Are the markets crumbling or multiplying? With Josephine Kim, Director, Asia Pacific Electronic Sales – Global Execution Services, Bank of America Merrill Lynch.
How is market fragmentation changing and developing across Asia?
Japan had the first mover advantage in Asia as the first to welcome fragmentation, but it has not really blossomed, compared to Australia where we have seen a lot more volume growth on the new venues and execution channels. Hong Kong is quite interesting because most people really want to get fragmentation into that market, but there are limitations such as regulation and the proliferation of fees. Brokers do offer internal crossing engines, so a lot of clients are benefiting from those.
India is also interesting because it had two exchanges for years and recently got a third. We do not see that much activity, but at least it is creating a lot of buzz in terms of how fragmented the market is. It will be interesting to watch.
How are other Asian markets developing?
It typically depends on how the exchanges are preparing themselves. If you look at Japan for example, they opened up to various alternative venues and were pretty open-minded in terms of sharing their liquidity with other independent venues. But then the exchanges decided to merge to be more competitive as they realised that liquidity is something they want to keep. Australia is forcing the exchanges to share liquidity with best execution rules.
Korea and Singapore are the two markets most likely to come next, although there are other venues like Indonesia and Malaysia where a few broker-dealers offer limited crossing engines. Korea is a very active market, especially for futures and options, so a lot of people are interested in trading, but then again, they do have the investor ID restrictions and they are trying to implement a financial transaction tax so that is going to kind of hinder the attractiveness of trading.
Singapore is slightly different because the market depth is not as attractive as Korea. Singapore used to be a hub for US and European-based high-frequency trading firms, but it just seems to be losing its ground as a hub. Having the longest trading hours in Asia may open more doors for investors from different time zones but without the market depth, it will still be a challenge for Singapore to attract investors and independent liquidity providers. Minimum crossing rules also draws interesting opinions from people as some believe this will enhance and control the market participants and reduce toxicity of the pool albeit the overall reduction on actual crossing opportunity.
How does that variation across markets affect the trading environment?
The buy-side used to choose an execution broker based on the level and quality of research and their trading ideas. Today, the buy-side tends to go with the broker that has the liquidity. So, the buy-side traders are often watching the market and watching their stock, so they can see who is on the order book panel, and they tend to put their entire orders on the brokers with the most liquidity. A lot of this change is tied to unbundling, but it is also to do with the liquidity and facilitation as they often find it difficult to trade when there is less liquidity available in the market.
Are changes such as CSAs enabling that unbundling and enabling that separation between research and execution? Are these tools coming into existence to meet that desired change or are these tools enabling that change?
These tools are definitely opening the doors for the traders to choose from. It is simply an option that, because of this policy, the buy-side head traders have the independency to choose the best execution brokers and feel less obligated to trade with the best research providers.
On the sell-side everyone is becoming more liquidity sensitive. The buy-side trading instructions are becoming more complicated; the buy-side still want to have the baseline of a simple VWAP or POV as their first and second algos, but when the liquidity comes in, they do not want to miss out that opportunity, so a liquidity seeking type of smart algo, with a combination of base benchmark, seems to be being used more commonly.
How are the liquidity profiles of those venues changing?
Let’s use Australia, as an example. It is mandatory there to provide best execution to the client. That means that it is the broker’s responsibility to find the best execution price, across the dark or lit; it is not a choice anymore, it is an obligation. So, because of that, we started looking at the quality of liquidity pools. The number of liquidity pools has gone up a bit such as Chi-X Australia launching in 2011; and there are numerous exchange provided dark and lit pools – see Table A-1. The quality of each venue has risen as various enhancements or improvements have come online. We now care more about where orders are getting crossed within the dark liquidity, whether it is getting crossed at mid or better, or whether it is having any price reversion after the fill has been made. So, I think that quality is top of mind now.
If you look at the market share, we are still talking about a small portion of the pie – see Table A-2. One has to also have context of this and understand how dark liquidity is performing in the US and Europe – see Tables B1 and B2.
In Japan, 94% of volume is still on the primary exchange – see Graph C1. But if you look at the quality, and the savings gained by going through alternative venues, you can save about seven basis points. So again, the focus has been more on the quality of the executions.
Are the buy-side asking a lot more questions and are the brokers able to provide that data and what are the exchanges able to give back?
We have seen a lot more requests from buy-side firms; they’re a lot more aware and highly educated. They want to know where their orders are getting filled. Many of the top tier brokers, including ourselves, offer a real-time ping, so we can tell the client where the order got executed, whether it is an alternative venue or the primary exchange. From static data, providing a TCA of where orders had been executed, we are now moving into providing real-time intra-day pings back to the client. The broker always had access to where the pool was, even for the dark liquidity venues where we used a heat map so we could tell where the liquidity was gathering, but the buy-side did not have direct access. Now they really want to see what is happening on a real-time basis. It is faster and more dynamic because they can pull in, they can add more and they can change on a real-time basis.
How do you use all of that data and how do you keep track of everything?
It is quite interesting because although the market has been fragmented, we are now hearing about aggregated pools. I think it is just because a lot of people are thinking, “OK, liquidity is fragmented, so how are we going to capture all the pools?”
Ideally clients would like to get access to one venue where all the liquidity is aggregated. Whether it is the brokers coming up with a single pool solution or independent liquidity providers coming together as a hub for all the brokers’ pool, I think we are at that stage where we are really looking to optimise. There are a couple of things that, individually, as a broker, we are doing. A lot of broker-dealers are struggling because liquidity is centred on the big banks and everything is following the liquidity, so it is becoming less attractive for the buy-side to use broker-dealers for pure execution. As the market is becoming more fragmented and complex, we have seen more tools developed to keep up with the complexity, both buy-sides and sell-sides are investing more time to analyse data and achieve best execution.
Analysis such as toxicity analysis can identify where the clean and deep liquidity is before accessing the venue or, as the brokers, we can come up with an aggregated pool as well. But it is easier said than done. Having an aggregated pool is a great idea for the buy-side and it will probably work out cheaper for the sell-side than getting connected to multiple venues. However, a lot of the decisions around the standards, platforms and cost sharing models used to build it are hard to tackle.
Is there an opening as well in there for vendors?
When it comes to an aggregated pool, a third-party non-broker could be useful; Chi-East used to be in that role. They could pull the brokers together because it always helps to have an independent moderator. It should be a coordination and partnership between vendors and the broker as well.
Where do we go from here and what sort of timeline?
On the regulatory side, I think we will continue to see a lot of reviews. It is already happening in Australia. Singapore started looking at DMA as well. Hong Kong just released Consultation Conclusions. India is always looking at how it can best control electronic trading. Regulation will be a continuous topic over the next couple of years. Authorities want to look at how best to control electronic trading and the risks associated with it.
With exchanges, we are seeing consolidation. Australia and Singapore failed, but within Japan, the TSE and OSE merger, it is already happening. We are going to see a lot of the exchanges trying to find a way to stay competitive as the exchange is no longer the only place for execution. I do not think we are going to have a massive increase in the number of alternative venues. Having too many places will make the trading mechanism too expensive and inefficient. We might see some new countries joining the fragmented market structure but the actual venues are unlikely to grow in number any time soon. However, we will see more transparency in the quality of the source.
Some alternative venues tend to have smaller tick sizes than the exchanges. In Japan, for example, if the TSE changes tick sizes to be in line with alternative venues that will mean exchanges are becoming as attractive and we might see liquidity coming back to the exchange. That is something that the exchanges can look at and, if they do it, the alternative venues will have to look at something else that can attract more buy-side.