Search

Liquidity in the National Balancing Point (NBP) forward market during 2010-14 is examined using liquidity measures adopted from the financial literature. Since the sample period includes the date when the EU Regulation on Market Integrity and Transparency (REMIT) became in force, the question of whether changes in these measures reflect REMIT is also investigated. There is evidence of increased market transparency and competition, but no significant differences in the level of liquidity in the NBP one-month-ahead market appear to have followed the introduction of REMIT.

This paper develops an efficient approach to analytical learning of Asymmetric Stochastic Volatility (ASV) models through nonlinear filtering, and shows that they are useful for practical risk management. This involves the derivation of a Nonlinear Quadrature Filter (NQF) that operates directly on the nonlinear ASV model. The NQF filter makes Gaussian approximations to the prior and posterior density of the latent volatility, but not in the observation space which makes possible easy handling of heavy-tailed data. Experiments in Value-at-Risk (VaR) assessment via an original bootsrtapping methodology are conducted with NQF and several ASV learning algorithms. The results indicate that our approach yields models with better statistical characteristics than the considered competitors, and slightly improved VaR forecasts.

Previous studies that tested for integration of European electricity spot markets employed standard unit root tests. This study extends the existing literature about electricity market integration by adopting a fractional cointegration analysis and providing empirical evidence that the classical unit root test framework, which tests for common stationary trends, may be inadequate because of long range dependencies in electricity spot prices. Hourly data from APX-ENDEX (UK and Netherlands), EPEX (Germany, Switzerland), Nordpool (Finland, Denmark, Norway) and Powernext (France) are employed between January 2009 and April 2011. Results based on parametric fractional ARIMA models suggest that long memory and cointegration in some but not all markets exists. Germany and the Netherlands seem to share the strongest cointegrating relationship. France and Germany as well as France and the Netherlands also share a common stationary trend. Hence, price shocks can be lasting and the EU policy goal of market integration is yet to be achieved.

This paper proposes several techniques for avoiding overfitting in the genetic programming (GP) of polynomials. The model specification flexibility is increased by: (1) a polynomial block reformulation, which reduces the statistical bias, and, (2) complexity tuning using local ridge regression and regularized weight subset selection, which reduce the statistical variance. Another contribution is the designed fitness function for search navigation towards highly predictive models. Experimental results on time-series forecasting show that these techniques help GP to find accurate, less complex and better forecasting polynomials than traditional Koza-style GP (J.R. Koza, 1992) and the previous Stroganoff system (H. Iba et al., 1994, 2001).