StockMax and competition for HKEx
7 August 2011
There's been much ado recently about the pending launch by HSBC's securities arm, HSBC Securities Brokers (Asia) Ltd (HSBA), of an internal crossing engine called StockMax Crossing (StockMax) for securities listed in HK, to be made available to retail investors. The media coverage so far has failed to make the connection between this issue and that of minimum trading spreads on the stock exchange.
There's nothing new about crossing engines - these have been operating on the fringes of the market for years, accounting for less than 3% of the market turnover in 2010, but until now, the Securities and Futures Commission (SFC) has only allowed these systems to be available to institutional investors. Some of these platforms are "lit pools", where orders are displayed to participants (albeit anonymously), and others are "dark pools", where orders are not displayed, but participants hope that the system will match their order with one in the opposite direction at a mutually acceptable price and size.
Systems announced to date include CLSA's BlocSec, Deutsche Bank's DBATS, Goldman Sachs' SIGMA-X, Instinet's CBX Asia, ITG's Posit, Nomura's NX and UBS PIN (Price Improvement Network). There are also pools of pools, or aggregators, which seek to connect these internal crossing systems to combine the fragmented liquidity. These include Chi-East (a JV between Chi-X and Singapore Exchanges) and ITG's Posit Marketplace. Some of the pool operators have also established bilateral arrangements to join their pools together.
In regulatory terms, the systems are known as "Automated Trading Systems" (ATSs). Banks and SFC-licensed brokers can provide an ATS as a "Type 7" regulated activity. Those entities which are not licensed intermediaries, such as Bloomberg, can be separately authorised to run an ATS under Section 95(2) of the Securities and Futures Ordinance. A list of these is published on the SFC web site. HSBA is a licensed intermediary, so its ATS license is granted as Type 7, and there is a lengthy set of conditions on the SFC web site. Presumably the SFC has to be satisfied that all these conditions have been met before StockMax can launch.
Webb-site believes that the exclusion until now of retail investors from ATSs has been an unnecessary and protectionist measure for Hong Kong Exchanges and Clearing Ltd (HKEx, 0388), the listed company which owns The Stock Exchange of Hong Kong Ltd (SEHK). HSBC's StockMax, and others like it, should move ahead with allowing retail investor participation in their ATS platforms, because it will allow retail investors to benefit from better pricing opportunities. The SFC should allow it rather than U-turn and restrict StockMax to institutional investors.
As detailed in our article Arculli's rant against competition for HKEx (26-Jun-2010), SEHK currently has a statutory monopoly for the operation of a "stock market" in Hong Kong, so HKEx has been relatively sanguine about the small amount of volume conducted by ATSs. That's because ATS operators in HK have to report their trades to SEHK anyway, and HKEx also owns the clearing house, Hong Kong Securities Clearing Co Ltd (HKSCC) which clears and settles the trades.
ATSs are not on a level playing field with the biggest lit pool of all, namely the Automated Matching System (AMS/3) run by SEHK. As explained in paragraph 1 of the SEHK document "Operation Procedures for Trading Fee", an Exchange Participant has to pay the 0.005% trading fee on each side of a trade "whether in the Trading Hall and whether the transaction is arranged or concluded in Hong Kong or elsewhere". Of course, over 99% of trades are not conducted in the Trading Hall, so the wording is somewhat antiquated, but what they mean is that the Exchange Participant pays the fee "whether it trades on our system or not".
This "take-or-pay" arrangement is reminiscent of the anti-competitive practice which Microsoft once used to license MS-DOS to PC makers. Although other per-copy terms were available, the cheapest deal was a "lockout" deal which said that "you are free to use other operating systems, but you must pay us a license fee based on the number of PCs you sell, regardless of which operating system you use". In a 1994 settlement with the US Fair Trade Commission, this "per processor" license fee arrangement was banned because it had the effect of stifling competition. The requirement that all ATS transactions result in a trading fee to SEHK has a similar effect.
If and when HK gets a competition law, this take-or-pay fee arrangement might well be legally unsustainable, as it would be an abuse of SEHK's dominant market position.
What HK needs is a Consolidated Tape System (CTS) like that in the US, to which all operators of trading platforms (including SEHK) report their trades. In the US, the CTS is run for the Consolidated Tape Association (CTA) by the NYSE, but there is no reason why HK's could not be run by a neutral third party, perhaps established by the SFC and run not-for-profit, with operators contributing to the running costs proportionate to their turnover. Whoever runs the CTS, ATS operators should not have to pay SEHK a fee for each trade they report.
Trades matched by ATS operators have to be reported to SEHK within 15 minutes under Trading Rule 520 and 526, but that is an outdated rule designed for manually-matched trades. Automated systems should be able to do this much faster. To ensure that ATS participants do not have an information advantage, the SFC should require that no trade is reported to a customer of an ATS before it is reported to the CTS, and the absolute deadline for reporting automated transactions should be reduced from 15 minutes to say 1 second. If a participant in an ATS knows that his order has been filled before it is reported to the market via SEHK (or the future CTS), then he might take advantage of that by adjusting his bids and offers on SEHK or on other ATSs.
To avoid fragmenting market liquidity, HK also needs a Consolidated Quotation System (CQS) so that all displayed orders on any platform are unified, allowing brokers to satisfy their duty to find best execution for client orders. Again, there is no need to reinvent the wheel on this - the US has a CQS run by the CTA. For more detail on that, see our earlier article.
Slipping in dark pools
One of the perceived attractions of dark pools to institutions is the possibility that large blocks of stock can be traded without significantly impacting the market price, or what the professionals call "slippage". You might wonder whether that is more perceived than real. Intuitively, what should matter is how much real supply or demand there is at different price points, regardless of whether or how it is displayed. If there is a large block for sale, and a large potential buyer, then there's no reason to think that the price would move down any more than it would move up. If there is a large sell order but not much latent demand, or vice versa, then of course the slippage will be greater, dark pools or not. Even if you break the sell order into chunks, that won't create demand for the stock that isn't already there. So what the dark pools are about is not so much minimising slippage but accelerating execution. Rather than spending a long time breaking up your demand into displayed orders, you can probe the market with a dark order for the whole lot.
SEHK should counter this by introducing a "dark order" type of its own. Naturally, these orders would have to rank below displayed orders in terms of priority for execution at a given price (but ahead of displayed orders at inferior prices). Introducing this order type would allow SEHK to be more competitive with the dark pool operators.
Going inside the spread
But there's another big attraction of ATSs: if a stock exchange is inherently archaic and inefficient by requiring large minimum price-movements or "tick size", then that dictates a large minimum spread between the best bid and best offer price in the market. Competing platforms, lit or dark, can go inside that spread and allow people to meet in the middle. And SEHK is indeed archaic and inefficient when it comes to spreads. Currently, if a stock is on offer at $0.51, you can either take it or bid at $0.50, but nothing inbetween. That is a 2% spread. You might be willing to pay $0.505, and the seller might be willing to accept it, but you cannot do that on AMS/3 on SEHK. So off-exchange you go, seeking out a 1% price improvement on an ATS if you split the difference.
It is not clear whether StockMax would actually allow retail investors (rather than just institutions) to place bids and offers inside the spread - but it should, if it wants to attract extra business. A retail investor willing to pay $0.505 for a stock would see a lot of value in that. Equally, HSBC should allow retail investors to specify whether their order is displayed or dark. In other words, they should allow the same types of orders to retail investors as they do to institutional participants. Finally, a retail investor should be allowed to choose whether his order is routed to SEHK if it cannot be filled on StockMax, and of course, if the order is at an intra-spread price like $0.505, then it would have to be rounded down to $0.50 when it is passed to SEHK.
Small brokers are opposing StockMax, because they fear it will draw business away from them and away from SEHK's AMS/3. They fear that the sheer size of HSBC's retail investor base will produce a more liquid ATS in which significant volume could be matched and price-improvement offered. Institutions could also "jump the queue" by offering the same price as the top of the queue on AMS/3, undermining the time-priority that people receive when posting bids and offers in AMS/3. This is an exchange-centric way of looking at it. If you regard AMS/3 as just another ATS, albeit the largest one, with a 97% market share, then there is no reason why participants should not be free to join any queue in any other ATS.
HKEx is probably also concerned, because although (for now at least) it gets paid either way, it doesn't want to lose substantial volume to other platforms. But if its service was more competitive, then it wouldn't have so much cause for concern. Ironically it is the same factions which oppose reduction of spreads on SEHK that oppose the innovation being offered by ATSs. In an answer to the Legislative Council on 3-Mar-2010, Secretary for Financial Services Ceajer Chan Ka-Keung said:
"Dark pools have the benefit of bringing down trading costs, improving efficiency of trade execution and providing significant innovation in terms of trading services offered."
This is indeed true - and HKEx needs to respond to that innovation.
Cut the spread
Spread-reduction has been a long-running battle in HK, which Webb-site has been campaigning for since early 2002. Some small brokers fear that it leads to reduced rates of commission, because those who seek to day-trade by moving from one side of the spread to the other will make smaller profits and will therefor demand lower transaction costs. This is rather a false concern now that commissions have already been competed down to 0.08% or less for online brokers. Another reason that some oppose narrower spreads is that, with wide spreads, there is an order-posting race that goes on when the market opens, to get to the front of the queue on either side of a wide spread. With narrower spreads, there is less advantage in being at the front of the queue.
Your editor David Webb was an elected independent director of HKEx the last time this issue came up. We pushed hard, and the board of HKEx agreed to cut spreads across all price points, but broke it into two phases, then after pressure from small brokers, the board split the second phase into phases 2A and 2B, and then after more pressure, abandoned phase 2B in 2007 (2B or not 2B, that was the question). The result was the current spread table:
That left reasonably narrow spreads above $5, but left the majority of companies (by number) in the inefficient price band below $5. As of Friday (5-Aug-2011), only 279 of the 1460 listed equities were priced above $5 per share (a table ranked by price is here, updated daily). Not all low-priced stocks are small companies - Bank of China, for example, closed at $3.33 per share on Friday and was the 12th largest stock by market capitalisation.
HKEx should respond to this competition by reducing the spread table, particularly below $5, and making smaller reductions above $5 to make the whole table more consistent. Small brokers should support this, because it will reduce the attraction of ATS systems and stop the big retail banks and larger brokers from drawing clients away from the smaller firms in search of price improvement on ATSs. We propose the following spread table:
The small brokers
Out of the 526 Exchange Participants, The top 65 have about 89% market share between them, with the other 461 holding 11% (including 35 who don't trade at all). Given their small combined market share, you might wonder why small brokers are so influential on policy-making at HKEx and the SFC.
The answer lies partly in HK's political system. Participants of SEHK, together with the 184 participants in the Hong Kong Futures Exchange Ltd (which is also owned by HKEx) and the 171 members of the Chinese Gold & Silver Exchange Society together form the "small circle" electorate of the so-called "Financial Services" constituency in the Legislative Council. They also elect 18 of the 1200-member Election Committee which chooses Hong Kong's Chief Executive. Each firm has 1 vote. Asset management companies, and the 30,000+ humans licensed by the SFC, are not electors. So in effect, this is the "Small Brokerage" sector, not the "Financial Services Sector". The legislator also has a vote in the Election Committee. 19 votes out of 1200, or 1.6%, might not sound like much, but any Chief Executive candidate naturally panders to all such sectors to amass support. The legislator, and others in these narrow "functional constituencies" are also important to the Government in getting its laws enacted.
Another reason for the small brokers' influence is that the HK Government, particularly since the mass protest of 2003 which dislodged HK's first Chief Executive, Tung Chee Hwa, shrinks away from protest of any form. That's partly because their Beijing masters hate to see public protests for fear that mainlanders may mimic them more often than they already do. So if street protests are threatened or actioned by any organised group then the Government tends to concede, sometimes over-compensating by flinging cash at the problem.
But wait, you might think, "isn't HKEx an independent company, free to make decisions in its business interests"? No, because the shareholders of HKEx can only elect 6 out of 13 directors, the rest being appointed directly (6) or indirectly (1) by the Government. The Chairman of HKEx has to be approved by the Chief Executive of HK, and is currently a cabinet member. So although the Government only holds a 5.9% stake, HKEx is in most respects a Government-controlled company. In an earlier example, of this influence, HKEx on 20-Feb-2002 delayed the abolition of minimum brokerage commissions by a year. The Government cited the unemployment rate as an excuse, even though the sector employs far more people now than it did then.
For the sake of its competitive position in international capital markets, Hong Kong cannot afford to continue being held captive to a small but influential vested interest, whether it is the small brokers in relation to the Trading Rules of SEHK, or the tycoons in relation to other areas of reform, such as the Listing Rules.
HSBC's StockMax should be seen as a wake-up call to the Government, the SFC, HKEx and small brokers. If HKEx is not allowed to innovate, cutting spreads, introducing dark order types, and generally becoming more competitive, then we can expect more platforms like StockMax to emerge. The Government and SFC should not cave into pressure and resist this competition by prohibiting retail access to ATSs. The take-or-pay arrangement on SEHK trading fees should be abolished and the statutory monopoly on a "stock market" should be removed. Both are inconsistent with the principles of a level playing field and a competitive market in trading services. To facilitate this competition while maintaining fair and orderly markets, the SFC should oversee the establishment of a Consolidated Tape System for trades and a Consolidated Quote System for displayed orders.
© Webb-site.com, 2011