The Road from Tokyo to Osaka


Mizuho Securities’ Spyridon Mentzas discusses the status of the Japanese exchange merger and offers thoughts on how well the two systems will merge and the benefits investors can expect.
The merger of Tokyo Stock Exchange (TSE) and Osaka Securities Exchange (OSE) is not yet finalized, but it appears they will merge in the beginning of 2013, with the details yet to be specified. The first impression is that they have nearly identical trading rules with some minor differences, such as the OSE trading until 3:10, while the TSE closes at 3:00. When TSE decided to shorten the lunch time in November, the OSE did the same. When one of the exchanges (usually, the TSE) changes the rules, then the other moves in tandem: for example, changing the tick sizes. If the merger does go ahead, it is likely that they are going to use the TSE’s cash system, arrowhead, and OSE’ J-GATE for derivatives. They will use the old systems in parallel, which will achieve a reduction in cost because they will not have to maintain two systems.
Further Industry Consolidation
The ECN’s in the US enjoyed technological superiority versus the classic exchanges, where NYSE’s latency was significantly slower than Arca’s. This would have been reason enough for TSE to consider buying a PTS, but with arrowhead’s current latency of less than 2 milliseconds (and another upgrade in the next few months to target less than a millisecond), simply buying a PTS would not give them a noticeable advantage because the TSE and OSE are on par with the PTSs. The reason why PTSs are increasing their market share is that, unlike in the UK and US, where Reg NMS and MiFID have required trading on the exchange with the best price, in Japan the PTSs draw volume through decimal points and smaller tick sizes than the incumbents.
For example, Mizuho Financial Group might trade on the TSE at 105 yen bid, 106 yen offer. That one yen spread is close to 100 basis points or almost one percent, whereas the PTS trades at 0.1 yen. This is a major incentive for investors to buy and sell on the PTSs with their smaller increments to reduce market impact and trading costs. From the beginning, the regulators have not been overly concerned with the PTSs deciding to trade in decimal places and have 0.1 yen ticks. It was always up to the PTSs to decide and the TSE could do the same. If anything, I think the new exchange would rather reduce their tick sizes, than merge again.
However, not all participants would be happy to see new tick sizes, for example, some of the proprietary houses or small firms that trade with retail, as altering their downstream systems to handle decimal places would be costly.
This will also create a fragmentation of liquidity in tick sizes. The bids and offers on the TSE are often thick, with something like 50 billion shares sitting on the bid side, so with 0.1 yen ticks, the average order size might move to 3 million or 1 million shares. Traders who want to buy a large lot will have to scroll up and down to find out how much they have to go up to absorb the available liquidity. I think for the traditional long-only traders, this might mean an increased scattering of liquidity. There is sufficient liquidity in the market at present; even for stocks trading at a low price – there are market makers trying to make 1% during the day. If smaller tick sizes are introduced, that liquidity will likely be scattered or disappear.

Regulatory Hurdles

The regulators are not likely to be overly concerned about this merger. As a listed company, OSE has shareholder requirements that the TSE does not. There is also a great level of confidence because both operators are Japanese. With a foreign operator, the concern for the regulators would be that they may pull out of the market if they do not realize the expected profits.
The exchanges in Japan also participate in self-regulatory corporation activities, where they monitor trading, market participants and stock listings. A PTS, on the other hand, is not self regulatory, so while they perform market surveillance, they are not required to perform evaluations on listed companies; they simply track which stocks are listed on the TSE and OSE. The exception to this is BATS who added this function once they were approved as an exchange. For the TSE and OSE, a merger will create efficiency and a cut in the overheads for these self-regulatory activities as they will not have to duplicate these efforts.
Improved Trading Conditions
As a broker servicing domestic, international and retail clients, we have to consider all of our clients’ needs. If there was only one exchange, we would have to connect to that exchange; if there were 50, then we would have to connect to all 50. From a broker’s point of view, if the exchanges do consolidate into one and that one exchange covers equities, futures, derivatives and commodities, then it will reduce our investment in connectivity and trading systems, not to mention membership fees. Also, when you consider high frequency trading and colocation, currently most brokers colocate at TSE for cash equities and at OSE for futures, but a merger will hopefully mean we only have to borrow one rack for both equities and futures. For the OSE, the data center for futures is in Tokyo and cash is in Osaka, so the best case scenario is to have one primary data center for both products and one for backup creating efficiency and cutting costs.
One other point to note is fragmentation. There is not much fragmentation in Japan compared to the US or Europe, where the incumbent exchanges are trading less than the majority of the volume. In Japan, adding together the TSE and OSE’ volume would result in more than 95% of total volume and marginally less fragmentation, which might in turn result in new foreign investment firms and brokers entering the market and thereby increasing liquidity.