A trade blotter is simply a record of trades made over a period of time. The details of those trades are kept at hand so that the trader can review and confirm the trades at the end of the day. I don’t mean to draw a literal comparison here – I’m not going to discuss specific trades. This blog is rather intended to be a blotter in the metaphorical sense of having a scratch pad for ideas related to finance and trading, with a focus on useful tools.
I have more than two decades of experience with investing in and implementing alternative asset management strategies, in both hedge fund and proprietary trading businesses. I currently work with a Chicago-based hedge fund of funds, where I am responsible for identifying investment opportunities among hedge funds across a wide range of markets, strategies, and trading styles. I am also a member of our Investment Committee, responsible for selection and allocation decisions for all of our products. That said, this blog has nothing to do with my current employment and should not be construed to be related. See the standard disclaimer.
Given that I’ve become a dedicated user of R, this blog will include quite a bit of code. R is a free, open source statistical software program that provides a wide variety of statistical and graphical techniques and is highly extensible. R is widely used in finance, particularly in econometrics, quantitative finance and risk analysis. As a result, there’s a huge user community that has contributed hundreds of functions to the R language covering a wide variety of applications.
I’m glad to be a part of that community, and I’m hoping that this blog will help me expand on some the packages I’ve written, particularly about the features and functions that might get overlooked by users just reviewing the documentation. I expect to focus on topics related to finance, including performance and risk measurement, portfolio construction, and strategy development, among others.
Performance and Risk Measurement. I firmly believe that using quantitative tools to uncover information about performance and risk through time can help an investor to become more informed, ask better questions, and make better quality decisions.
To that end, I developed the R package PerformanceAnalytics, which applies current research on return-based analysis of strategy or fund returns for risk, autocorrelation and illiquidity, persistence, style analysis and drift, clustering, quantile regression, and other topics.
Portfolio Construction. Choosing the size of an investment is a complementary process to choosing the instrument to invest in. The key to doing it successfully is to make sure that the method fits the nature and objectives of the portfolio.
I developed the R package PortfolioAnalytics to provide an extensible business-focused framework for portfolio optimization and analysis. The package focuses on numeric approaches for solving non-quadratic problems useful, for example, in risk budgeting. I’ve used multiple optimization approaches to improve single or multi-style portfolios of hedge funds and other asset classes. I’ve found random portfolio sampling particularly helpful for building intuition about optimization.
Strategy Development. Being able to assess the underlying economic rationale of a given investment strategy is critical. I have worked closely with research and investment teams in the past to develop and evaluate investment models, portfolio strategies, and risk parameters for tactical asset allocation, trend following, equity market neutral, statistical arbitrage, and bond basis strategies. Allocating among strategies presents another level of challenge. As a result of that experience, I’ve been actively developing a set of tools around trading simulation and backtesting in R.
One package is blotter, an R package that provides transaction infrastructure for defining transactions, portfolios, and accounts for trading systems and simulation. It provides portfolio support for multi-asset class and multi-currency portfolios.
Another is FinancialInstrument, an R package for defining and storing meta-data for tradable contracts (referred to as instruments, such as for stocks, futures, options, etc.). It can be used to create any asset class and any associated derivatives, across multiple currencies.
A package called quantstrat ties those pieces together, allowing users to specify, build and backtest quantitative financial trading and portfolio strategies. These packages remain in heavy development.
In all of these cases, these packages have benefitted very much from the collaborative nature of the R community in finance. There are some very talented people who are co-authors and collaborators on these and other packages, and I’m pleased that they have taken such a strong interest. In particular, Brian Peterson and Kris Boudt have been key co-authors and collaborators. None of this would be possible without Jeff Ryan‘s excellent xts and quantmod packages. Other very important contributors include Josh Ulrich, Garrett See, Eric Zivot, and Diethelm Wuertz. Many other people contribute both directly and indirectly, such as Guy Yollin, Dirk Eddelbuettel, etc. Those active on the email list at r-sig-finance have been particularly helpful.
Beyond finance, economics, and technology, I have a strong interest in decision-making, social network analysis, and process development, and I hope that some of those topics come up here every once in a while as well. I expect that this blog will be a living document, and I expect that it will change through time as my time and interests migrate. But I hope you find it interesting at least, and even perhaps useful.
I’m one of the original organizers of R/Finance conferences in Chicago, working with a fantastic group of finance professionals and R package authors to run a conference each year. This is a two-day conference that covers portfolio management, time series analysis, advanced risk tools, high-performance computing, econometrics and more. Everything is discussed within the context of using R as a primary tool for the analysis. We started in 2009, and are about to hold our fourth conference in May 2012.
Who I Am
Peter Carl
Through my career, I’ve garnered experience in all aspects of the investment process, including strategy research and development, portfolio and risk management, execution, operations and technology development.
I had the good fortune to begin my career in 1986 at O’Connor and Associates, an options trading boutique that was purchased by Swiss Bank Corporation in 1991. While at O’Connor, I worked closely with the Sante Fe-based Prediction Company, an independent research and development firm, to create a large-scale proprietary statistical arbitrage business. UBS and SBC merged in 1998, and UBS purchased Prediction in 2005.
After I left UBS in 2001, I developed and launched a global macro hedge fund that used a quantitative investment strategy based on a cyclical asset allocation process that gradually moved around various asset classes. I spent enough time discussing risk management issues with funds of hedge funds that when an opportunity came up to move to the other side of the table, I found the prospect very intriguing. It has been a very interesting place to be during the past few years, to say the least.
I think that it is important, particularly in this environment, to keep learning. I started with a BSc in Industrial Engineering from Northwestern University and an MBA from the University of Chicago, but I have probably learned the most from those nearby.
tradeblotter 