research and other premium services, with execution at broker-dealers generally sold as a package deal. "Given this ability to add additional value, it is not surprising that (institutions) are demanding the freedom to chose their own counterparties rather than be forced to deal with whoever it is that provides them with research," says Liquidnet's Klinger.
The increased emphasis on execution is clearly visible in the deviation from Asian broker vote results, which have traditionally valued research comparatively highly. The Greenwich Associates survey shows that Asian broker votes typically allocate 65% of average fund commissions to paying for brokers' research and sales services, compared to 40-45% in the US. However, in response to questions about actual payments made to brokers, the results varied substantially from the broker vote-determined amounts by 20-25% on average, indicating that buy-side traders are placing more emphasis on execution.
"The difference between the actual payments to the broker and the broker vote result is primarily driven by concerns about execution quality, backed by trade cost analysis," says John Feng, managing director at Greenwich Associates in Stamford, Connecticut.
Indeed, some firms have begun to hide broker-dealer vote results from their trading desks to keep the emphasis on best execution. "A number of our clients operate a policy whereby information about broker votes is deliberately held back from the trading desk, allowing them more freedom to trade with whichever sell-side party offers the best results," says Liquidnet's Klinger.
This separation of research and execution is further driven by the increasing popularity of commission-sharing agreements (CSAs) among buy-side firms in Asia. In essence, a CSA is an agreement whereby a broker receives a full-service commission rate, but keeps only the execution rate as payment for its execution services. The remainder is paid out at the client's instruction to several independent research providers, brokers or other service vendors of its choice. By using CSAs, the buy side can concentrate on best execution and pay for research separately.
"It used to be the case that the broker offering the research, would also carry out the execution, but this is no longer the case," says ITG's Corcoran. "Managers are looking for best-of-breed execution as well as research. CSAs provide them with an a la carte system of broker services, allowing investment management firms to cherry-pick different services from different providers."
The Greenwich Associates survey found that CSAs are used by around 40% of the buy-side community, but this figure is expected to grow significantly. "From a buy-side perspective, CSAs make a lot of sense," says Klinger. "It allows them to work out what a particular piece of research is worth to them and pay accordingly. I have no doubt the use of CSAs will increase throughout Asia."
For the sell side, however, these agreements bring added pressure to deliver top-notch execution and/or research. For those firms unable to compete with their more successful, often larger peers, the outlook is not pretty, and consolidation is expected. "What is troubling many brokers (broker-dealers) is the possibility that the value they attach to their research is considerably more than their clients think it is worth," says Klinger. "There will be consolidation on the sell side. We're already seeing a trend whereby buy-side firms are paying more for top brokers and specialist research or execution houses and less for mid-tier brokers offering second-rate research and execution. The increased use of CSAs will only enforce this trend."
Schroders IM's Wood echoes this view. "The top firms that can supply pools of liquidity, whether from other asset management firms or through risk capital and secondary placements, can still add a lot of value, not to mention their ability to supply algorithms and DMA platforms," he says. "But the more mediocre brokers (broker-dealers) will certainly struggle."
Yet some believe there will still be a future for institutions that offer specialist services across the region. "Asia remains a very fragmented market place, with many diverse countries and a wide range of stocks to cover," says Feng at Greenwich Associates. "It would be challenging for any buy-side institution to cover the entire continent efficiently, and the outsourcing model - whereby the buy side utilises broker insight and access - has proven valuable."