From the Trader’s Desk: India and Hong Kong
RCM’s Head of Asia Pacific Trading, Kent Rossiter, points out some of the good and bad of Indian SOR and reflects on Hong Kong market structure.
Are Smart Order Routers (SORs) in India working well?
SORs sure are working in India. I am not sure what is more of a raging success in the Asian equity SOR world, India or Japan, but the cost savings estimate numbers we are hearing are evidence enough to suggest that Indian SOR development is a big plus.
For ages, there have been two meaningfully big markets; the Bombay Stock Exchange (BSE) and National Stock Exchange (NSE). Up until a year ago, when Securities and Exchange Board of India (SEBI) opened the playing field up, investors who wanted the liquidity of both had to do so by manually monitoring their screens. This was painfully labor intensive and with the thin displayed liquidity of bids and offers, difficult to actually execute. You would often find fills from one exchange or another being executed at inferior prices to the other as a dealer had their eyes off the ball. Those executions were inevitably followed by a conversation with a dozen excuses. I would be told what I was seeing on my screen was not the real situation, but a latency delayed picture.
For the most part we are only using brokers with SOR for our Indian executions, and these brokers co-locate servers so latency is no longer a concern. We are getting fills at the best prices available and from two pools of liquidity where we may have only had one in the past. Only if the order is really small would we limit ourselves to one exchange in an effort to save on ticketing charges.
SOR is just the most recent visible step in the broader trend of the evolution of markets. Accordingly, the buy-side and sell-side traders have to educate themselves and keep up.
What are the issues with Indian SOR?
It is the lack of interoperability at the post-trade clearing level that has limited the true savings many investors would have benefited from otherwise. This is a challenge that SEBI continues to address. The lack a central clearing counterparty for the NSE and the BSE causes settlement costs to be about twice what they would be if only one exchange were used, and this is a consideration for most institutions when deciding whether or not to use two exchanges. If the exchanges and SEBI could reach a solution in terms of interoperability arrangements for SORs, the cost savings and benefits of SOR usage could be passed to the end users. Until then, its true potential remains yet to be uncovered.
What do you hear from your brokers in India?
India is a unique market with the way stocks are quoted out to two decimals, yet market depth is often extremely thin. SOR and algo usage have greatly tightened those spreads as systems can refresh and adjust quote prices quickly. Being able to access liquidity on both venues increases the total liquidity at hand as compared to other markets where flow is concentrated on the primary exchange. One broker’s estimate of India’s NIFTY50 stocks puts savings in a range of 5-8 bpts on average for clients using their SOR.
Many trading systems used by buy-side trading desks are unable to split an order that is executed across two venues for confirmation purposes. While RCM’s EMS can handle dual markets seamlessly, some of our peers are using systems which do not accommodate multiple exchange fills. One of the largest brokers with SOR connectivity in India has told me that only about 10% of their clients are demanding dual exchange fills, which is surprising as that same broker says using a SOR in India has lead to an average performance improvement of about 10 bpts. This broker says the worst case savings is as low as 2 bpts, and they have had discussions with some clients as to what the minimum order value needs to be before it makes sense using both exchanges.
That is dependent on the clearing and custodial fees of the client, but generally the order will be $250k or more. Unfortunately it is not as simple as setting a minimal threshold because it would vary widely on a stock by stock basis, and is not consistent as to which stocks provide the least or most savings.
What is the status of broker internalization engines in Hong Kong?
Internalization engines have been under much controversy in Hong Kong of late. HSBC’s StockMax internalization engine was to include retail flow, but their license was amended at the eleventh hour limiting use to only professional investors. Many institutional investors feel this is too bad, and hope it is a temporary delay as regulators study the pro’s and con’s of off-exchange liquidity. As I understand it myself, HSBC’s retail investors themselves had been expecting to have their orders benefit from StockMax pricing. They expect their trades to get the best price, not just what’s being quoted on the primary exchange.
Hong Kong Exchanges and Clearing (HKEx) currently enjoys a statutory monopoly for the operation of a ‘stock market’ in Hong Kong. Alternative Trading System (ATS) operators in Hong Kong have to report their trades to HKEx, and also exchange partner’s have to pay 0.5 bps for each side of the trade regardless of whether or not the transaction was conducted in the HKEx or through an alternative liquidity venue. Moreover, HKEx owns the clearing house, Hong Kong Securities Clearing Co (HKSCC), which clears and settles the trades, another expense to ATS. With the thin margins of many dark pools these costs can add up quickly.
In Hong Kong the Securities and Futures Commission (SFC) has begun requiring brokers with internalization engines (Type 7 license) to identify those executions done in their internalization engines. I think I speak for my peers in saying we would like the HKEx to make this information public. At the moment there is no differentiation seen on the reporting tape for the types of off-exchange prints.
How does this compare to broker internalization engines in other markets?
The amount of liquidity in the various broker’s internalization engines really differs. In Japan and Australia for example, there is no stamp duty, so frictional costs are lower. Depending on the broker, there could be a lot of HFT strategy flow included in those engines. Hong Kong, on the other hand, even with its wider spreads, does not have as much HFT participation, but still has plenty of dark pool activity. Another incentive for market-makers to trade in Japan is their ability to price inside the spread with just a touch. The cost of getting price priority in Japan is relatively cheap.
All investment involves risks. Past performance is not indicative of future performance.
The views and opinions expressed in this document, which are subject to change without notice, are those of RCM Asia Pacific Limited and/ or its affiliated companies at the time of publication.
Some of the information contained herein including any expression of opinion or forecast has been obtained from or is based on sources believed by us to be reliable, but is not guaranteed and we do not warrant nor do we accept liability as to adequacy, accuracy, reliability or completeness of such information obtained from or based on external sources.
The information contained herein including any expression of opinion is for information purposes only and is given on the understanding that it is not a recommendation nor investment advice and any person who acts upon it or otherwise changes his or her position in reliance thereon does so entirely at his or her own risk without liability on our part.
This is not a recommendation or offer to buy or sell or a solicitation or incitement of offer to buy or sell any particular security, strategy or investment product.