Remember me

Register  |   Lost password?


Chi-X Australia take off?

Mon, 07 Oct 2013 02:00:39 GMT

With the expiry of ASIC's transitional arrangements for best execution I think the dynamics of the Australian cash equities market is about to shift dramatically.

Thus far, Chi-X AU as been trapped in the first phase of liquidity capture: market makers being the predominant quoter of resting orders on the venue. While other brokers have been taking better prices when available, they predominantly quote on ASX.

The key barrier to Chi-X at this point is the reluctance for brokers to quote orders on Chi-X in preference to ASX. This motivation is understandable - up until now not all brokers have been connected to Chi-X, so you risk being left out of a trade if you don't quote on ASX. But this could change very quickly.

Now the tardy brokers have connected to Chi-X (well, just about all), we enter phase II of liquidity capture. In this phase a key advantage for Chi-X is activated: queue jumping.

Traders performing analytics on their trades will notice that resting orders on Chi-X yield a significantly higher return, as the queues are much shorter. Shorter queues mean higher probability of execution, and faster execution - with presumably the same adverse selection risk. In other words, you will extract more alpha from the market by resting orders on Chi-X from this point on. Remember there is an economic incentive for brokers on the other side of the trade to preference Chi-X as a venue when crossing the spread, due to lower fees.

As brokers move to adopt this approach, the advantage will be diluted to some extent - but will only reach equilibrium once the queue lengths at ASX and Chi-X are identical. Presumably the advantage for Chi-X will be present for some time to come.

Do resting orders on Chi-X actually produce more profit than ASX? I'm not sure about last month, but I would expect this month the answer is certainly yes. I am currently putting together the data set to provide the actual profit numbers to contrast the two venues now - but the only question in my mind at this point is the quantum of the advantage.

These dynamics are interesting in light of the recent market design changes proposed by ASX: removal of priority crossings and CentrePoint preferencing. I've commented in the past that I believe these changes will result in an increased participation in CentrePoint, to the detriment of the lit order book. I guess stealing orders from your own venue is better than letting another venue do it, so how well will brokers configure their smart order routers for mid-point (hidden) orders? This will be crucial in influencing where the orders will go - lit Chi-X or ASX, or mid-point orders on Chi-X or ASX. Four options - sounds like we need HF-TCA to be able to answer these questions with evidence!

What will happen from here? In a post-connected market, ASX will need to work hard to keep firms quoting on their venue - rather than getting cheaper fees and shorter queues somewhere else. In this situation - might we actually see a trade appear on ASX PureMatch, which could be thought of as their own "ASX QueueJump"? :)

I've been thinking about the fairness aspect of queue jumping. Whilst ASIC's supervision levy structure made Chi-X very cranky about it being "unfair" to their startup (by increasing costs on those who split orders across venues), I feel that the queue-jump effect will be of a higher quantum - and this time to their advantage. I do feel that queue jumping is unfair, this time to ASX, as it does hand a free-kick to Chi-X for traders wanting to get an order done. Of course the opportunities for queue jumping are wide and varied and include broker crossings and dark pool execution. The only way to address this particular unfairness is to mandate a central pool of liquidity (I dubbed NLN). Don't think that's going to happen any time soon. In fairness to Chi-X, ASX also have the opportunity to do exactly the same thing with PureMatch, as mentioned above.

by filmackay" />filmackay

, , , , , , , , , , , , ,