Changing Tick Sizes In Japan
With Japan FIX Trading Community Regional Committee co-chairs, Junya Umeno, Director, Head of Trading, Tokyo, Blackrock, and Hiroshi Matsubara, Marketing Director, Japan, Fidessa
The implementation schedule of TSE’s “small tick-size program” has been divided into three phases. Phase 1 began on 14th January this year with narrower tick-size for TOPIX 100 instrument universe priced at and over 3,000yen. Phase 2 was implemented on 22nd July, introducing decimal pricing for stocks priced under 5,000yen in the TOPIX 100 universe. Phase 3 will review the results of the Phase 2 and decide if they are to implement optimized tick-sizes outside the TOPIX 100 names in line with the launch of the new arrowhead system, scheduled in mid-2015.
JPX summarized changes observed with the 39 issues that were subject to tick-size changes out of the TOPIX100. After the implementation of Phase 1, JPX saw significant drops in spreads for those 39 issues. The more highly liquid an issue, the more drastically the spread reduction was affected. The number of orders for Phase 1 issues increased significantly (by 2-3 times). The number of executions also increased but not as much as number of orders. There was a drop of the execution ratio immediately after Phase 1 but it is gradually coming back. No significant increase of trading values itself was observed after January as Phase 1 issues still account for around 25% on average of all the listed stocks after the tick size reduction.
From a market data point of view, it has become quite difficult to view the whole picture of market depth with smaller tick sizes. From market data users’ perspective, if client orders are algo traded then there is no need to follow market depth with human eyes. Though there is no significant impact on how to look at the data, the results of algo execution should lower transaction costs with smaller spreads. JPX says that there is no intention to price all listed stock in decimal tick-size, as they are adopting the phased periods to determine optimal tick sizes for the whole universe.
In terms of buy-side trading, as manual DMA trading becomes more and more difficult because of of these small ticks, more demands for DSA and broker discretionary algo trading is expected.
If you look at the FFI chart on the 30 names affected by the small tick program, they show around 0.15 constant drop in FFI since the small-tick implementation in January (see chart). This clearly means that the trading share of those 30 names is coming back to the main market from PTS. JPX says that the goal is to create a healthy trading environment that benefits all the retail and institutional investors with lower transaction costs. In the US, tick sizes are being broadened for illiquid stocks. In fact, traditional US tick spreads have been considered too narrow while Japanese ticks had been too broad compared with actual prices. Both markets are expected to land at the optimised tick-sizes despite those two opposite directions.
Some media reported that the tick-size reduction on JPX was targeted to increase liquidity from HFT. However in reality, many HFT firms (especially in regard to market-making strategies) have seemed to be refraining from trading in Japan since Phase 1. Those market-making HFTs are said to be reviewing and re-tuning their algos as they cannot make money out of the smaller spreads.
The ongoing debate around HFT in general does not make much sense as different HFT strategies are affected differently by market microstructure changes. Smaller tick sizes make it difficult to create profits for market-making strategies while HFT, with statistical arbitrage strategies, should be benefiting from this latest change in tick sizes.
The outcome of Phase 2 with the introduction of decimal tick sizes on JPX should be a big milestone for the direction of future liquidity fragmentation in Japan and it will be interesting to see what happens with PTS trading and HFT liquidity after this hot summer.