Fidessa’s Hiroshi Matsubara, Co-chair of the FPL Japan Regional Committee, discusses the changing picture of Japanese best execution and the role technology (particularly TCA) will play in improving market efficiency.
The importance of PTS aggregation is certainly increasing, but it still faces several critical issues before it can develop further in Japan. The share of Japanese equities trading on the PTS venues has now reached around 6% and even exceeds that of the cash equities trading on Osaka Securities Exchange (OSE). The Fidessa Fragmentation Index (FFI) for the Nikkei 225 index constituents is now around 1.4 compared with 1.15 in July 2010.
The main PTS venues for institutional traders are Chi-X and SBI Japannext. The main benefit PTS provides as a trading venue is smaller trade tick sizes in decimals, which enable price improvements which cannot be achieved on the Tokyo Stock Exchange (TSE) as there are no decimal tick sizes on TSE. High frequency trading (HFT) models can make money more easily on a PTS with decimal tick sizes and a low latency trading platform, as offered on Chi-X and SBI Japannext.
So far, only a small number of domestic buy-sides are using PTS – mainly via broker SOR s (Smart Order Routing). Yet, even the more conservative asset managers in Japan are finding it difficult to ignore this opportunity for potential price improvement. More Japanese buysides are interested in trading on PTS venues, but the following factors are hampering on-shore buy-side traders in Japan from active usage of the lit alternative venues:
- The restriction of trading more than 5% of a stock off the exchange, in connection with the Takeover Bid (TOB ) rule. As a result, some domestic asset managers only use PTS for sell orders.
- Not all brokers’ back office systems are ready to accept and place decimal limit orders on PTS.
- The volume of shares being traded on PTSs is increasing, but it is not yet recognized that PTS venues have acquired the traditional price discovery function, thereby assuring achievement of best execution.
What is the next trend in algorithmic trading in Japan? Is smart order routing increasing or is the emphasis more on dark pool/execution algos?
In the two years since the start of arrowhead on the TSE, use of algorithmic trading has become a standard of buy-side trading, even for most domestic asset managers in Japan. Now, mid-tier domestic brokers are implementing algo execution services, as accepting Direct Strategy Access (DSA) orders has become a de facto prerequisite for attracting order flow.
I believe that the most typically used algo models are still based on VWAP and IS, but liquidity-seeking algos used to identify a trading opportunity for illiquid stocks are being used more.
Given the recent solid development of PTS venues in Japan, Smart Order Routing (SOR) technology is now a ‘must have’ tool for agency brokers, following closely after the algo engine itself. I anticipate that mid-tier domestic brokers will start adopting SOR in the next few years.
What are the best ways to access liquidity in these market conditions? How does technology help to meet this demand?
The development of alternative venues has altered the trading landscape in Japan. The role of SOR will become increasingly important in order to be competitive in this new environment. In terms of technology, SOR is needed in a multi-venue environment, but both the buy- and sell-side need to get more acquainted with tools for analyzing their execution quality. Each broker provides their transaction cost analysis (TCA) but from a buy-side point of view, they cannot rely solely on the broker’s tool for measuring execution quality. Buysides will be forced to adopt their own neutral TCA execution quality measure, which is another aspect of the technology required.
It is also evident that since the launch of arrowhead, order sizes are getting smaller in Japan, according to data provided by the TSE. This may create problems for traditional asset managers, where their liquidity requirements are more for block orders. You cannot avoid situations where some liquidity is missed, but those changes should be absorbed by the development of further alternative venues. This specific block order requirement will be absorbed by a dark pool like a Liquidnet.
Traders will notice very little difference for cash equities trading, as Osaka only handles a small amount of cash equity trades. Moreover, exchange trading commission is traditionally very low in Japan, so it is unlikely to benefit investors significantly, unlike Europe where consolidation reduces fees.
The main impact will be on derivatives trading, as the merger should mean consolidation of the major Japanese derivatives instruments. With the Nikkei 225 futures and options traded on the OSE and JGB Futures and options on the TSE it will be good to see the major derivatives instruments in Japan integrated onto one trading platform.
A single co-location centre between the TSE and OSE will create the lowest latency trading environment between TSE’s cash and OSE’s futures and options platforms. This will provide HFT firms with opportunities to exploit cash/futures arbitrage. Clearing and settlement for derivatives are currently separated between TSE and OSE and the merger will provide a more efficient platform for the back office flows of derivatives trading.
What should be done to prevent another trading halt similar to 2 February 2012? What other market structure questions still need to be addressed by FSA Japan and Japan Securities Dealers Association (JSDA)?
In reality, you cannot expect a completely trouble-free trading system on any exchange. Therefore, it is imperative to be ready for another situation where trading is suspended on the incumbent exchanges. The JSDA has announced that from 1 April 2012 off-exchange trading (including PTS) will NO T be halted in such cases. This should give the market greater confidence that trading on alternative venues will be possible if such a situation occurred again.
Even though this is the correct reaction, it does not mean that everyone will necessarily shift their order flow to PTS venues the next time there is a problem on arrowhead. Many of the market participants do not yet recognize PTS’s price discovery function as it is difficult to prove that potential executions to be made on PTS are the ‘best executions’.
It is interesting to note that when the London Stock Exchange (LSE) halted trading in November 2009, all trading in the market ceased without there being a significant increase of trading in MTFs. Despite any reservations, the ongoing trend of increasing liquidity in alternative venues (including broker dark pools) will not reverse in Japan.
In terms of the regulatory situation, I think that we will eventually see those restrictions currently levied on PTS trading relaxed. At the same time, I would expect the Japanese regulatory authority to create a more solid regulatory framework, which will encourage market participants to compete to achieve best execution.