Freaky Fridays – 4 March 2010

A few of us here at Fidessa Towers have been wondering about the spiky path fragmentation has taken over the past year or so. Whilst the overriding trend is upwards there have been a number of times when the rate of fragmentation has suddenly dipped only to recover its upward trajectory the next day. This was particularly evident with the FTSE 100 last month where the LSE’s average market share for February was around 40% and its average trade size was around 2,600 shares. Yet on Friday the 19th its market share suddenly jumped to 53% and its average trade size rose to 3,858. The Fragulator shows similar patterns in previous months when the LSE’s market share suddenly jumps on the third Friday of every month.

The clue as to why this happens could well be in the timing of the jumps, with the third Friday of every month coinciding with the expiry for equity options contracts. It’s likely that anyone in possession of an “in the money” option will seek to take - or force delivery of - the underlying asset (in this case, shares) on that day. This is then associated with a desire to crystallise the gain by liquidating the cash position and so the net result is a huge jump in equity volume traded on these days. The LSE actually facilitates this by having a special intra-day auction on these days and so, as a result, its market share gets a significant boost. Holders of such options seem uninterested in looking at alternative venues for better executions as they have already factored in their gains based upon the LSE’s price. I guess it’s a bit like finding out you’ve got a winning lottery ticket - you don’t really care what the cost of the cab ride is to go and collect your winnings.

If this analysis is correct then it’s another example of how the primary exchanges can use the concept of auctions to their advantage and effectively close out the alternative community. Obviously auctions are a useful and necessary part of an orderly market but the problem is how to make them work in a way that is fair to all venues concerned. BATS Europe, for example, has tried to introduce its own concept of an opening auction although this has met with limited success to date.

Anyway, this seems like another good “ask the audience” opportunity:

 Loading ...

Finally, thanks to Electronic Broker for your comment on the last blog, Tick Tax. It raises some interesting points about the risk of overburdening the provision of market data and the source of competitive differentiation between the primary and alternative venues which I’ll be exploring in more detail shortly - stay tuned.

Tick Tax – 25 February 2010

Just returned from a very stimulating week in Japan where I was asked to present on High Frequency Trading and global fragmentation at the Japan International Banking & Securities Systems Forum. There was a great point made by Chuck Chon, CTO of SBI Japannext, when he was asked how his alternative venue differentiated itself from the mighty TSE. Amongst other points, Chuck confirmed that SBI was committed to lower tick sizes as this has the effect of narrowing the bid-offer spread for any particular stock. He went on to say that crossing the bid-offer spread is like a tax that is levied every time anybody (retail or institutional) participates in the market. The challenge, however, is that traditional market makers actually like a wide spread as they are the ones who are effectively charging the “tax” in the first place. High velocity HFT firms, on the other hand, prefer very low spreads as they are constantly in and out of the market.

So primary markets have to decide whether to maintain a higher tick size, which keeps their traditional market makers happy, or bow to the pressure from the HFT community. This reminded me of a similar debate in Europe and especially in the Nordic market where NASDAQ OMX Nordic only reluctantly agreed to reduce tick sizes when they came under stiff competition from the MTF community.

This whole issue seems to highlight the differences between the traditional business models for equity trading and the new world order that is emerging. The bizarre thing is that, at least in the case of tick sizes, the interests of the HFT community and the retail punter seem very well aligned.

Anyway, my thanks to everyone at Fidessa kk in Japan and especially to Hiroshi Matsubara for being such a great host and guide.

Jumping the Queue for Order Flow – 19 February 2010

A couple of announcements have caught my eye over the past few days. This week saw the launch of a smart routing service called CYCLE from BATS Europe and, last week, I was at a seminar in Stockholm where Turquoise’s Adrian Farnham presented its own liquidity aggregation service - TQ Lens. Both of these announcements highlight how venues are starting to overlap with other players in their continual quest for order flow. The rationale is simple - by offering to onward route any order that doesn’t match on their own platform, these operators hope to encourage more people to use their service in the first place. It’s a tricky game though and potentially takes them into the space occupied by the broker community, i.e. finding the optimum destination for their customers’ orders. Of course, the brokers are offering matching platforms too - in the form of their own dark pools - that aim to provide more effective executions for some types of order flow. The net result is confusion over whether you are using any such service as a destination, as a smart route to other destinations, or both.

This raises some interesting technical challenges as any particular order may run through a number of potentially different pieces of smart routing kit before it reaches its final destination. It’s not obvious how the “smart” decisions made at the beginning of the order’s lifecycle will persist through these other systems (especially if they’re from different suppliers). This is different from the situation in the USA where Reg NMS forces venues to onward route any order that it cannot match at the NBBO (National Best Bid and Offer) and so there is an agreed standard for how venues operate in this space.

Whether you are a broker or a venue the objective is the same – to move ahead of your competitors in the queue for order flow. In the absence of any defined rules as to how to play the game it’s not surprising that both brokers and venues are looking to jump the queue any way they can.

Latency Wars – The Empire Strikes Back – 5 February 2010

Few can have missed the announcement this week that the LSE’s new low latency platform, Millennium Exchange, will be up and running in September. This comes just months after its acquisition of the Sri Lankan firm, Millennium IT, which supplies the technology and is testimony to the new thinking now taking place at the LSE. The same news item also mentions that the LSE has adopted a “self certification procedure” for the transition which will place the onus on the trading community to ready itself for the new platform. This is a canny move by the LSE as it enables them to regain the initiative and, at the same time, set the agenda for its members and technology partners. This is because both will need to focus on ensuring an orderly transition to the new platform between now and the go-live date.

Jeremy Grant’s article on FT Trading Room earlier this week highlights what all the fuss is about: latency and the growing number of High Frequency Traders who supply more and more of the liquidity in today’s markets. Whilst we can all debate whether the HFT phenomenon is a good or a bad thing, they are most definitely here to stay and have been a fundamental force in fragmenting liquidity both in the US and in Europe. Now that Tokyo has joined the low latency debate with the TSE’s new arrowhead platform it will be interesting to see if this ignites the fragmentation fuse in Asia too.

This headlong rush to low latency must of course be accompanied by due consideration for resilience and failover. Most technologists agree that there is a basic trade-off between speed and resilience - the faster you go the harder it is to put the car back on the track if something goes wrong. In the US, this is much less of an issue as the rest of the market provides resilience in the event of a glitch at any one venue. As we have seen in Europe, however, the lack of a consolidated tape and other factors mean that trading seems to simply stop in the event of an outage at a primary exchange.

On this point, it was interesting to see the results of our own poll on how best to solve the current issues surrounding the lack of a European consolidated tape. 60% favoured a collaborative approach as opposed to more regulation. With the stakes so high, and so many vested interests, it will be a great achievement if we can solve these issues ourselves without more “help” from the regulators.

The Luck of the Irish – 27 January 2010

Looks like it’s Ireland’s turn to run the gauntlet with the MTF community. Pretty much all the MTFs - Chi-X, BATS, NASDAQ OMX Europe and Turquoise - have announced that they will be trading Irish stocks from now on. For this reason we have now included the Irish Stock Exchange (ISE) and its accompanying index (ISEQ) on the FFI website. A quick look revealed that many Irish stocks are in fact traded on the LSE and so Dublin’s market share of its main index is less than 30 percent already.

This is a good pointer to the decoupling that is occurring between indices and their parent exchanges. In Dublin’s case it’s got nothing to do with MiFID but the simple fact that the LSE is the bigger marketplace for some of the large cap Irish stocks (which make up the ISEQ index). For other indices, such as the FTSE 100, the CAC 40 and the DAX, we are slowly but surely seeing them decouple from their parent exchanges and start to exist in their own right. The more this continues to happen, the happier the MTFs will be as this helps to break apart the concept that any given stock or index should “belong” to a specific exchange in terms of trading.

As this process continues, country-specific indices may become less relevant as we are all increasingly forced to adopt a more pan-European perspective. In time the S&P Europe 350, the FTSE Developed Europe Ex UK Index or sector-specific indices like the Dow Jones EURO STOXX® Telecommunications index may well become more relevant than the current crop of country-specific indicators.

I guess when this has happened you could say that MiFID will have done its job in terms of breaking down the national monopolies in equities trading across Europe.

Is it all over for Dublin as far as keeping hold of its current market share of ISEQ stocks? So far it’s too early to tell and, of course, there is nothing to stop the ISE reinventing itself as a pan-European venue and taking on the MTFs at their own game.

The T Word – 20 January 2010

It was interesting to watch Andrew Bowley’s video interview on Finextra regarding the launch of Nomura’s dark pool - NX. It helped crystallise the love/hate relationship between dark pools and the T word - Transparency.

Andrew Bowley made a good point about how the pursuit of MTF status for NX had been going on for a long time and wasn’t just a knee jerk reaction to the increasing regulatory pressure on the dark pool phenomenon. He also talked about how greater transparency was a good form of advertising for his brand and would help attract customers. This would seem to be the case, too, for Instinet’s dark pool - BlockMatch - which went down a similar MTF route. Other dark pool operators don’t quite share Andrew’s view on transparency although this is a function of their diversity rather than any attempt to be necessarily opaque. On both sides of the Atlantic there is a wide array of activities being labelled dark pools, all of them offering different services aimed at different market sectors. In the US, for example, dark pools operated by GETCO and Knight are market maker dark pools where the pool operator is always taking the other side of the trade. Others, like Liquidnet, are buy-side only crossing networks. In yet another category are the venue operated pools such as Chi-Vision or the LSE’s Turquoise (nee Baikal) that offer aggregation services by providing dark SOR connectivity to other dark pools. And, finally, there are the broker-owned dark pools such as SigmaX (Goldman Sachs) and Crossfinder (Credit Suisse). This is what is causing the regulators such a headache as it’s impossible to regulate effectively in such a diverse environment.

This is all part of a wider transparency issue in Europe that revolves around the lack of agreed standards for pre- and post-trade market data. I have commented a few times that an industry derived solution would be better than anything imposed from above. This, however, requires that industry players adopt a grown-up approach and collaborate when perhaps their instinct is to try and compete. With this in mind, it was interesting to read the European Data Consolidation paper sent to me by Andrew Allwright at Thomson Reuters. While it naturally has a Reuters sales bias it does succinctly highlight the issues and make a number of sensible recommendations that could form the basis of a long term solution. I’m not sure that all the proposals outlined will work in practice but this is a community problem and I believe we should be working together to resolve it. Naturally it will require some discussions about the detail.

Anyway, this seems like a good time for another “ask the audience” moment.

 Loading ...

Fragmentation Fever Goes East - 6 January 2010

Looks like 2010 is going to be a pretty interesting year in the battle for liquidity between the established exchanges and the more recently established MTFs and dark pools. It also looks like fragmentation fever is likely to spread eastward, too. This is being driven by a number of factors that include technology, regulation and commercial opportunism. The introduction this week of arrowhead by the Tokyo Stock Exchange will bring the performance of the TSE more in line with global standards and so open the door for the High Frequency community and brokers that can exploit microsecond price movements. This, in turn, will encourage the growth and use of PTSs (the Japanese equivalent of America’s ATSs or Europe’s MTFs). Currently PTSs only account for a fraction of Japanese equity order flow but the ability to arbitrage between them and the TSE may well provide the stimulus the market needs.

Meanwhile, the regulatory picture in Australia is set to change now that the Australian Securities and Investments Commission will have the ability to grant new exchange licences. This will open up its domestic market to other competitors and so allow Chi-X Australia, AXE and other alternative venues a realistic shot at gaining market share. Given that trading volumes are smaller than in Europe or the US, however, it will be interesting to see how much fragmentation the domestic Australian market can take and it would seem that any alternative venues will need to attract flow from other Asia Pacific countries in order to be truly viable.

The Singapore Exchange and Chi-X Global have joined forces to create the first exchange-backed dark pool in the Asia-Pacific region which will compete with the global dark pools already operated by the big banks. Put all these things together and they might be enough to set off a wave of fragmentation in that region. As we’ve seen in the US and Europe, once the change happens there’s no going back. Countered against this, of course, is the argument that without a single regulatory mandate for change across the whole region, the domestic incumbents should be able to fight off the newcomers one by one. This was certainly the case in Europe pre-MiFID where a number of well constructed initiatives (anyone remember Jiway, for example?) failed to wrest liquidity away from the primaries without the help of a regulatory imperative behind them. On the other hand, maybe the events of the past few years in the US and Europe have changed the trading landscape forever and so it is simply a question of time before we see a similar, fragmented, situation across the Asia Pacific region. Perhaps the biggest driver for change, however, will come from the big banks and brokers themselves. Having invested such huge sums in SOR and dark pool technology they will be keen to leverage this investment every way they can.

Happy Holidays – 17 December 2009

Now that everyone at Fidessa Towers is starting to wind down and prepare themselves for the Christmas break, I decided to look back at the predictions I made at the beginning of January 2009. As you can see from the table below, it looks like the FFI is a pretty good predictor of trends.

If the 2010 predictions are accurate you can expect the LSE’s market share in lit trading of FTSE 100 stocks to fall further from its current level of around 60% to something like 40% by the end of 2010 (excluding any volume from its imminent acquisition of Turquoise or from the revamped Baikal). It’s not all bad for the LSE, however, as it has finally acquired enough pieces on the board to take on both the MTF community and the other primary European markets too. Its widely anticipated acquisition of Turquoise will give it a pan-European MTF (already equipped with live order flow) right from the get-go and its decision to launch FTSE index derivatives will challenge the other primaries’ derivatives markets as well. Xavier Rolet should be applauded for getting the LSE to act quickly and let’s hope that his Christmas stocking is overflowing with the energy he’ll need to integrate such a wide array of different technology platforms through 2010.

This is in contrast to NYSE Euronext which has bet the ranch on its single UTP infrastructure. If you read the adverts, UTP allows the trading community to connect from anywhere and trade any asset class within its network. This will be great when it’s finally complete but NYSE Euronext will still need to prove that it can offer the best of all worlds rather than a compromise between everything. This is particularly true in the multi-asset space as the different nature of derivatives and equities trading make it pretty hard to tune one single engine to be best of breed at both. Meanwhile, Deutsche Börse has launched its own pan-European market (XIM) as its first foray in the battle for non-German liquidity in the equities space. Like the LSE, it will also be listing derivatives based on UK stocks. Looks like the big boys have worked out where the next battle is going to be fought. Unlike equities, however, derivatives contracts are created and owned by the exchanges that list them and so it’s much harder to wrestle liquidity away from the incumbent (NYSE Euronext Liffe).

Meanwhile, in MTF land, the prediction at the beginning of the year was that we would see the number of viable alternative platforms reduce to three. Over the past few months Chi-X, BATS Europe and Turquoise seem to have comfortably filled this space and trade between 15-20% of the FTSE 100 and other primary indices such as the CAC 40 and DAX. Assuming that Turquoise is acquired by the LSE then it looks like BATS and Chi-X will be left to fight it out in the alternative venue space together with NEURO (which is the only venue right now that also offers a smart routing service alongside its matching platform).

One of the predictions that didn’t seem so accurate concerned the impact of dark pools although, if the volume traded on dark pools matched anything like the column inches they receive, then perhaps things would be different. This point was made at a Dark Pools debate organised by the Centre for the Study of Financial Innovation on Tuesday where someone commented that if dark pools had been given a more benign name (e.g. Added Liquidity Venues) then maybe there would not be so much fuss about them amongst the media and regulators. This is true but only up to a point. Whilst the volumes in dark pools registered as MTFs are still modest there has been a significant (34%) upswing in other not-lit trading activity. This is where the regulators really need to focus since it’s increasingly difficult for anybody to see exactly how and where this liquidity is being traded.

Anyway, my thanks as always to the guys at Fidessa Labs for all their hard work this year and to everyone else who has participated in this site.

Happy Holidays - see you in January!

Whose Liquidity Is it Anyway? – 10 December 2009

Two events this week crystallised the need for better pre- and post-trade information.

Firstly, BATS joined ranks with Chi-X to question the LSE’s decision to set its market to auction mode during its recent outage.

In a related white paper on the subject BATS also claims that “it is questionable whether the market was orderly on the LSE given publication of data which did not represent legitimate trading interest or available liquidity”. It would be interesting to find out, therefore, why the LSE decided to put its market into auction this time rather than suspending it altogether (as it has when it has suffered previous outages). But, whatever the LSE’s motivations were, you only have to look at the widening gulf in fragmentation between the LSE and other European primaries to see how critical the battle for liquidity in London has now become.

The crucial issue, however, is agreeing a mechanism to provide the market with a pre-trade tape of prices that has an agreed “market outage protocol”, an agreed standard for deciding on fungibility, and a means of determining which venues to include or exclude from the tape. Without this it would seem that effective price formation when a primary market is down is still some way off. This is especially the case when there is uncertainty over if or when the primary market will reopen (a point that is acknowledged in the BATS white paper). Anyway, rather than deliberate on this, the guys at Fidessa Towers and I thought we would take the “ask the audience” option and allow you to make your views known to the trading community at large.

 Loading ...

The second development was the announcement that Nomura is going to reclassify its dark pool (NX) as an MTF and adopt a more transparent approach to publishing post-trade information by publishing its trades to Markit BOAT. Whilst Nomura is to be applauded for playing the game, it’s still only a partial solution to the problem. Other broker dark pools, such as BlockMatch from Instinet, have taken the MTF high ground, too, but still report in a different way (BlockMatch trades are printed to the Chi-X OTC tape, for example). The net result is that it is still pretty hard for the chaps back at Fidessa labs to assign all these dark trades to the right categories. This point was highlighted by CESR chairman Eddy Wymeersch who commented in the FT Trading Room article “we have very contradictory figures with regard to dark pools”. Maybe I’ll ask Santa to put a Fragulator in his Christmas stocking.

A point that all venues (primaries, independent MTFs and broker dark pools) need to remember, however, is that it’s not actually their liquidity in the first place. Markets have always been about trying to bring together willing buyers and sellers in order to meet the needs of both. So, in reality, liquidity belongs to them and not to the venues. What we need, then, is a clear set of rules for both pre-trade price dispersal and post-trade reporting. Only then will the real liquidity owners (market traders) be able to get a fair deal out of MiFID.

Should I Stay or Should I Go? – 27 November 2009

Few can have missed the furore over the LSE outage yesterday.

During the outage the MTFs made repeated attempts to convince the trading community that it should go and trade on them whilst the LSE was unavailable. The chart below, however, shows that they were pretty unsuccessful in this and that, instead, traders simply stayed away from the market altogether until trading resumed. This shows that, in London at least, traders are still reluctant to use MTFs without the comfort of knowing that the primary market is open at the same time. Market makers, too, are reluctant to make prices when the primary is down and so this further encouraged traders to stay at home rather than go and play on the MTFs.

Another factor in play this time, though, was technology. The majority of smart routers are (correctly) configured so as to direct all orders to a primary market when it is in auction. During yesterday’s glitch, however, the LSE put its market into auction which had the effect of creating an “artificial auction” that sucked up available liquidity from smart routers. This then led to the huge spike in trading on the LSE when it reopened at 2 pm. The net result was that LSE market share jumped by 7% compared with its daily average for the rest of November. This last point highlights, yet again, the ironic interplay in the post-MiFID world whereby the MTF community is still dependent on the LSE being open in order to try and increase its own share of UK stock trading.

Finally, as some of you may have guessed, it’s been “Clash week” here at Fidessa Towers. For those that don’t know, The Clash were pioneers of punk rock in London during the 70s. Their aim, together with other similar bands, was to disrupt the hegemony of the established music industry. Only time will tell whether the MTFs will be as successful in changing the face of European equities trading.