An Algo for the Hard Orders
By Hitesh Mittal, Founder and CEO, BestEx Research
A major trend we increasingly see is that fundamental discretionary managers are focusing most of their attention on their "hard orders”, those greater than 5% of ADV. While there is always room to improve, even on lower difficulty trades, those are moving toward schedule-based algos like VWAP, while time and energy are preserved for the hard orders. The hard orders are generally mid- and small-cap names with size above 5% ADV, or large-cap orders where the notional size creates real execution risk and PMs are demanding speed.
For these large trades, ATSs offering natural liquidity with minimal impact is more of a pipe dream than a reality. There is increasingly more volume in ATSs, but the diverse mix of market participants and their focus on market share over quality make information leakage virtually guaranteed and resting orders are easily picked off.
In our conversations with fundamental discretionary asset managers, we hear this is one of the most frustrating aspects of algorithmic execution. Traders are finding that the moment they put orders into "dark" algorithms, prices seem to move immediately, orders fill quickly when prices are moving in their favor, and the reversion after the orders are executed is obvious. So traders have adapted. Because they cannot trust most dark algos, they use them as tools rather than algorithms. They manage orders with tight limits, large minimum quantities, or both, submitting only a small slice at a time. The trader is doing the real work of execution.
While ATS-N filings have indeed made ATS operating procedures more transparent, transparency isn’t a trading outcome. The segmentation mechanisms now offered by many ATSs are complex, and in some cases, are designed to serve HFT market makers rather than institutional asset managers (as detailed in the filings and is visible in the results). Further, the internal "policing" at ATSs often ignores obvious metrics like measuring price impact after a conditional order is faded. The price change you feel is real, some of the time; we’ve measured it. And execution markouts published in segment definitions often measure the tip of the iceberg, since they are calculated over milliseconds with undisclosed time horizon. The filings themselves are a wealth of information and careful reading can reveal critical "opt-out" features, but it is practically impossible for buy-side traders to read 30+ ATS rule books, evaluate them, and then lead the process of algo change with each of their brokers.
This is a problem worth solving. Existing algorithms are not built with enough rigor for this use case. We know this is true, because when we started building a solution, the data, frameworks, and rigor we needed simply did not exist.
Curator is our solution. It is built on the idea that aggregating dark liquidity is no longer enough. It must be curated.
At BestEx Research, we believe the best solutions combine market structure understanding with quantitative, empirical analysis. Neither is sufficient on its own. Data alone is not enough because there are hundreds of ways to access liquidity and there will never be sufficient data to evaluate all of them. And just understanding market structure is not sufficient, because 90% of the narratives don't hold water when you look at the data.
And there are a lot of narratives. "Higher spread capture is better." "Trade more when price moves in your favor." "Use a large minimum size for every child order to minimize impact." "Trajectory crosses are the best." "Conditional orders help you access passive resting liquidity." All of these can be tested—we’ve done it—and not all hold true.
With Curator, we’ve read every ATS filing, talked to ATS operators, and optimized each access point. We’ve designed tactics for each ATS, providing vs. taking vs. conditional. We’ve built specialized tools to evaluate every tactic through the lens of information leakage and adverse selection, and then optimized the mix. We’ve optimized within each tactic using the appropriate venues and segments, minimum quantities, order duration, and then added a layer of proprietary alpha signals.
And we’ve discovered a lot of inefficiencies along the way. We’ve found that accessing the "best segment" is not always the most effective strategy. We’ve found that segmentation applies to only some of the flow in some pools and all of the flow in others, and that some order types lead to great markouts but substantial information leakage. And we’ve seen that a few seconds of post-trade markout is only the tip of the iceberg when evaluating fill quality. We’ve even found that trading in locked markets is still a thing. The list goes on.
Curating liquidity is chaotic; the rules and the liquidity mix keep changing, and they always will. For that reason, dark execution must be evaluated systematically based on core measures that represent our client’s definitions of success. We’ve documented those principles in two research papers, one on curating firm orders and one on conditional orders.
Importantly, we do not have 100% of the answers. But we believe we have built a framework meaningfully ahead of where the rest of the market is.
Traders using Curator should be able to trust the algo again, submit the order, and spend less time manually managing every fill. The problems we described above are not going away on their own, and we think this is a meaningfully better way to handle those hard orders.
We would love to walk your desk through our Curation Framework. We think the conversation itself will be valuable. The hard orders are not getting easier, and dark liquidity is not getting simpler. But the tools for navigating it can get meaningfully better. Curator is a step in that direction, a step toward replacing narratives with data and aggregation with curation.
Coverage in Markets Media: Why do Traders Babysit Dark Algos?
