The Broken Market's Latest Creation: An Algo To Offset The Impact Of Other Algos

Define paradox: in a world in which market liquidity has become non-existent due to the pervasive presence of algos in every product, from stocks, to FX, to commodities and now to Treasurys (as the Oct 15 Treasury crash showed) who do nothing but churn all day long to collect liquidity rebates and just wait to front-run and subpenny whale orders, and yet the second a major sell order appears they all scurry as the "machines are all turned off" and we get a flash crash, what is one to do? Why put all their faith in an algo of course.

Define paradox #2: just whose algo is supposed to provide you will the much needed liquidity? Why that of the firm which blew up because its trading wires got crossed and on August 1, 2012 inverted bid and ask, promptly pushing itself into insolvency as its own "liquidity" evaporated in minutes and ended up being bought for pennies on the dollar by the like of Jefferies and Citadel. The firm, of course, is Knight Capital, or as it is now known, KCG.

And because two paradoxes make an unparadox or something, Trader's Magazine reports that "as part of KCG's continued push into the institutional trading side of the business"... which is a euphemism for please trade with us: we won't blow up again, we promise... "the well-regarded and historically focused market-maker [ZH: if you keep repeating that it magically comes true, just ask world-renowned trader Dennis Gartman] has built its first brand new algorithmic trading tool - Catch."

What does the algo known as Catch do? Well, supposedly it offsets the impact of all other algos who have crushed market liquidity. Translation: this is a "good" algo.

The firm has developed Catch - an algorithm designed to find what Susi and KCG term "higher quality" liquidity. Higher quality liquidity, he noted, could be defined as order flow that reflects little to no market impact after and execution - no market movement, reversion or footprint. Thus, Catch is meant for buysiders who are quietly searching for that elusive block trade and/or natural fill - not the small predatory orders or high-frequency traders who look to sniff out institutions larger orders and get out in front of them.

"Elusive" being the key word of course. And we won't even comment on the irony of an algo admitting its sole purpose is to offset what createors of every other parasitic, predatory algos deny exists: HFTs which "sniff out institutions of larger orders and get out in front of them." Maybe there is another irony somewhere that offsets this one. Just like the paradoxes.

So, how does the algorithm work?

In simple terms, Catch casts a wider net for an order, leveraging a broader set of tools when chasing liquidity. Utilizing thoughtful, passive order placement logic and an advanced fair value model, Catch is empowered by KCG's Big Data analytics and low-latency routing technology.


Catch is guided, but not governed, by a market-aware participation strategy that uses adaptive participation guidelines. This strategy influences the urgency of trading, but will not force trading at inopportune times, as can be the case with algorithms coded with hard bands.


Susi explained that the algo is designed to act aggressive early on. That is, volume participation will be greater at the beginning of the order, getting a user closer to the fill sooner. As the algo runs, it continuously recalibrates. Catch manages opportunistic passive and aggressive trading by considering urgency, inventory, and market conditions in real-time. Therefore, its participation is continually recalibrated as an order progresses to ensure proper exposure and optimal execution performance.


He added that Catch is rooted in the firm's market making technology, market latency experience and now big data technology. The result is an algorithm which helps the buyside find alpha at all levels - especially incremental or "micro" alpha at the child order level.


Uh, Knight Capital - which again blew up due to its market making going horribly wrong - pitching something as being rooted in the firm's "market making technology, market latency experience" is probably not the best approach.

The said, we can't help but note the biggest irony of all: first of all HFTs destroy market liquidity, and now they "sell" products meant to circumvent the zero liquidity they themselves have created. Does that qualify for a market monopoly yet?

In conclusion, however, we are very much relieved to note that when even HFTs begin offering products that seek to offset the impact of other HFTs, than the market structure wars are finally coming to an end as the cannibalization among the parasites has reached the endgame.

Now if only central banks could follow suit.