In multivariate time series systems, key insights can be obtained by discovering lead-lagrelationships inherent in the data, which refer to the dependence between two timeseries shifted in time relative to one another, and which can be leveraged for thepurposes of control, forecasting or...

This paper proposes two new approaches to improve estimation of the coefficients of the multivariate HAR (MHAR) model, and in turn improve forecast performance. A robust estimator of the covariance matrix is adopted to replace the realized covariance (RCov) matrix while estimating the MHAR...

We study a general class of utility processes V(c)=(V_{t}(c)), where V_{t}(c), a dynamic utility operator, is a decision criterion that quantifies a decision maker's evaluation of uncertain consumption streams c. We call this dynamic utility operator robust and its distinctiveness is that it...

Portfolio selection is often faced with large noisy data sets of strongly correlated asset returns, and so is prone to unstable portfolio weights and serious estimation error. To attenuate these problems, this paper proposes a new latent factor model equipped with both a suitable robust...

The objectives of option hedging/trading extend beyond mere protection against downside risks, with a desire to seek gains also driving agent's strategies. In this study, we showcase the potential of robust risk-aware reinforcement learning (RL) in mitigating the risks associated with...

Principal Component Analysis (PCA) is an important methodology to reduce and extract meaningful signals from large data-sets. Financial markets introduce time and non stationarity aspects, where applying standard PCA methods may not give stable results. We propose robust rolling PCA (R2-PCA)...

We revisit the relationship between pre-primary education and achievement test scores in the primary school, by employing a Bayesian Model Averaging (BMA) approach combined with Gini regression estimates. The data are taken from the OECD Programme for International Student Assessment (PISA) for...

We show that in all (whether generic or nongeneric) binary-action supermodular games, an extreme action profile is robust to incomplete information if and only if it is a monotone potential maximizer. The equivalence does not hold for nonextreme action profiles