Galen Stops takes a closer look at the new latency mechanism being introduced to FX and certain equity options products by Eurex, in a bid to improve the order book.
Speed bumps seem to be the topic de jour in the listed derivatives markets just now, with the Intercontinental Exchange (ICE) winning approval to implement a new asymmetrical delay on orders in certain metals futures contracts amidst much debate and controversy.
Eurex is planning to implement something similar for FX products and German and French equity options, but if there’s one thing that Jonas Ullmann, head of market functionality at the exchange, wants people to know, it’s that this is not a speed bump.
“It’s a latency differentiation mechanism,” says Ullmann, of what has been officially branded as Passive Liquidity Protection (PLP) by Eurex.
Of course, this might sound like splitting hairs, which is why Ullmann is keen to explain the details of how the PLP mechanism actually works in comparison to a traditional speed bump.
“The PLP can be enabled on a product level from our side, and what it does is check every incoming transaction shortly before it arrives in the order book to see whether it is aggressive or passive – aggressive meaning that the order is executable at entry in our matching engine and passive meaning that it is not. The aggressive orders are then parked on the sideline for a certain period of time before being put back into the order book to interact with it. Passive orders are just adding liquidity to the order book, so they are immediately processed and not delayed,” he says.
Crucially, no one in the market can see delayed orders, meaning that there is no possibility of anticipating those orders and no “last look” type functionality.
The PLP mechanism will be implemented for all FX products on May 27 and for German and French equity options on June 3. There is a crucial difference between the two, however, which is that since 2017 the FX products have already been subject to a “dual-gateway solution” which effectively implemented the same deferral mechanism on orders. Thus, for these products Eurex is simply replacing a hardware-based solution with a software-based one and not significantly changing how this market functions.
Improving the order book
By contrast, this latency mechanism is completely new to the equity options products, and it is being rolled out as a pilot programme to test its effectiveness in improving the order book.
“Right now, a lot of liquidity providers need to invest more into technology in order to protect themselves against other, very fast liquidity providers, than they can invest in their pricing for the end client. The end result of this is a certain imbalance, where we have a few very sophisticated liquidity providers that are very active in the order book and then a lot of liquidity providers that have the ability to provide prices to end clients, but are tending to do so more away from the order book,” explains Ullmann.
“Do we discriminate or differentiate between market participants? Clearly, the answer is no, because we treat every transaction, every order, the same”
He continues: “With the PLP mechanism we are trying to give this second group an easier way to provide liquidity in the order book, in order to improve both the size, spread, visibility – and therefore attractiveness – of the order book for everyone interested in trading our products, especially the end clients.”
The point here being that currently an end user looking to fill 500 lots might be looking at a screen and only seeing 100 lots available, when in fact the amount they want is available in the market, it’s just that many liquidity providers are reluctant to put their pricing on-screen for fear of getting run over by faster players. The hope is that the PLP mechanism will alleviate this fear and thus encourage more liquidity to migrate into the order book.
Ullmann acknowledges that such a mechanism has, as already noted, proven controversial in the derivatives industry, but he categorically rejects the notion that it creates an uneven playing field for market participants.
“Do we discriminate or differentiate between market participants? Clearly, the answer is no, because we treat every transaction, every order, the same. It’s like a matching algorithm, everyone needs to act within the rules of the matching algorithm, so if you have a price-time algorithm, it’s different to a pro-rata algo, and people might behave differently according to which is being used, but it’s very transparent and everyone is treated the same,” he says.
“On the options side, we will carefully evaluate this pilot to understand its impact. If the impact is positive, then we will discuss this with the market and try to understand if the positive effects can be repeated in options in other asset classes”
If this is true, then why are some firms so seemingly dead-set against latency mechanisms of any kind in the listed derivatives markets? The answer, according to Ullmann, is simply that right now some of these firms are using strategies whereby they benefit from being faster than everyone else because they can hit stale quotes or orders in the market and profit from this. And therefore, such a latency mechanism is simply not in their interests.
“But as an exchange we cannot consider all the different business strategies that firms might employ,” says Ullmann. “What we have to do is treat everyone equally, setting rules and mechanisms that benefit the overall market and ensure that the end-client is in a better position.”
As to the actual length of the deferral times on trade, Ullmann insists that these were not randomly plucked out of thin air.
Given that Eurex’s FX futures market is relatively nascent, the exchange initially proposed a 3 millisecond delay but, following further consultation with its initial clients, agreed to start with a deferral time of 8 milliseconds instead, which it claims reflects the lifecycle phase of the products. This deferral time will be regularly reviewed by the Eurex FX product committee, with a view to ensuring that it remains optimal for the order book as it develops.
On the options side – where the deferral time is 1 millisecond for the German equity options and 3 milliseconds for the French ones – Ullmann points to the very data-driven approach taken by the exchange when deciding these numbers.
For these, Eurex started out by broadly looking at reaction times and how long it takes for transactions from market participants to go through its matching engine. It then began to look at cases where there was a trade but the passive side tried to cancel it but was too late to do so because the price had already been hit, examining the gap in time between when the transaction occurred and when the cancel order subsequently arrived.
“We also looked at index options and fixed income options, but there was a strong signal from market participants that we should start with equity options”
“To put this into an example, let’s say that two people want to trade, one of them is passive and the other, who is much quicker, comes in and hits them with an order. Now the passive side decides that they won’t want to do the trade, but they’re not quick enough to remove their price before they get hit with the other side’s order. We’ve looked at the statistics around such events, which market participants call “too late to cancel”, looking at how often they’ve tried but been unable to cancel a quote. Now the numbers aren’t static, because some market participants might be just a few microseconds too late while other less sophisticated firms might be a few milliseconds too late, but based on analysis of this data plus other changes in behaviour that we anticipate, we have been able to arrive at the current delay times for the products in the equity options space,” says Ullmann.
Looking at other products
Indeed, it was data analytics and extensive consultation with market participants which led to Eurex picking these particular contracts for the PLP pilot, adds Ullmann. Part of the decision was just logic – testing out this mechanism on products that are not traded enough would prove little about its effectiveness, implementing it for one of the most widely traded ones risks disrupting a key market – but the exchange also looked at a number of different data points, and then shared this data with market participants in order to make it.
“We looked at a lot of different metrics: how is liquidity in the order book at the moment, how many participants are using it, what is the overall interest in the market for the product, how many pick-offs do we see and how prominent is a certain behaviour in the market. Then we elicited market feedback to find out in which products the market participants would be most bullish in supporting this initiative. We also looked at index options and fixed income options, but there was a strong signal from market participants that we should start with equity options,” says Ullmann.
One of the concerns expressed by some market participants regarding the aforementioned changes implemented by ICE is that such latency mechanisms could start to proliferate across exchange traded products, but this seems unlikely to happen at Eurex, at least any time soon. While Ullmann refuses to categorically rule out the prospect of one day implementing the PLP mechanism for other futures products, he’s also pretty clear that he doesn’t see any other applications for it on the futures side where it would provide a material benefit to the quality of the market.
“There is nothing planned on our side and, from the data, not much evidence that it’s needed,” says Ullmann.
He adds: “On the options side, we will carefully evaluate this pilot to understand its impact. If the impact is positive, then we will discuss this with the market and try to understand if the positive effects can be repeated in options in other asset classes. But again, the evaluation will be very data-driven.”