Regulatory prohibitions on direct market access within Asia-Pacific markets add to the legal and operational challenges and costs faced by international investors. The promotion of best practices across APAC markets is necessary to facilitate the standardization and control of algorithmic electronic trading strategies (“Algos”). Below we analyze certain structural features of the Asian Algo market, with a closer look at Hong Kong and Singapore to highlight key regulatory hurdles to the regional development of Algos.
Algos in APAC
High transaction costs
High transaction costs and transaction taxes persist as a barrier to market development, which adversely affect Asia’s competitiveness relative to global markets. Even the region’s free-market petri dishes (i.e., Hong Kong and Singapore) have unacceptably high transaction costs relative to other developed markets.
Current obstacles to becoming members of exchanges (eg., requiring a local presence) persist, resulting in many international brokers avoiding membership in Asian markets; instead, they must route their trades through locally incorporated brokers, which increases:
- credit risk for international brokers facing a smaller broker;
- operational inefficiencies due to additional back-to-back settlements; and
- a lack of transparency in technical platforms.
China comes out first on both the buy-side and sell-side, with India coming in second on both sides. Other markets that feature prominently on the buy-side are Singapore, South Korea, and Malaysia; on the sell-side, it’s Japan, Singapore, Malaysia and Indonesia.
In terms of what asset classes are trading with Algos, it is useful to highlight a few key points below:
- Actively Algo trading on the buy-side: cash equities; index futures/options; exchange traded funds (“ETFs”); single stock futures/options; and commodity derivatives.
- Actively Algo trading on the sell-side: warrants; fixed income derivatives; FX futures/options; and spot FX.
- Nobody trades over-the-counter (“OTC”) derivatives with algorithms.
The main reasons for Algo trading in Asia are: improving trading performance; pure profit; too many orders to trade manually; minimizing impact cost; and liquidity.
In terms of what Algo strategies are being used in Asia, the following are the most popular: volume weighted average price (“VWAP”); arbitrage; participation; implementation shortfall; pegging; list/basket; dark Algos; trend following; mean reversion; delta neutral; and pairs trading.
China and India remain the standout destinations for future Asian Algo trading, with rapidly growing interest in certain asset classes: OTC; ETFs; debt; and index futures.
Alternative Trading Systems (“ATS”) would help cut impact costs by operating as non-displayed venues. This is because details of orders are only put into the public domain after a trade is completed, which minimizes the potential for leakage of investor intentions alongside mid-point matching—both buyers and sellers benefit from such improved prices.
Hong Kong Algos
- Regulatory developments
Hong Kong’s Securities and Futures Commission’s (“SFC”) electronic trading rules came into force on 1 January 2014. The SFC’s electronic trading rules surprised the market in that they require the sell-side to make judgments about the buy-side’s ability to use electronic trading tools such as Algos, while also requiring the buy-side to make judgments about the sell-side. Requiring market participants to conduct this level of due diligence may backfire, resulting in a smaller roster of sell-side Algo providers. The SFC does not stipulate the level of effort required under the proposed due diligence obligations, which could be unduly onerous and inject an undesirable level of uncertainty into the market.
- Market structure
Hong Kong is primarily a long-only market. International players open up regional offices to access the China market, the region and to enjoy a favourable tax regime. There is no high frequency trading (“HFT”) on the Hong Kong exchange (“HKEx”) per se, but there is a large market in derivative warrants. The secondary market is weak and HKEx has been overly focused on its position as a gateway to China IPOs.
However, since Hong Kong is a regional financial hub, respondents trade in a number of markets. Singapore, Japan and South Korea are the most common markets on the buy-side, whereas Japan, India, Singapore and Taiwan are preferred on the sell-side.
- Asset classes
The most prevalent asset classes traded using Algos are cash equities, warrants, ETFs and index futures/options. Although there is a high level of algorithmic trading in exchange products, many of these are not HFT in nature. Spot FX is popular on the buy-side, as compared with FX derivatives on the sell-side.
Of the asset classes currently trading without algorithms, OTC, fixed income/debt, commodity derivatives, fixed income derivatives and index options are the most popular. How OTC and fixed income/debt asset classes evolve in their use of Algos going forward will be of particular interest.
There is clearly high demand for derivatives trading in Hong Kong, due primarily to the exemption from expensive stamp duties that impede HFT in the cash market. Furthermore, ETFs have been hugely successful in Asia, particularly in China and Hong Kong with the launch of cross-border Renminbi Qualified Foreign Institutional Investor (“RQFII”) ETFs.
Regulatory concerns over HFT appear too focused on minimum resting periods. Arguably, a greater focus on and enforcement of market abuses would probably be a more effective allocation of scarce regulatory resources.
In summation, while Hong Kong remains a regional financial center providing market access throughout APAC and into China, competitive concerns remain for the SFC to address, including:
- the HKEx monopoly appears to be stifling certain types of Algo trading;
- the local brokerages are engorged due to local presence requirements; and
- stamp duties are hindering the development of the secondary market.
- Market structure
The Singapore Exchange (“SGX”) leads APAC in terms of cross-border derivatives trading. Importantly, the SGX lists Asian benchmark indices on its own exchange. By offering such global access, Singapore is at the center of the ASEAN Trading Link. However, liquidity remains a major challenge and trading costs remain high.
There is a large presence of international firms, which offers global connectivity with Singapore as its base. Although China remains the most popular market, Singapore is also a major gateway to Indonesia and South Korea.
- Asset classes
The most prevalent asset classes for Algo trading are cash equities, ETFs, index futures/options, single stock futures/options and warrants.
The asset classes that are currently traded without Algos include fixed income/debt, OTC and index futures/options, and spot FX, FX futures/options, single stock futures/options and cash equities.
Most market participants build their own proprietary Algos in Singapore and across Asia. However, ‘white-labeled’ brokered algorithms are also active in Singapore.
Clearly, many strategies are being used in Singapore over a wide range of asset classes. The biggest challenges in Singapore’s Algo market are: liquidity; alternative trading venues; tighter spreads; and lower fees. Lower trading costs would help improve volumes (liquidity) at SGX. Ultimately, in order to promote increased competition and liquidity and allow further development of the Algo market, Singapore needs a true ATS.
From a regulatory standpoint, the minimum resting period for orders at SGX are not an ideal method for combating HFT abuses and this is probably too blunt of an instrument for monitoring and preventing market abuses. By contrast, reduced tick sizes are a welcome move in the right direction and the ASEAN Trading Link appears to be an excellent move toward improved regulatory consistency.
For the Asian Algo market to continue expanding in tandem with regulatory best practice, greater regulatory consistency and enhanced competition (ending trading monopolies) are essential. Both the buy-side and—especially—the sell-side recognize the need for better regulation of algorithmic trading, such as moving away from blunt tools (such as minimum resting periods) and increasing liquidity and competition. In particular, Asian regulators need to collaborate (like the US and EU) to allow greater substituted regulatory approval, whereby approval and regulation under another regional regulator is taken into consideration when assessing remote trading membership.