Accounting for Ambiguous Modes in Historical Data: A Bayesian Approach

2016 ◽  
pp. 247-271
Keyword(s):  
Bayesian Approach ◽  
Historical Data

scholarly journals A Bayesian Approach for Asset Allocation

2020 ◽  
Vol 9 (4) ◽  
pp. 1 ◽  
Author(s):  
Mihnea S. Andrei ◽  
John S. J. Hsu
Keyword(s):  
Sensitivity Analysis ◽  
Covariance Matrix ◽  
Test Data ◽  
Bayesian Approach ◽  
Asset Allocation ◽  
Historical Data ◽  
Bayesian Framework ◽  
Data Set ◽  
Model Section ◽  
And Performance

The Black-Litterman model combines investors’ personal views with historical data and gives optimal portfolio weights. In this paper we will introduce the original Black-Litterman model (Section 1), we will modify the model such that it fits in a Bayesian framework by considering the investors’ personal views to be a direct prior on the means of the returns and by including a typical Inverse Wishart prior on the covariance matrix of the returns (Section 2). We will also consider an idea of Leonard & Hsu [1992] for a prior on the logarithm of the covariance matrix (Section 3). Sensitivity analysis for the level of confidence that investors have in their own personal views was performed and performance of the models was assessed on a test data set consisting of returns over the month of January 2018.

scholarly journals A Bayesian Approach to Measurement of Backtest Overfitting

Risks ◽  
2021 ◽  
Vol 9 (1) ◽  
pp. 18
Author(s):  
Jiří Witzany
Keyword(s):  
Bayesian Approach ◽  
Performance Indicators ◽  
Multiple Testing ◽  
Historical Data ◽  
Broad Class ◽  
Investment Strategies ◽  
Robust Estimates ◽  
Expected Performance ◽  
Mcmc Simulation ◽  
Naive Approach

Quantitative investment strategies are often selected from a broad class of candidate models estimated and tested on historical data. Standard statistical techniques to prevent model overfitting such as out-sample backtesting turn out to be unreliable in situations when the selection is based on results of too many models tested on the holdout sample. There is an ongoing discussion of how to estimate the probability of backtest overfitting and adjust the expected performance indicators such as the Sharpe ratio in order to reflect properly the effect of multiple testing. We propose a consistent Bayesian approach that yields the desired robust estimates on the basis of a Markov chain Monte Carlo (MCMC) simulation. The approach is tested on a class of technical trading strategies where a seemingly profitable strategy can be selected in the naïve approach.

scholarly journals Bayesian Approach to the Assessment of Geological Risk in Oil and Gas Exploration

2020 ◽  
Vol 8 (7) ◽  
pp. 203-210
Author(s):  
Bruno Rafael Dias de Lucena ◽  
Leonardo Junqueira Lustosa
Keyword(s):  
Bayesian Approach ◽  
Oil And Gas ◽  
Historical Data ◽  
Oil And Gas Exploration ◽  
Oil Resources ◽  
The Bayesian Approach ◽  
Illustrative Application

When assessing undiscovered oil resources, an important step is the assessment of geological risk, which is usually defined as the probability that there will be no accumulation of hydrocarbons. Some important authors have traditional ways of obtaining this probability, but these classic models are not developed on a rigorous basis. Therefore, they may present conflicting results, which are not always compatible with reality and are not able to take into account historical data from similar situations already studied. This article aims to propose a Bayesian approach to the determination of geological risk with advantages over classical approaches. The positive aspects and limitations of the Bayesian approach are discussed and an illustrative application using fictitious data is presented.

scholarly journals An Application of Bayesian Method in Packaged Food Quality Control

1970 ◽  
Vol 9 ◽  
pp. 41-48 ◽  
Author(s):  
R. P. Khatiwada ◽  
A. B. Sthapit
Keyword(s):  
Experimental Data ◽  
Quality Control ◽  
Prior Knowledge ◽  
Food Quality ◽  
Bayesian Approach ◽  
Prior Information ◽  
Historical Data ◽  
Kernel Density ◽  
Predictive Distribution ◽  
Food Quality Control

Conventional method of making statistical inference regarding food quality measure is absolutely based upon experimental data. It refuses to incorporate prior knowledge and historical data on parameter of interest. It is not well suited in the food quality control problems. We propose to use a Bayesian approach inferring the conformance of the data concerning quality run. This approach integrates the facts about the parameter of interest from the historical data or from the expert knowledge. The prior information are used along with the experimental data for the meaningful deduction. In this study, we used Bayesian approach to infer the weight of pouched ghee. Data are taken selecting random samples from a dairy industry. The prior information about average weight and the process standard deviation are taken from the prior knowledge of process specification and standards. Normal-Normal model is used to combine the prior and experimental data in Bayesian framework. We used user-friendly computer programmes, ‘First Bayes' and ‘WinBUGS' to obtain posterior distribution, estimating the process precision, credible intervals, and predictive distribution. Results are presented comparing with conventional methods. Fitting of the model is shown using kernel density and triplot of the distributions. Key words: credible interval; kernel density; posterior distribution; predictive distribution; triplot DOI: 10.3126/njst.v9i0.3163 Nepal Journal of Science and Technology 9 (2008) 41-48

scholarly journals A Bayesian Approach for Large Asset Allocation

2020 ◽  
Vol 10 (1) ◽  
pp. 58
Author(s):  
Mihnea S. Andrei ◽  
John S. J. Hsu
Keyword(s):  
Sensitivity Analysis ◽  
Bayesian Approach ◽  
Asset Allocation ◽  
Historical Data ◽  
Factor Models ◽  
Memory Allocation ◽  
Multivariate Normal ◽  
Confidence Levels ◽  
Conduct Sensitivity Analysis ◽  
Bayesian Factor

The Black-Litterman model combines investor’s personal views with historical data and gives optimal portfolio weights. In (Andrei & Hsu, 2020), they reviewed the original Black-Litterman model and modified it in order to fit it into a Bayesian framework, when a certain number of assets is considered. They used the idea by (Leonard & Hsu, 1992) for a multivariate normal prior on the logarithm of the covariance matrix. When implemented and applied to a large number of assets such as all the S&P500 companies, they ran into memory allocation and running time issues. In this paper, we reduce the dimensions by considering Bayesian factor models, which solve the asset allocation problems for a large number of assets. In addition, we will conduct sensitivity analysis for the confidence levels that the investors have to input.

Information technology and the newsprint demand in Western Europe: a Bayesian approach

2003 ◽  
Vol 33 (9) ◽  
pp. 1644-1652 ◽  
Author(s):  
Torjus Folsland Bolkesjø ◽  
Michael Obersteiner ◽  
Birger Solberg
Keyword(s):  
Information Technology ◽  
Bayesian Approach ◽  
Western Europe ◽  
Historical Data ◽  
Bayesian Models ◽  
The United States ◽  
Consumption Pattern ◽  
Sufficient Information ◽  
Demand Model ◽  
Demand Models

This paper focuses on the impacts of new information technology on newsprint demand in a sample of West European countries (Germany, Italy, Spain, and the United Kingdom). It is hypothesized that information technology, through the ready and free availability of news content on the Internet, could induce a structural shift in the newsprint consumption pattern in these markets. Econometric analyses based on historical data for the four countries mentioned above do not yet support this hypothesis. Based on evidence from the United States, where Internet penetration is higher, and several recently published market studies, there is, however, reason to expect stagnating newsprint consumption in Western Europe. By using Bayesian demand models, we try to incorporate prior information from these market studies in the econometric analysis. A classical demand model, based solely on historical data from 1971 to 1999, is estimated for comparison with the Bayesian models. Predictions for newsprint consumption based on the Bayesian approach show lower future consumption levels than those predicted by the classical models, which are commonly used in forest product demand studies. We conclude that Bayesian models carry the potential to improve the quality of forest products demand analyses when a structural break can be expected and sufficient information on its dynamics is available.

Sign in / Sign up

Export Citation Format

Share Document