August 12, 2025
A process for developing and testing quantitative strategies should focus first on knowing when and why something won’t work, as well as providing a foundation for later enhancement.
September 15, 2025
Investors can achieve better returns overall when the portfolio’s risk is diversified. In this paper, we developed a method for optimizing the contribution to the tail risk of the entire portfolio by diversifying that tail risk across the portfolio.
September 15, 2025
Complex portfolios also tend to have complex objectives and constraints. These complexities lead to “non-convex” (non-smooth) optimization surfaces. Differential Evolution is an optimization technique that can find optimal solutions even to high-dimensional complex portfolio problems.
September 15, 2025
Some portfolio selection problems give rise to complex risk objectives. Assets like cryptocurrencies have “non-normal” return distributions. Utilizing “modified” risk metrics that consider the skew and fat-tailed nature of cryptocurrencies may improve portfolio objectives.
August 12, 2025
We believe in releasing our tools for quantitative analysis as open source code for others to examine, evaluate, and apply. quantstrat originated as a framework for developing and testing signal based strategies at time frequencies ranging from nanoseconds to years. We believe there is more value for generating alpha in the open source community than in secrecy for advanced tools. Explore, utilize, and contribute to the quantstrat package with us.
September 15, 2025
Performance and risk analysis is critical for navigating the digital asset space. We’ve developed a toolkit of cutting edge performance and risk tools for these purposes. The open source PerformanceAnalytics package is used globally by quants, fund managers, and banks. In addition to standard risk and performance metrics, this package aims to aid practitioners and researchers in utilizing the latest research in the analysis of non-normal return streams.
September 15, 2025
The blotter package is our open source toolkit for recording and reconciling modern electronic transactions in multiple asset classes, multiple currencies, and is capable of handling millions of transactions in a backtest or a production algorithm.
September 15, 2025
Modern portfolios with diverse assets, including digital assets, alternatives, and global exposures, need better tools than the classic mean-variance optimization framework. Since 2008, we’ve been developing PortfolioAnalytics to include support for the construction, optimization, and analysis of portfolios using some of the most advanced optimization solvers, objectives, constraints, and risk metrics available. The code is widely used in both academia and industry for complex portfolio analysis.
August 12, 2025
GARCH models are useful for modeling volatility. Hidden Markov Chains allow for the probabilities of “hidden” states to be assessed with high accuracy. Combining the two allows for the detection and prediction of volatility “regimes”. First prototyped as part of our work on Differential Evolution in 2008-2009, we expanded this into the MSGARCH package, starting with a Google Summer of Code project in 2017.
September 15, 2025
Analyzing high frequency data comes with a host of complexities. Since 2013 we have mentored multiple GSoC projects and contributed to the highfrequancy package.
September 15, 2025
Using LLM AI models to help search large bodies of research can improve productivity and insight into the work being searched. This 2024 GSoC project uses open LLM’s to create a RAG search model for scientific research.
September 15, 2025
osQF started as R/Finance in the fall of 2008 in Chicago, IL. It was founded by a small group of open source package authors, finance industry practitioners, finance professors, and members of quantstrat. The open source community is at the heart of osQF because it is uniquely unparalleled in both its breadth and depth.