Case Study: Trading Risk Management
The work described in this case study can be categorized as Business Reviews and Diagnostics
The Client and Scope of Work
EG Consults had a number of clients focused on trading risk management in different areas of a major investment bank. These included Complex Derivatives Trading, Corporate Credit Portfolio and Capital Analysis. This highly quantitative and data intensive work required partnership with:
traders that owned the risk of the portfolio
quantitative analysts that developed and calibrated the models used in valuation
technology and operations teams that managed the supporting risk infrastructure and produced results
EG Consults developed accessible and intuitive analysis of risk and profit leading to agreement on findings, recommendations and next steps.
The Value Proposition
Trading risk management aims to establish a framework for disciplined risk taking. Traders, the actual risk takers, execute specific transactions and manage a focused portfolio of positions. They are directly responsible for profit and loss. There are a number of trading oversight functions, each with a specific focus. For example, a risk manager may be responsible for hedging the risk of several trading desks. At the corporate level, a firm-wide view (across many types of trading desks) is required to manage the risk of the bank as a whole, to allocate capital among competing interests and to ensure that various regulations are followed. EG Consults has worked both directly for risk takers, for trading management and also for corporate functions.
Trading risk management at every level must look forward and also backwards. We look forward to outline possible future portfolio performance under a variety of scenarios and thereby establish our risk tolerance. How much risk are we willing to take for potential returns and how much is too much? We also look backwards to explain realized profits and losses. Was actual portfolio performance anticipated in our analysis and our metrics? What are the lessons learned and how do we do better going forward?
Pulling It All Together
Pulling it all together for a trading risk management story means:
working in a complex and dynamic data environment, understanding the intricacies of the supporting systems and operational processes
partnering with subject matter experts including risk takers and the quantitative analysts that develop models
outlining a narrative that clearly evidences the current state of a portfolio and explains projected risk and past performance
developing findings through a highly rigorous but flexible approach
On Managing Large Data Sets
Trading risk management depends on the ability to navigate large data sets. At a major bank, there are thousands of positions owned by each trading desk and hundreds of thousands, or even millions, of positions firm-wide. An established glossary of terminology and precise definition of all data elements including metrics, units and timestamps is critical. There are always trade-offs between processing short-cuts such as a sensitivity based approach versus a full valuation approach, which may take hours to generate results. There are always assumptions with implications that must be clearly defined and well understood, particularly under distressed market conditions, when trading risk management matters most. EG Consults exhaustively defined the operating environment and the data flows that are the foundation for analysis.
Drivers of Portfolio Value
Portfolio value at any instant in time comes from the interaction between trading positions and the multitude of rates and prices that define the market environment. Trading positions change moment to moment and the market environment is highly dynamic.
A fundamental challenge of trading risk management is understanding correlation, or the relationships between rates and prices. For example, what is the relationship between the benchmark rate set by the Fed and the market price of U.S. 10-year treasury bond ? When the Fed announces that it will keep its benchmark rate at zero, can the 10Y bond rate rise? If it does rise, how dramatic an impact will there be on U.S. equity prices? Will the impact on technology stocks be greater than the impact on cyclical stocks? These relationships are dynamic and context specific. There is no guarantee that relationships that existed in the past will continue in the future.
In many cases, complex models, governed by context specific model parameters, are used. As the environment changes, these models need to be recalibrated and the model parameters reset.
The scope of EG Consults’ work included thinking through risk scenarios that might arise and assess the ability to hedge positions under those scenarios. The goal was to learn from the past but also think creatively to predict the future, under normal and stressed market conditions. Analysis of results framed a number of questions. What can we discern from trends or breaks in trends? Can we clarify what seem to be anomalies or concentrations? Can we isolate the impact of various value drivers, singly and in various combinations? Can we study the past but also think creatively to predict the future? EG Consults has deep experience understanding the drivers of portfolio value, working with large data sets and generating high quality analysis.
EG Consults Work Product
Narratives were presented in an accessible manner, often with graphics and data visualizations to illustrate key concepts. Presentation standards included consistent look and feel of work product, clearly defined and consistently used terminology, specification of all calculations and units and use of illustrative graphics and data visualizations. EG Consults presentation standards promoted management confidence in the quality of the analysis and the underlying data, laying the groundwork for focussed and constructive discussion of risk.
A Sample Data Illustration
To illustrate how successfully a portfolio of loans has been hedged, we compare the overall returns of our portfolio to the returns of an index of loans (LCDX) traded in the market. From the charts at right, we see that week to week there is poor correlation between index versus portfolio returns, but, over a longer horizon (bi-monthly) the correlation is significant. This implies that hedge ratios can be set to minimize long term or daily P&L volatility, but not both.
A Sample Scenario Report
A scenario report, outlined at right, that:
describes the scenario and the portfolio sensitivity being flexed
identifies units and terminology
has a layout that naturally guides the eye and highlights trends and breaks in trends
Commentary discusses current versus prior scenario results, clarifying changes due to market data, positioning and model calibration.
Supporting material includes detailed scenario descriptions and explicit identification of the parameters being flexed, the relationship between parameters, if any, and the precise amounts these parameters are being flexed.