More trading in multiple asset classes has led to proliferation of data. The varying quality of information makes tasks such as transaction cost analysis difficult, finds Nicholas Pratt.
Asset managers have long been trading in multiple asset classes but the term “multi-asset class trading” now refers to more than simply a firm that has a fixed income team as well as an equities team. Due in part to the search for greater returns in a relatively flat market, managers are trading across several asset classes as part of the same transaction.
A number of buy-side firms are looking at reflecting this multi-asset class focus in their trading operations.
The first move that many managers have made is to re-organise the trading desk so that all the dealers in all the asset classes are sitting together and trading through one desk.
According to Christoph Mast, managing director and global head of dealing at RCM Allianz Global Investors, such a set-up “is the best way to use a team of professional traders and it is a better way to manage risk”.
RCM has a mix of four different teams – equities, fixed income, foreign exchange and derivatives – all trading from a single desk. “By having a multi-asset desk we are trying to encourage more interaction between the teams. Not only will this hopefully help to develop more effective strategies, it will also improve execution quality,” says Mast.
Proving quality
The problem, though, is that proving this improved quality of execution is not quite so straightforward. As trading strategies become more complex and investors demand more reporting, asset managers are being challenged to keep up in terms of obtaining adequate post-trade data.
“Asset managers want consolidated systems with consolidated data models so that they can understand the relationships between the underlying securities,” says John Mitchell, senior vice president for global sales, at data management vendor Asset Control. “This involves bringing data together from external brokers, market data providers and their own internal order books. And once you try and do this in a multi-asset environment, it becomes exponentially complex.”
The next step, then, is to employ electronic trading as far as the various asset classes will allow. “When you are voice trading in an over-the-counter market, it can be difficult to tell exactly when trades were executed and, therefore, what the price of the execution was,” says Robert Kay, head of analytics at Trading Screen, a vendor of execution management systems (EMS).
He believes that the challenge of providing best execution for multi-asset trading is only likely to increase the case for electronic trading. “One of the advantages of a multi-asset trading EMS is that a single electronic time stamp will capture all that information electronically.”
The system is also used for providing RCM with its multi-dealer platform for the quote-driven over-the-counter markets, another process that has typically been voice-based and lacking transparency.
“The main problem is that without an exchange it is very difficult to follow the price. By using a multi-dealer platform, we can get dealers to send us quotes and then we are able to select the best one. The fact that this is all done electronically means we are also able to demonstrate to our clients that we executed at the best price. That electronic audit capability is very important.”
Varying quality
But while the upgrades in EMSs have helped asset managers to pull together their own data, there is little they can do when it comes to market data, the quality and availability of which varies from asset class to asset class.
The equities market, as would be imagined, is relatively straightforward and are where most transaction cost analysis (TCA) initiatives have been focused up to now. It is similarly straightforward with futures contracts.
But in the over-the-counter markets the lack of central counterparties has made the post-trade data challenge far greater.
Consequently, some multiple asset classes are easier to follow than others. An equity trade mixed with a sector trade (where the trader is betting that, for example, Vodafone will outperform the telecoms sector but remain market neutral) is well catered for in terms of data, especially if the futures contract comes from a well-known derivatives exchange. Other multiple asset class trades with complex integrated hedges are less straightforward.
Efforts are being made to provide TCA for the foreign exchange market but it isstill in its infancy. The fixed income market is even further behind. Some fixed income instruments, such as sovereign debt, are relatively liquid and blessed with high levels of market data, but this is not the case for the asset class in general. There is no tick data available for fixed income and while a bond can be actively traded at certain times, it can also lie dormant for six months. There are some vendors offering a TCA services for fixed income, such as Livevol, based in the United States, but many other providers have chosen to leave that market alone.
The notion of providing TCA in a packaged format combining several asset classes is not something providers have yet seen much demand for.
“Most buy-side firms would prefer to see the TCA for separate legs of a multi-asset trade rather than as an aggregated average,” says Ian Domowitz, managing director of ITG, a provider of TCA. “For example, if I’m trading an inter-listed stock, there isn’t a cost to the security and a cost to the overlay, so there is no average.”
The technology issues around post-trade data and multi-asset trading are not the biggest issue for asset managers right now, says Domowitz, given that most of them will outsource the task of producing TCA data.
But is there any concern that the cost of this burden will be passed down to asset managers?
“My hope is that the information technology will become smarter at managing the data and that the market will be competitive enough to equal out any extra cost incurred,” says Mast.
Sheer weight
With the demand for both pre and post-trade data only likely to increase, there is some concern about the sheer weight of data. “We are close to breaking point in terms of data,” says Robin Strong, director of buy-side market strategy, Fidessa, a developer of order and execution management systems. These systems are typically processing more than 100,000 prices a second.
Is there, then, a point at which asset managers may say enough is enough and look to limit the amount of data they receive and the complexity involved?
It is unlikely says Mitchell. “Whether they like it or not, the data challenge will continue to go further. Data is increasingly being seen as the way for asset managers to gain competitive advantage as well as a key requirement for both regulators and investors.”
Nor is it likely that asset managers will look at toning down the complexity of their trading strategies because of the intricacy involved. As Kay says: “The front office always drives the back office and it would be a brave chief financial officer that said otherwise. So they won’t stop trading multi-asset, they will simply spend whatever is needed to keep up.”
Having said that, it is the brokers, not the asset managers, who will feel this burden. The sell-side has been slow to adapt to a multi-asset trading environment and is still segregated by asset class. This has shown in the underwhelming range of services that have been made available to buy-side firms, though Mast does say that brokers have shown some co-operation in terms of adapting to his firm’s multi-dealer platform by sending their quotes electonically.
Dedicated technology
According to Strong, it is specialist hedge funds that have developed dedicated technology to deal in multiple asset classes and exploit pricing inefficiencies. “The key for these hedge funds is to focus on bespoke pricing models that address very specific instruments. The rest of the supporting technology is relatively vanilla.”
Consequently, a number of traditional asset managers are acquiring large hedge funds in order to bring those niche capabilities in-house. “The traditional model has been for asset managers to employ brokers for these services. This is what we saw with algorithmic trading, where brokers would offer asset managers access to their algorithmic trading tools in exchange for a few extra basis points in commission.”
So, are we likely to see similar technology emerging for multi-asset trading?
“Brokers are absolutely capable of doing this but I suspect that there has only been a modest demand from the buy-side and not enough for brokers to devote resources to developing these kinds of tools,” says Strong. “But we do hear of brokers who are looking to be innovative and develop this kind of multi-asset capability so in time we will see how serious they are.”
©2011 funds europe
