Jumping the queue (in a good way)
Published on May 12, 2008.
Institutional participants in Asian markets, used to sitting in long queues at the exchange, are natural targets for crossing solutions, says Brook Teeter, director, equities, Advanced Execution Services, Credit Suisse in Hong Kong.
Although viewed as a region for purposes of geographical breakdown, Asia is home to diverse markets and equity trading environments. One common factor, however, is a reputation for wide spreads and long queues. Investors are therefore naturally receptive to services that brokers can offer to improve the trading conditions for individual orders.
One technique, honed in US and European markets, is internal crossing. Several major brokers have recently been promoting their own internal crossing engines in individual markets within the region, while third party and agency crossing facilities have also been introduced to parts of the region over the past year. CrossFinder, Credit Suisse’s internal crossing engine, is currently available in Japan, Hong Kong, Singapore, Australia and Korea.
Anonymity
While internal crossing is hardly new in itself, the automation of the process removes one of the drawbacks of traditional telephone- based broker services: the need to reveal trading intentions to a human being outside one’s own trading desk.
From the client’s point of view, participation in internal crossing is essentially part of a quest for a better price and faster completion, while for the broker it is a tool to help provide best execution. To garner credibility as a value-added service, any internal crossing facility offered by a full service broker must be demonstrably free of influence from within. At Credit Suisse, for example, all cash, program and other electronic flow goes directly into a black box that no one, including Credit Suisse staff, can see into. Clients are aware if an order has been crossed, since they’re not getting hit on the exchange, but we do not receive a report on crossing activity until the end of the day. That shows the number of orders and execution value that have been crossed. It provides an aggregated view and does not include information on the specific clients or the exact orders that have been crossed.
In these circumstances, the participation of proprietary flow in the crossing process may be seen as a less sensitive issue. However, while some clients are sufficiently comforted by the ‘black box’ approach, others are strategically opposed to any interaction with proprietary flow. Where we have introduced CrossFinder in Asia, therefore, we have turned off prop flow participation.
CrossFinder operates as a continuous limit order book without displaying the book. Executions occur according to price and then time priority. A client with a large order in a particular market may, for example, find themselves at the back of a queue on the exchange with effectively no chance of getting done by the end of the day or, alternatively, completing an order but at an unattractive price. An effective crossing facility will provide a realistic chance of an earlier fill at a price specified by the client on submitting the order or better, depending on what is available in the crossing engine from the other side.
Competition
An obvious issue in markets where demand for liquidity outstrips supply is whether individual crossing engines should be encouraged to interact. In Europe and the US, we have hooked into several other pools of liquidity and are open to doing the same in Asia. We are not proactively opening up our pools to other full service providers, but if it were to provide a better price for our customers, we’d be open to that conversation. Not surprisingly, other full service firms are reticent at this stage.
We do, however, believe that there is room in the Asian markets for third party alternative trading systems, whether in the form of ECNs or other matching facilities. CrossFinder, for example, integrates with Instinet’s CBX in Japan. An order entering CrossFinder will search for a match in both systems and the order will be filled in whichever system provides a better price for the client. If there is no price on either, the order is then routed to the exchange.
Volume expectations
It is at this stage not possible to suggest a typical percentage of an order that clients should expect to be able to cross before going out to the exchange. Estimates will vary from broker to broker and market to market in Asia.
In general, we see 12-15% of our flow being crossed in Japan on a daily basis. On an average trading day in Hong Kong, the percentage is lower – roughly 5-6% – reflecting the relatively lower volumes in the market, though on an index rebalancing day, of which Hong Kong has quite a few, that number shoots up significantly.
Obviously, the more volume that is pumped through an individual crossing engine, the better chance each participant has of finding a match. Clients should perhaps look to see what percentage of volume each broker accounts for in a market as an indication of their chances in a particular engine. There are at the moment only a handful of broker crossing platforms available across the region, but as take up of internal crossing expands in Asia, the volumes that can be crossed with price and time improvement is only likely to increase.
