fBasics-package: Portfolio Modelling, Optimization and Backtesting

Description

The Rmetrics "fbasics" package is a collection of functions to explore and to investigate basic properties of financial returns and related quantities. The covered fields include techniques of explorative data analysis and the investigation of distributional properties, including parameter estimation and hypothesis testing. For Explorative Data Anlysis several plot functions are available to explore the time series themselves, to show their distributional properties, and to encover correlations and dependencies. Functions to compute Distributional Properties of financial returns and derivated and related series are enclosed. The in detail considered distributions include the normal, the student-t, the stable, the family of generalised hyperbolic, and max drawdown distributions. Moment and log-likelihood estimators allow to estimate the distributional parameters. Functions to compute moments and modes ar also avialable. The functions from Hypothesis Testing one sample and two sample tests. Most of the one sample tests deal with testing normality. The two sample tests allow to compare two series to find out if the series are correlated or their distributions have the same distributional parameters.

Arguments

Overview:

The following chapters give a brief introduction how to optimize and analyze portfolios. 1. Explorative Data Analysis 2. Distributional Properties 3. Hypthesis Testing

1. Explorative Data Analysis

Explorative data analysis of financial return series and related series Exploratory Data Analysis is an approach for data analysis that employs a variety of techniques most of graphical nature to maximize insight into a data set, to uncover underlying structures, and to detect outliers and anomalies. For this several functions are implemented to plot the series, to plot the distribution from different views, and to visualize correlations and dependencies.

ll{ Time Series Plots: seriesPlot Returns a tailored return series plot, cumulatedPlot returns a cumulated series given the returns, returnPlot returns returns given the cumulated series, drawdownPlot returns drawdowns given the return series, Density Plots: histPlot Returns a tailored histogram plot, densityPlot returns a tailored kernel density estimate plot, logDensityPlot returns a tailored log kernel density estimate plot, qqPlot returns a quantile-quantile plot, scalinglawPlot returns a scaling law plot, Box Plots: boxPlot Returns a side-by-side standard box plot, boxPercentilePlot Returns a side-by-side box-percentile plot, CorrelationPlots: acfPlot autocorrelation function plot, pacfPlot partial autocorrelation function plot, lacfPlot lagged autocorrelation function plot, teffectPlot Taylor effect plot. }

2. Distributional Properties

The distributional properties can be investigate for several types of distribution functions which are important in the investigation of financial returns and related time series. Stable and Skew-Stable Distribution ll{ [dpqr]stable the stable and skew-stable distribution, stableMode the stable and skew-stable mode, stableSlider interactive stable distribution display. }

The family of Generalized Hyperbolic Distributions ll{ GH: [dpqr]gh GH, Generlized hyperbolic distribution, ghFit GH parameter estimation, ghMode mode of the GH distribution, HYP: [dpqr]hyp HYP, hyperbolic distribution, hypFit HYP parameter estimation, hypMode mode of the hyperbolic distribution, hypSlider HYP distribution Slider, NIG: [dpqr]nig NIG, hyperbolic distribution, nigFit NIG parameter estimation, nigMode mode of the NIG distribution, nigSlider NIG distribution Slider, GHT: [dpqr]ght GHT, generalized hyperbolic Student-t, ghtFit GHT parameter estimation, ghtMode mode of the GHT distribution, ghtSlider GHT distribution Slider.}

The max-Drawdown Distribution

The functions compute compute drawdown statistics. Included are density, distribution function, and random generation for the maximum drawdown distributions. In addition the expectation of drawdowns for Brownian motion can be computed. ll{ [dpr]maxdd the max drawdown distribution, maxddStats the expectation of drawdowns. }

3. Hypthesis Testing