Neural and wavelet network models for financial distress classification

This work analyzes the use of linear discriminant models, multi-layer perceptron neural networks and wavelet networks for corporate financial distress prediction. Although simple and easy to interpret, linear models require statistical assumptions that may be unrealistic. Neural networks are able to discriminate patterns that are not linearly separable, but the large number of parameters involved in a neural model often causes generalization problems. Wavelet networks are classification models that implement nonlinear discriminant surfaces as the superposition of dilated and translated versions of a single "mother wavelet" function. In this paper, an algorithm is proposed to select dilation and translation parameters that yield a wavelet network classifier with good parsimony characteristics. The models are compared in a case study involving failed and continuing British firms in the period 1997-2000. Problems associated with over-parameterized neural networks are illustrated and the Optimal Brain Damage pruning technique is employed to obtain a parsimonious neural model. The results, supported by a re-sampling study, show that both neural and wavelet networks may be a valid alternative to classical linear discriminant models.

The relationship between membrane damage, release of protein and loss of viability in Escherichia coli exposed to high hydrostatic pressure

The aim of this work was to examine a possible association between resistance of two Escherichia coli strains to high hydrostatic pressure and the susceptibility of their cell membranes to pressure-induced damage. Cells were exposed to pressures between 100 and 700 MPa at room temperature (~20C) in phosphate-buffered-saline. In the more pressure-sensitive strain E. coli 8164, loss of viability occurred at pressures between 100 MPa and 300 MPa and coincided with irreversible loss of membrane integrity as indicated by uptake of propidium iodide (PI) and leakage of protein of molecular mass between 9 and 78 kDa from the cells. Protein release increased to a maximum at 400 MPa then decreased, possibly due to intracellular aggregation at the higher pressures. In the pressure-resistant strain E. coli J1, PI was taken up during pressure treatment but not after decompression indicating that cells were able to reseal their membranes. Loss of viability in strain J1 coincided with the transient loss of membrane integrity between approximately 200 MPa and 600 MPa. In E. coli J1 leakage of protein occurred before loss of viability and the released protein was of low molecular mass, between 8 and 11 kDa and may have been of periplasmic origin. In these two strains differences in pressure resistance appeared to be related to differences in the ability of their membranes to withstand disruption by pressure. However it appears that transient loss of membrane integrity during pressure can lead to cell death irrespective of whether cells can reseal their membranes afterwards.

The role of competition and state aid policy in financial and monetary law

During the financial crisis, companies and lenders found themselves in distressed situations. Competition authorities across the globe had to deal with controversial issues such as the application of the failing firm defence in merger transactions as well as assessment of emergency aid granted by states. This article considers competition policy in periods of crisis, in particular the failing firm defence in merger control and its state aid policy.

Modulation of pro-survival Akt/protein kinase B and ERK1/2 signaling cascades by quercetin and its in vivo metabolites underlie their action on neuronal viability

Much recent interest has focused on the potential of flavonoids to interact with intracellular signaling pathways such as with the mitogen-activated protein kinase cascade. We have investigated whether the observed strong neurotoxic potential of quercetin in primary cortical neurons may occur via specific and sensitive interactions within neuronal mitogen-activated protein kinase and Akt/protein kinase B (PKB) signaling cascades, both implicated in neuronal apoptosis. Quercetin induced potent inhibition of both Akt/PKB and ERK phosphorylation, resulting in reduced phosphorylation of BAD and a strong activation of caspase-3. High quercetin concentrations (30 microM) led to sustained loss of Akt phosphorylation and subsequent Akt cleavage by caspase-3, whereas at lower concentrations (<10 microM) the inhibition of Akt phosphorylation was transient and eventually returned to basal levels. Lower levels of quercetin also induced strong activation of the pro-survival transcription factor cAMP-responsive element-binding protein, although this did not prevent neuronal damage. O-Methylated quercetin metabolites inhibited Akt/PKB to lesser extent and did not induce such strong activation of caspase-3, which was reflected in the lower amount of damage they inflicted on neurons. In contrast, neither quercetin nor its O-methylated metabolites had any measurable effect on c-Jun N-terminal kinase phosphorylation. The glucuronide of quercetin was not toxic and did not evoke any alterations in neuronal signaling, probably reflecting its inability to enter neurons. Together these data suggest that quercetin and to a lesser extent its O-methylated metabolites may induce neuronal death via a mechanism involving an inhibition of neuronal survival signaling through the inhibition of both Akt/PKB and ERK rather than by an activation of the c-Jun N-terminal kinase-mediated death pathway.

On the utility of input selection and pruning for financial distress prediction models

Analyzes the use of linear and neural network models for financial distress classification, with emphasis on the issues of input variable selection and model pruning. A data-driven method for selecting input variables (financial ratios, in this case) is proposed. A case study involving 60 British firms in the period 1997-2000 is used for illustration. It is shown that the use of the Optimal Brain Damage pruning technique can considerably improve the generalization ability of a neural model. Moreover, the set of financial ratios obtained with the proposed selection procedure is shown to be an appropriate alternative to the ratios usually employed by practitioners.