It should be easy to explore and analyze any data you think might result in an opportunity, and to quickly turn your insights into the prototype of a signal, factor, or model. RapidQuant gives you tools that "get out of your way" so you can focus on your ideas instead of their implementation.
The high-productivity Python environment of RapidQuant is uniquely suited to signal generation, with rich interactivity, intelligent data alignment and handling of missing data, flexible reshaping, powerful data structures, and libraries to accomplish sophisticated financial analysis.
RapidQuant fits the iterative workflow of financial research, letting you rapidly backtest your ideas and understand the results. Computations have been comprehensively optimized for performance so you won't have to wait, yet there's no compiling or mode-switching.
The backtesting framework of RapidQuant makes it easy to construct a portfolio, backtest, and understand performance & risk, while the powerful analytics library can be used to gain unique insight into the results.
RapidQuant's powerful and performant in-memory objects extend the open source pandas library to provide data structures for regular and irregular financial time series, data frames, and panel data, plus the tools to easily interface between memory and multiple storage formats, including Microsoft® Excel®, SQL databases, HDF5, web APIs, CSV and text files, even the clipboard.
The pandas library also includes fast and efficient tools to manipulate data, including intelligent data alignment and handling of missing data, flexible reshaping and pivoting, label-based slicing, SQL-like operations, hierarchical indexing, date range generation and frequency conversion, date shifting and lagging, customizable time offsets, and much more.
The open source NumPy and StatsModels libraries provides a huge range of features, including linear algebra, high-performance array processing on multidimensional arrays, fast Fourier transform, random number generation, plus many statistic and econometric models and tests, including regressions, linear models, descriptive statistics, moving window statistics and linear regressions, non-parametric tests, and much more.
The factor transformation library allows you to standardize, winsorize, volatility adjust, volatility shrinkage, find industry/country mean and volatility adjustments, and more.
RapidQuant includes a breadth of tools to understand risk, including ex-ante portfolio volatility and covariance, ex-ante CAPM beta, total and marginal contribution to risk, systematic and idiosyncratic contribution to risk, portfolio risk decomposition by factor category, and more.
The framework for backtesting allows you to run a backtest on your portfolio against historical prices with your transaction cost assumptions.
The performance analytics library gives you the tools to understand summary and moving window performance, including summary statistics, covariance, correlation, skewness, kurtosis, semivariance, drawdowns, realized CAPM beta and alpha,historical VaR, historical expected shortfall, hit rate, average positive & negative return periods, Sharpe ratio, information ratio, Sortino ratio, Treynor ratio, updown ratio, upside potential ratio, Calmar ratio, and more.
The performance attribution library allows you to deeply understand your model and portfolio, with BHB, Brinson-Fachler, ex-ante factor exposures, ex-post multi-variate factor loadings, regression based factor attribution, and return decomposition along size, volume, risk, and turnover quantile ranges.