Giving customers the inside edge

Investors in Asian markets are beginning to experience the benefits of internal crossing as global brokers bring their version of the service to markets in the region. Takayuki Saito, executive director, head of Direct Execution Sales Asia Pacific, UBS, explores what clients should expect from the introduction of these services.

Internal crossing is making visible inroads into the Asian equity trading environment. In Japan, for example, over 70% of UBS agency flow passes through our IXT-PIN crossing engine; our presence in Australia is wellestablished, while planning for expansion into other markets is at an advanced stage.

As a global operator, the question we have to ask ourselves as we roll out our service in Asia is how far our internal crossing facility should be seen as a single global offering, given the different conditions pertaining across the region.

We begin with a general crossing model that allows for customisation to take into account the local crossing rules for each market. The crossing logic allows, in principle, for the possibility of liquidity sharing with different brokers and with emerging dark pools, as might be expected in US or Europe, but the degree of customisation required will depend on a number of marketspecific factors.

Learning from experience

Our approach has been not to rule out functionality designed for other markets even where it is not currently feasible to offer it. However, solutions are tightly correlated to the regulations and trading conditions in each individual market.

Even where market norms and regulations are broadly similar, differences in specific aspects can have a significant impact on the design of the service. In Japan, for instance, it is possible to cross within the spread; in Australia, it is not. The Australian market also has unique regulations on both the timing of crossing transactions and the percentage of overall flow that can be included. These needed to be taken into consideration at the outset in building a service for the Australian market.

Other peculiarities can be anticipated within each market, such as minimum transaction sizes and the appropriate venue for reporting the results of crossing activity. Firms that have tried to import and tweak systems essentially built for US and European markets have found such an approach ultimately less cost-effective than designing and building from scratch while drawing where appropriate on existing components.

Differentiation

Beyond the functional differences that reflect local market conditions, there are strategic decisions that providers of internal crossing facilities need to address. One is the nature of the internal flow that is allowed to interact with the crossing engine. Typically, we restrict access to agencyonly flow from within UBS. It is an almost universal observation that institutional customers are more comfortable in the knowledge that proprietary activity is strictly segregated.

The question then arises as to whether or not to allow a degree of interaction with external liquidity venues so that a portion of our crossing flow is visible to them and vice versa. Here, the decision may vary from market to market.

In Japan, sharing agreements with other brokers are unappealing, given our dominant position in the institutional market. In such an arrangement, we would effectively be net providers of liquidity to our competitors. There are, however, alternative independent execution venues that are entering the Japanese market.

One whose launch is imminent is kabu.com. We will be involved in executing flow on kabu.com from day one since that will expose our clients to the retail flow that makes up some 20% of the Japanese market. While sharing liquidity with rival brokers is, in our judgement, not in the best interests of our clients in Japan, sharing with an external provider that offers a feature not available anywhere else, such as tighter spreads or different types of flow is something that we feel we owe to our clients.

Measuring success

There is a ‘default’ perception that more crossing results in better performance overall, but this obviously depends on the type of flow involved. Ultimately, the deciding factor is whether or not best execution can be demonstrated as a result.

Two different brokers may, for example, be crossing 10% of their client flow. However, if one is essentially dealing with small orders that then go to the exchange, while the other is basing its crossing activity on parent VWAP orders or allowing clients to post directly into dark pools and achieve complete fills without ever being exposed to the market, the results on the same names are inevitably going to be different.

Ultimately, a broker should be able to go back to its clients and tell them how much they have saved in trading costs as a result of the crossing strategies adopted. Typically, in Australia, we aim to save clients 10% of their commission on a daily basis. If you demonstrate that to a client or show them with some confidence that you will save, on average, 5bps on every Telstra trade, for example, that is both a powerful validation of internal crossing and a persuasive sales tool.