How Should a Firm’s Transaction Cost Analysis Framework Adapt for RFQ-Based Trades with a Systematic Internaliser?

Concept

A firm’s Transaction Cost Analysis framework confronts a fundamental architectural challenge when applied to Request-for-Quote (RFQ) based trades with a Systematic Internaliser (SI). The analytical tools designed to measure execution quality in transparent, continuous lit markets are structurally inadequate for the bilateral, on-demand liquidity environment of an SI. The core of the issue resides in the nature of the interaction.

A public exchange offers a continuous, observable stream of data against which performance can be benchmarked. An SI interaction, governed by the MiFID II framework, formalizes a previously opaque Over-the-Counter (OTC) relationship, creating a distinct trading venue with unique obligations and behaviors.

Engaging an SI through an RFQ protocol is an active, discreet inquiry for liquidity. It is a bilateral negotiation, even when automated. The firm transmits its trading intent to a specific counterparty that is dealing on its own account. This act of inquiry is the first and most critical event in the trade lifecycle.

A traditional TCA framework, which typically anchors its analysis to the moment an order is sent to a broker or hits a public market, misses the entire pre-trade discovery process. The quality of execution is determined not just by the final fill price relative to a market benchmark, but by the quality of the quote received, the time taken to receive it, and the market impact generated by the inquiry itself.

What Defines the Systematic Internaliser Environment?

A Systematic Internaliser is an investment firm that executes client orders using its own capital on a frequent, systematic, and substantial basis, operating outside the confines of traditional exchanges like Regulated Markets (RMs) or Multilateral Trading Facilities (MTFs). The regime was engineered within MiFID II to bring structure and transparency to the vast OTC space. SIs have distinct obligations; for liquid instruments, they must provide firm quotes when requested, making them a reliable, albeit bilateral, source of liquidity. This structure creates a hybrid venue that is part dealer, part exchange, demanding a bespoke analytical approach.

The interaction is fundamentally different from routing to a lit market. In a lit market, an order interacts with an anonymous central limit order book. In an SI RFQ, the order is a direct request to a counterparty who knows the identity of the firm requesting the quote.

The SI’s decision to price a certain way is based on its own inventory, its hedging costs, its assessment of the client’s intent, and its relationship with that client. Therefore, analyzing the trade requires a framework that can model counterparty behavior and the strategic implications of the bilateral exchange.

A TCA framework must evolve from measuring passive execution against a public benchmark to actively analyzing the quality of a negotiated, bilateral outcome.

This environment is further complicated by the composition of the SI landscape. The universe of SIs includes both large investment banks and specialized Electronic Liquidity Providers (ELPs), a classification that includes high-frequency trading firms. Each type possesses a different risk appetite, technological architecture, and hedging strategy. An effective TCA framework must be capable of differentiating between these counterparty types, as their quoting behavior and potential for information leakage will differ materially.

Strategy

Adapting a TCA framework for SI-RFQ trades requires a strategic pivot from post-trade measurement to a holistic analysis of the entire trade lifecycle, beginning with counterparty selection and the RFQ event itself. The objective is to build a system that quantifies the quality of the bilateral negotiation, a concept that traditional TCA is unequipped to handle. This involves designing new metrics that capture the unique risks and opportunities of the SI environment, specifically information leakage, quote quality, and counterparty performance.

The foundation of this adapted strategy is the recognition that every RFQ is a release of information. The primary risk is that a counterparty, upon receiving the request, may pre-hedge in the open market, causing adverse price movement before delivering a quote. A successful strategy, therefore, is one that measures this leakage and optimizes for counterparties and protocols that minimize it. This moves the analysis from a simple price improvement calculation to a sophisticated, game-theory-informed assessment of counterparty behavior.

How Do You Measure What You Cannot Directly See?

The central strategic challenge is quantifying events that are not explicitly recorded in a standard execution report, such as information leakage and the true opportunity cost of a rejected quote. The solution lies in building a data-rich analytical environment that uses high-frequency market data to create proxies for these invisible costs. The framework must measure the market’s state microseconds before the RFQ is sent and compare it to the state just before the quote is received and after the trade is executed. This allows the firm to build a statistical picture of each SI’s footprint.

This requires a significant enhancement of data capture capabilities. The TCA system must ingest not only the firm’s own order and execution data but also every quote received from every SI, including rejected quotes. Each quote is a valuable data point.

It reveals the SI’s pricing at a specific moment in time. Aggregating this data across thousands of trades allows the firm to build a proprietary view of each SI’s quoting behavior, response times, and fill rates under various market conditions.

The strategic shift is from asking “What was my slippage?” to “What is the total cost of my interaction with this specific counterparty?”

A core component of the strategy involves segmenting analysis by SI type. Bank SIs and ELP SIs operate under different models. Bank SIs may be managing a large, diverse inventory and have broader client relationships to consider.

ELP SIs are specialists in automated market making, operating with a high-speed, technologically driven approach. The TCA framework must categorize them accordingly, as their performance on metrics like hold time and fill rates will likely show distinct patterns.

A New Arsenal of TCA Metrics

An effective strategy introduces new performance indicators designed for the bilateral nature of RFQ trading. These metrics supplement, and in some cases replace, traditional benchmarks like VWAP or arrival price. They provide a multi-dimensional view of execution quality that reflects the nuances of the SI interaction.

• Quote Quality Score (QQS) ▴ This composite metric evaluates each quote received. It is calculated as a function of the quoted price versus the contemporaneous mid-point on the primary lit market, the response time in milliseconds, and the size of the quote relative to the requested size. A high QQS indicates a fast, competitively priced quote at the desired size.
• Information Leakage Index (ILI) ▴ This metric measures adverse price movement in the lit market between the time an RFQ is sent to an SI and the time a quote is received. It is calculated by comparing the lit market’s VWAP during the quote window to the pre-request VWAP. A consistently positive ILI for a specific SI is a strong indicator of information leakage.
• Discretionary Latency (Hold Time) ▴ This measures the time an SI holds an RFQ before responding. By analyzing thousands of trades, a baseline systematic latency (network and processing time) can be established for each SI. Any time significantly above this baseline is classified as discretionary latency, or “last look,” representing a period where the SI is deciding whether to fill the order based on market movements.
• Fill Rate & Rejection Analysis ▴ This involves tracking not only the percentage of quotes that result in a fill but also the market conditions at the time of rejections. Analyzing when an SI chooses not to quote or when a firm rejects an SI’s quote provides deep insight into the counterparty’s risk appetite and the firm’s own decision-making process.

Comparative Framework Traditional TCA Vs SI-RFQ Adaptation

The following table illustrates the necessary evolution of the TCA framework. It contrasts the standard metrics used for lit market analysis with the adapted metrics required for the SI-RFQ environment.

Execution

The execution of an adapted TCA framework for SI-RFQ trades is a data engineering and quantitative analysis project. It requires building the technological and procedural architecture to capture, process, and analyze the unique data streams generated by bilateral trading. This operational playbook moves from the strategic “what” to the granular “how,” providing a blueprint for implementation. The goal is to create a closed-loop system where TCA output directly informs and improves routing decisions and counterparty selection in real time.

The first step is a comprehensive data audit to ensure all necessary points are captured. The firm’s Order Management System (OMS) and Execution Management System (EMS) must be configured to log every stage of the RFQ lifecycle with microsecond-level timestamps. This includes the moment the decision to trade is made, the moment each RFQ is sent to each SI, the moment each quote is received, and the final execution details. Missing any of these data points renders a true analysis of leakage or hold time impossible.

What Is the Operational Blueprint for Implementation?

Implementing a robust SI-RFQ TCA framework is a multi-stage process that integrates technology, data science, and trading desk workflow. It is a systematic build-out of analytical capabilities.

1. Data Infrastructure Enhancement ▴ The process begins with the core infrastructure. This involves establishing a dedicated time-series database capable of storing and querying vast amounts of high-frequency data. The EMS/OMS must be configured to tag every RFQ event with a unique parent order ID, the specific SI counterparty ID, and the status of the request (sent, quoted, filled, rejected, timed out). All quote data, including price, size, and timestamp, must be captured for every SI polled, not just the winning one.
2. Benchmark Construction ▴ The next step is to build the right benchmarks. This involves creating a “Composite Quote Benchmark” (CQB) for each RFQ event, which could be the size-weighted average price of all quotes received. The system must also calculate the “Contemporaneous BBO” by querying a historical tick database for the state of the primary market’s best-bid-and-offer at the exact microsecond a quote is received.
3. Metric Calculation Engine ▴ With the data and benchmarks in place, a calculation engine must be developed to compute the adapted metrics. This engine will run batch jobs post-trade to calculate the Information Leakage Index (ILI), Quote Quality Score (QQS), and other custom metrics for every trade. The logic should be transparent and well-documented so that traders and quants can understand and trust the outputs.
4. Counterparty Profiling System ▴ A dynamic database should be created to store performance metrics for each SI counterparty. This system will aggregate data over time, building a detailed profile that includes average QQS, median ILI, typical discretionary hold time, and fill rates across different asset classes, market volatility regimes, and order sizes. This profile becomes the foundation for intelligent order routing.
5. Feedback Loop Integration ▴ The ultimate goal is to use the analysis to improve future performance. The counterparty profiles and TCA outputs should be fed back into the EMS’s smart order router. The router can then use this data to make more intelligent decisions about which SIs to include in an RFQ for a given order, balancing the probability of a good price against the risk of information leakage.
6. Reporting and Governance ▴ Finally, a new suite of TCA reports must be designed. These reports will de-emphasize simple slippage numbers and instead highlight the new, more meaningful metrics. They should allow traders and compliance officers to drill down into individual trades and understand the full context of the execution, including which counterparties were polled and why a particular one was chosen.

Quantitative Modeling in Practice

The core of the execution lies in the quantitative models that power the analysis. The following tables provide a simplified view of what the outputs of such a system would look like. They translate the abstract metrics into concrete, actionable data points that a trading desk can use to evaluate performance.

Table 1 ▴ Sample SI-RFQ Transaction Cost Analysis Report

Effective execution transforms TCA from a historical report card into a predictive decision-making engine.

This detailed report provides a far richer picture than a simple slippage calculation. It shows that while the firm achieved price improvement against the BBO, a portion of that gain was eroded by adverse market movement during the quoting process. It also quantifies the hold time, giving the firm a data point to compare this SI against others. This level of detail is essential for true performance optimization.

Reflection

The architecture of a truly effective transaction cost analysis system is a reflection of a firm’s commitment to understanding the market’s deepest structures. Moving beyond conventional metrics to build a framework that can quantify the subtleties of a bilateral negotiation with a Systematic Internaliser is a significant undertaking. It requires a fusion of quantitative discipline, technological investment, and strategic foresight. The data points and models discussed here are components of a larger operational intelligence system.

Consider your own firm’s analytical capabilities. Does your current TCA framework provide a complete picture of your RFQ-based executions? Can it distinguish between different types of counterparty behavior? Can it identify the hidden costs of information leakage or the opportunity costs of rejected quotes?

The answers to these questions reveal the robustness of your execution architecture. The framework detailed here is a pathway toward transforming your TCA from a compliance tool into a source of genuine competitive advantage, providing the clarity needed to navigate the complex, interconnected world of modern liquidity.