Accessing and aggregating liquidity is standard for algo brokers, but institutions truly need liquidity curation. In this paper, we released our dark liquidity curation playbook to institutional clients. This white paper consolidates analysis of millions of executions and reviews of hundreds of pages of ATS-N filings.

‍

Table of Contents:

‍

‍1. Introduction

The original promise of Alternative Trading Systems was simple—two institutions wanting to trade the same stock in opposite directions could match anonymously at midpoint, avoiding market impact entirely. For a time, that promise held.

Then it broke.

As ATSs competed for market share, they opened their doors to increasingly diverse participants, including HFT market makers, proprietary trading firms, retail flow from wholesalers, algorithmic conditionals. Volume grew. Fill rates improved. But so did two critical risks for institutional investors—adverse selection and information leakage.

Adverse selection occurs when an institutional order trades against a counterparty with superior information about imminent price movements, systematically buying before prices fall or selling before prices rise. Information leakage happens when trading activity signals an institution's intentions to other participants, who then trade ahead of the remaining order flow. Both erode execution quality, and both become more likely as counterparty diversity increases.

These risks are related but distinct. Adverse selection is about the trade just made—a trader bought, then prices fell because the counterparty knew something they did not. Information leakage is about the trades not made yet—others detected the trader’s buying interest and prices rise before their order is complete. One costs the fill that was received; the other costs the fills still to come.

ATSs responded to these issues with segmentation, dividing their pools into sub-venues and letting participants choose which segments they are willing to trade against. But segmentation schemes vary wildly across venues and the details of the methodologies are buried in dense ATS-N filings. This leaves a fundamental question unresolved…does segmentation actually reduce adverse selection and information leakage?

In this research paper, we answer that question and go beyond segmentation to answer a bigger question—the most important question of all: How can institutions get the dark liquidity they need in a fragmented market while avoiding the challenges that increase their execution costs?

To these ends, we analyze the full spectrum of ATS segmentation approaches, present original data on their effectiveness, and introduce liquidity curation methods that go beyond what ATSs offer. We conclude with a practical playbook for measuring, monitoring, and adapting execution approaches as conditions change.

The framework in this paper is the framework behind Curator, our dark liquidity aggregation algorithm. We built Curator around a simple premise: the industry has largely solved for fragmentation, but it has not solved for curation. Most execution algorithms can access most ATSs, but connecting to a venue is not the same as knowing how to use it. Which segments to access, which tactics to deploy, when to provide versus take, how to handle conditionals, and when to let signals override default behavior are all important decisions, and these decisions compound. Most algorithms treat them as an afterthought or not at all. In building Curator, we learned that liquidity curation is the edge, not just liquidity access. Everything in this paper reflects how we think about the problem and how Curator operates.

In addition, fragmentation has not fully been addressed. Private pools and certain buy-side-only segments remain inaccessible to most execution algorithms. We have developed a broker-neutral approach that allows buy-side firms to access these pools directly through Curator, provided they are willing to disclose their identity to the pool operator. This unlocks liquidity that would otherwise require using multiple brokers' algorithms or giving up access entirely.

This paper will address tactical execution questions around firm orders only, which are multifaceted and complex. Conditional orders are a second, critically important piece of the dark liquidity curation puzzle. Conditional orders follow their own unique rules and require distinctive consideration around adverse selection and information leakage as a result. For example, while conditional orders do not result in immediate executions and thereby avoid some of the toxicity firm orders may be exposed to, their structure of generating potential invitations a trader can decline creates an even larger risk of information leakage.  A second paper will be released shortly after this one to focus strictly on conditional orders. It will cover the mechanics, benefits, and drawbacks of conditional orders, as well as how to ensure they contribute meaningfully to (and do not detract from) execution quality. 

The days of simple block crossing are over. Navigating today's ATSs requires understanding the fifty shades of grey between "toxic" and "non-toxic" dark liquidity.

Request the full paper on the link below to continue reading

‍