Seize the Advantage in the Race to Efficient Execution | Partner content | Asian investor

Transaction cost analysis (TCA) plays an important role in defining execution strategy in all environments of the trading arena.

Efficient trade execution can save clients and investors millions of dollars a year, dramatically improving the performance of fixed income portfolios and actively managed funds.

An evolving technology in a difficult market

At S&P Global Market Intelligence, a leading provider of fixed income market data solutions, we are at the forefront of delivering artificial intelligence (AI) technology, providing insights and essential analyzes for individual entities up to full portfolio analysis prior to execution, regardless of the market environment.

The emergence of cutting-edge cloud-based technology, internal contributor-based data, and state-of-the-art machine learning technologies, provides our subscribers with essential analytical tools to stay ahead.

Dynamic modeling for a changing landscape

Our AI uses a learning process that identifies changing market dynamics to select the best model to estimate execution prices and slippage, providing predictions and criteria for optimal execution against single traded entities to complex baskets, based on real-time market dynamics:

_{**For bonds with limited or no data, an approximation is calculated based on alternative bonds with similar yield pricing characteristics.}

Case study

A sample of data from 2020

We use a sample of 88,460 transactions from 1,669 unique investment-grade ISINs spread across 255 US Treasuries, 167 global sovereign bonds and 1,247 global corporate bonds from S&P Global Ratings B to AAA inclusive. There is a relatively equal weighting of 42,805 buys and 45,655 sells in the case study period:

Insight

The goal of S&P Global Market Intelligence Fixed Income Pre-trade is to provide essential and valuable tools to improve estimation of transaction costs and execution potential by understanding the dynamics behind effective execution in changing markets.

While the objective of OTC bond trading is to buy and sell as close as possible to the bid and ask prices, respectively, the mid-price quote will be used as a neutral marker. Costs in the form of price slippage will be represented as a negative value and conversely any improvement in the average price will have a positive value.

Hypotheses:

• The average price is calculated as the arithmetic mean between the best offer / offer from the average of the tenders provided by our Price Viewer

• Slippage is defined as the difference between the executed price and the average price in the price viewer Credit ratings are provided by S&P Global Bond Ratings\

• Total nominal size traded equal to $530 billion

• All data is converted to a base currency of US dollars for comparability purposes.

Market execution slippage density function

Using the currency-adjusted execution price as the standard for our Price Viewer’s average price slippage value, 39.13% executed better than the average price, comprising 41.78% of orders for buy versus slightly lower 36.78% sell orders.

The negative bias is very apparent, driven by this poorer performance over the period in terms of sell orders.

Overall, the trades were misexecuted based on an average 0.17% slippage of the execution price at the $0.15 price from the average price.

Zooming in on the interquartile ranges with the Box Plot illustration shows tighter execution between the 1st and 3rd quartile ranges, with a significantly narrower range between the absolute investment grades, UST, A and B.

Noise within the levels is expected based on the dataset, risk appetite impacted by Covid in 2020 and real-time liquidity.

The outliers in the dataset confirm that better execution could have been established with huge cost reduction results.

Fixed income securities before trading

An adaptive model for prediction

Changing market characteristics and drivers justify adaptation as essential in any predictive model, whether an estimate is required from a relative or independent perspective.

Removal of constraints and oversight freeing up the full capabilities of the AI ​​to self-calibrate without the need for endless optimization and targeting of parameters.

S&P Global Market Intelligence Fixed Income Pre-trade has one goal throughout its selection process: to arrive at a model that minimizes predictive error in real time.

Any number of inputs (independent variables) can be used in the calculation process, with the significance being determined internally and the optimal predictor being achieved.

Explore the percentage of variance explained

By transforming the available market data into a standardized space, removing any scaling bias, we can determine the importance of the inputs and their contribution to the definition of the regimes, thereby removing any multi-collinearity issues while isolating the drivers real-time market.

Transformed coordinates not only resonate in components with consistently similar characteristics by rating level, but also highlight trending factors within timelines.

Maximizing the percentage of explained variance along with minimizing the prediction error is the goal of obtaining the best estimator.

Mean square error and r-squared: quality controllers

To assess the prediction improvement using an adaptive learning model, we can consider the modeling accuracy by means of the normalized variance statistic, R-Squared.

Below are the relevant R-Squared values ​​between the variable dimensions and that of a selective mean squared error minimization model (called “R-Squared_Mininised”) in our machine learning process:

The advantage of an adaptive error minimization model over a pre-defined linear input model is clear: the result being the adaptive model that efficiently estimates execution price forecasts and hence slippage costs in real time. .

Estimation of the impact of execution on size

Using beta coefficients for a single ISIN, calculating the buy and sell aggressor estimates for a high quality A rated corporate bond, we can see the changing impact on the execution price and the cost estimate due to changes in run quantity only, using static input values ​​for all coefficients except run quantity.

For this bond, the relationship within the dataset yields a nearly linear relationship between the traded quantity and the execution price, when all other input variables are held the same.

For the smaller end of the quantity scale, this indicates that there is a possibility of a small gain in purchase execution relative to the middle, ie. an execution closer to the bid price, which would be considered to be executing at an average price from a risk aversion perspective, but the slippage increases at a gradient as the buy quantity increases.

In terms of sell execution, the situation is quite the opposite, with the impact of the quantity traded being absorbed, with negligible impact on the execution price as the quantity increases. This market could be called cheap.

The benefits of an adaptive learning model are not just limited to TCA prediction, but the generic construction of the model allows the end user to select inputs and outputs for any complex basket to perform a variety of essential pre-trade analysis, including a representation of current market status.

Identifying inherent dimensionality characteristics will undoubtedly improve returns and provide insight into changing market dynamics ahead of the curve.

Contact us for more information

_{S&P Global Market Intelligence
At S&P Global Market Intelligence, we understand the importance of accurate, in-depth and insightful information. Our team of experts deliver unparalleled insights and cutting-edge data and technology solutions, partnering with clients to broaden their perspective, operate with confidence and make decisions with conviction.}

_{S&P Global Market Intelligence is a division of S&P Global (NYSE: SPGI). S&P Global is the world’s leading provider of credit ratings, benchmarks, analytics and workflow solutions for the global capital, commodity and automotive markets. With each of our offerings, we help many of the world’s leading organizations navigate the economic landscape so they can plan for tomorrow, today. For more information, visit www.spglobal.com/marketintelligence.}

_{Copyright © 2022 by S&P Global Market Intelligence, a division of S&P Global Inc. All rights reserved. No content, including by framing or similar means, may be reproduced or distributed without the prior written permission of S&P Global Market Intelligence or its affiliates. The content is provided “as is”.}

_{spglobal.com/marketintelligence}