Search

Neural networks have shown considerable success when used to model financial data series. However, a major weakness of this class of models is the lack of established procedures for misspecification testing and tests of statistical significance for the various estimated parameters. These issues are particularly important in the case of financial engineering where data generating processes are very complex and dominantly stochastic. After a brief review of neural network models, an input selection algorithm is proposed and discussed. It is based on a multistep multiple testing procedure calibrated by using subsampling. The simulation results show that the proposed testing procedure is an effective criterion for selecting a proper set of relevant inputs for the network. When applied to Euro exchange rates, the selected network models show that information contained in past percentage changes can be relevant to the prediction of future percentage changes of certain time series. The apparent predictability for some countries which we analysed does not seem to be an artifact of data snooping. Rather, it is the result of a testing procedure constructed to keep the family wise error rate under control. The results also remain stable while changing the subseries length. [ABSTRACT FROM AUTHOR]

One of the main problems in managing multidimensional data for decision making is that it is impossible to define a complete ordering on multidimensional Euclidean spaces. In order to solve this problem, the scientific community has devolped more and more sofisticated tecniques belonging to the wide framework of Multivariate Statistics. Recently some authors [DR04] have proposed an ordering procedure in which the "meaningful direction" is the "worst-best". The aim of this paper is to extend this approach considering that, especially in financial applications, variables are quantified using different scales and, as we will show, this can lead to undesired results. As a matter of fact, we show that, without an appropriate rescaling, variables with a large range of variation (rv) are "overweighted" with respect to variables with a small one. [ABSTRACT FROM AUTHOR]

Policy makers often have to deal with extreme events. We consider the two main approaches for the analysis of these events, statistical and physical statistics, which are usually considered dichotomically. We quickly delineate both limits and advantages of their tools and some new tentatives for overcoming the historical limits of the latter kind of approach. We highlight that for dealing adequately with extreme events it seems necessary to make use of a new kind of Decision Theory and new tools. [ABSTRACT FROM AUTHOR]

The adoption of pension funds in the Italian social security policy has increased the offer of several investment funds. Workers have to decide what kind of investment to perform, the funds having a different composition and a subsequently different degree of risk. In this paper we propose the use of a distance between GARCH models as a measure of different structure of volatility of some funds, with the purpose of classifying a set of funds. Furthermore we extend the idea of equivalence between ARMA models to the GARCH case to verify the equality of the risk of each couple of funds. An application on thirteen Italian funds and fund indices is performed. [ABSTRACT FROM AUTHOR]

This paper deals with the calculation of suitable risk indicators for Life Insurance policies in a Fair Value context. In particular, after determining the quantile reserve for Life Insurance Participating Policies, the role of the term structure of mortality rates is analyzed in risk determinations. Numerical results are investigated in order to determine not only suitable risk indicators, but also the mortality risk impact in this context. [ABSTRACT FROM AUTHOR]

This contribution deals with Monte Carlo simulation of generalized Gaussian random variables. Such a parametric family of distributions has been proposed for many applications in science and engineering to describe physical phenomena. Its use also seems interesting in modeling economic and financial data. For low values of the shape parameter α, the distribution presents heavy tails. In particular, α = 1/2 is considered and for such a value of the shape parameter, different simulation methods are assessed. [ABSTRACT FROM AUTHOR]

In recent years, marked point processes have found a natural application in the modeling of ultra-high-frequency financial data. The use of these processes does not require the integration of the data which is usually needed by other modeling approaches. Two main classes of marked point processes have so far been proposed to model financial data: the class of the autoregressive conditional duration models of Engle and Rüssel and the broad class of doubly stochastic Poisson processes with marks. In this paper, we show how to model an ultra-high-frequency data set relative to the prices of the future on the S&P 500 index using a particular class of doubly stochastic Poisson process. Our models allow a natural interpretation of the underlying intensity in terms of news reaching the market and does not require the use of ad hoc methods to treat the seasonalities. Filtering and estimation are carried out by Monte Carlo expectation maximization and reversible jump Markov chain Monte Carlo algorithms. [ABSTRACT FROM AUTHOR]

In this paper, an analysis of the performances of an active and quantitative fund management strategy is presented. The strategy consists of working with a portfolio constituted by 30 equally-weighted stock assets selected from a basket of 397 stock assets belonging to the Euro area. The asset allocation is performed in two phases: in the first phase, the 397 stock assets are split into 5 groups; in the second, 6 stock assets are selected from each of the group. The analysis focuses: i) on the specification of quantitative approaches able to effect the group formation; ii) on the definition of a profitable active and quantitative fund management strategy; iii) on the quantitative investigation of the contribution individually provided by each of the two phases to the total profitability of the fund management strategy. [ABSTRACT FROM AUTHOR]

This paper deals with problem of analyzing uncertainty arising from both mortality and disability risks in Long Term Care (LTC) covers. To this purpose some projected demographical scenarios are developed. A biometric risks analysis is performed at portfolio level according to both loss function and risk reserve. The probabilistic structure adopted is consistent with multiple state models, based on a time-continuous Markov chain. [ABSTRACT FROM AUTHOR]

In the framework of demographic processes, different approaches can be identified. According to their aims, these approaches can be differentiated in extrapolative and structural methods. The first focus on the homogeneity of trends in order to obtain projection. The second are based on structural models relating demographic variables to other kinds of variables (geographical, social, economical, etc.). Nowadays, this distinction is not so clear and the joint use of explorative and explanatory approaches is increasing. In this paper, we focus on the extrapolative features of the Lee-Carter model (1992) and propose a reading of such method in the framework of the Multidimensional Data Analysis. Our aim is to propose a data analysis strategy exploiting the analytical and geometrical properties of the Lee-Carter method. [ABSTRACT FROM AUTHOR]

In this paper we consider the problem of bandwidth selection in local polynomial estimation of derivative functions. We use a dependent data context, and analyze time series which are realizations of strictly stationary processes. We consider the estimation of the first derivative of the conditional mean function for a non-linear autoregressive model. First of all, we emphasize the role assumed by the smoothing parameter, by showing how the choice of the bandwidth is crucial for the consistency of the non-parametric estimation procedure, through an example on simulated data. We then use a new approach for the selection of such a parameter, based on the neural network technique. Such alternative method presents several advantages with respect to the traditional approach used so far. [ABSTRACT FROM AUTHOR]

The notion of influence function was introduced by Hampel and it plays a crucial role for important applications in robustness analysis. It is defined by the derivative of a statistic at an underlying distribution and it describes the effect of an infinitesimal contamination at point x on the estimate we are considering. We propose a new approach which can be used whenever the derivative doesn't exist. We extend the definition of influence function to nonsmooth functionals using a notion of generalized derivative. We also prove a generalized von Mises expansion. [ABSTRACT FROM AUTHOR]

The nature of business cycles is a central and conflicting question in macroeconomics. We would like to stress, however, the importance of chaos, in the context of business cycle theories. In fact, those who believe in i.i.d. disturbances simply state that fluctuations are determined by exogenous factors. Chaos supporters, on the other hand, disagree with a linear world and believe that the source of fluctuations is endogenous to the economic system. The aim of paper is to highlight the power of chaos theory to analyze business cycles. [ABSTRACT FROM AUTHOR]

The object of the paper is to compare of two approaches for the analysis of financial durations. The first is the parametric approach (Autoregressive Conditional Duration model) implemented using the exponential, the Weibull, the Burr and the Pareto density functions. The second makes use of bivariate and trivariate copula functions. [ABSTRACT FROM AUTHOR]

The paper concerns the case of the insured loan based on an amortization schedule at variable interest rates, hooked at opportune rate indexes. Using the cash flow structure as a basis, the aim is the evaluation of the mathematical provision of a portfolio in a fair value approach. In this environment, the complexity of the life insurance contract market leads to practical valuation management focused on the choice of the most suitable mortality table and discounting process. A numerical application is proposed, for comparing the reserve fair values referred to two insured loans based on an amortization schedule, the first calculated at a fixed rate and the second, the alternative offered in the market, at a variable rate. [ABSTRACT FROM AUTHOR]

In the first part of the paper we recall the theory of iterated function systems and iterated multifunction systems. In the second part we show some applications in economics, statistics and finance. [ABSTRACT FROM AUTHOR]

This paper deals with the application of the Value at Risk of the mathematical provision within a fair valuation context. Through the VaR calculation, the estimate of an appropriate contingency reserve is connected to the predicted worst case additional cost, at a specific confidence level, projected over a fixed accounting period. The numerical complexity is approached by means of a simulation methodology, particularly suitable also in the case of a large number of risk factors. [ABSTRACT FROM AUTHOR]

In this paper we provide a microeconomic model to investigate the long term memory of financial time series of one share. In the framework we propose, each trader selects a volume of shares to trade and a strategy. Strategies differ for the proportion of fundamentalist/chartist evaluation of price. The share price is determined by the aggregate price. The analyses of volume distribution give an insight of imitative structure among traders. The main property of this model is the functional relation between its parameters at the micro and macro level. This allows an immediate calibration of the model to the long memory degree of the time series under examination, therefore opening the way to understanding the emergence of stylized facts of the market through opinion aggregation. [ABSTRACT FROM AUTHOR]

This paper shows an example of integrated use of three different approaches (Extreme Value Theory (EVT), two-dimensional Fast Fourier Transform (FFT) and Monte Carlo simulation) to model non-life insurance company aggregate losses, taking into account the need for Internal Risk Model development in the light of Solvency II European project. In particular EVT permits the definition of the truncation point between small and large claims. Two-dimensional FFT is used to model not only aggregate losses, but dependence between its basic components too. Finally, Monte Carlo simulation describes large claims behaviour. Collective Risk Model has been developed using Matlab software. [ABSTRACT FROM AUTHOR]

In actuarial literature the properties of risk measures or insurance premium principles have been extensively studied. We propose a characterization of a particular class of coherent risk measures defined in [Art99]. The considered premium principles are obtained by expansion of TVaR measures, consequently they look very interesting in insurance pricing where TVaR measures are frequently used to value tail risks. [ABSTRACT FROM AUTHOR]

In this paper we consider the problem of studying the gap between bounds of risk measures of sums of non-independentrandom variables. Owing to the choice of the context of where to set the problem, namely that of distortion risk measures, we first deduce an explicit formula for the risk measure of a discrete risk by referring to its writing as sum of layers. Then, we examine the case of sums of discrete risks with identical distribution. Upper and lower bounds for risk measures of sums of risks are presented in the case of concave distortion functions. Finally, the attention is devoted to the analysis of the gap between risk measures of upper and lower bounds, with the aim of optimizing it. [ABSTRACT FROM AUTHOR]

The scaling behaviour of both log-price and volume is analyzed for three stock indexes. The traditional approach, mainly consisting of the evaluation of particular moments such as variance or higher absolute moments, is replaced by a new technique which allows the estimation of the self-similarity parameter on the whole empirical distribution designed by any time horizon. In this way, the method we propose attaches its own scaling parameter to any two given time lags, so defining a scaling surface whose properties give information about the nature of the analyzed process. We conclude that, for the log-price process, self-similarity is rejected with a frequency much larger than that assumed by the confidence interval and, when not rejected, the scaling parameter heavily changes with the considered pair of time horizons. Opposite evidence is provided for the volumes, characterized by (generally low) self-similarity parameters which are somewhat uniform with respect to the pairs of time horizons. [ABSTRACT FROM AUTHOR]

The solution of a multistage stochastic programming problem needs a suitable representation of uncertainty which may be obtained through a satisfactory scenario tree construction. There is a trade-off between the level of accuracy in the description of the stochastic component and the computational tractability of the resulting scenario-based problem. In order to face such a trade-off which plays a crucial role in the determination of the optimal solution, we discuss methods that allow progressive reductions of a given scenario tree by means of spatial aggregation. In this process it is important to take into account the choice of proper aggregation criteria in order to try to preserve all the relevant information. [ABSTRACT FROM AUTHOR]

In this paper we extend the Least Squares Monte Carlo approach proposed by Longstaff and Schwartz for the valuation of American-style contingent-claims to the case of life insurance contracts. These contracts, in fact, often embed an American-style option, called surrender option, that entitles its owner to terminate early the contract and receive a cash amount, called surrender value. The additional complication arising in life insurance policies with respect to purely financial American contracts is that there is not a fixed date within which the option can be exercised, since its "maturity" is driven by mortality factors. This complication has been handled by very few papers, often at the cost of excessively simplified valuation frameworks. The aim of this contribution, that is not a specific valuation model but a methodological approach, is to allow a full exploitation of the flexibility inborn in Monte Carlo and quasi-Monte Carlo methods in order to deal with more realistic valuation frameworks. [ABSTRACT FROM AUTHOR]

Style analysis models aim to decompose the performance of a financial portfolio with respect to a set known indexes. Quantile regression offers a different point of view on the style analysis problem as it allows the extraction of information at different parts of the portfolio returns distribution. Moreover, the quantile regression results are useful in order to estimate the portfolio conditional returns distribution. [ABSTRACT FROM AUTHOR]

The forecasts generation from models that belong to the threshold class is discussed. The main problems that arise when forecasts have to be computed from these models are presented and, in particular, least squares, plug-in and combined predictors are pointed out. The performance of the proposed predictors are investigated using simulated and empirical examples that give evidence in favor of the forecasts combination. [ABSTRACT FROM AUTHOR]

During the course of evolution, genomes can undergo large-scale mutation events such as rearrangement and lateral transfer. Such mutations can result in significant variations in gene order and gene content among otherwise closely related organisms. The Mauve genome alignment system can successfully identify such rearrangement and lateral transfer events in comparisons of multiple microbial genomes even under high levels of recombination. This chapter outlines the main features of Mauve and provides examples that describe how to use Mauve to conduct a rigorous multiple genome comparison and study evolutionary patterns. [ABSTRACT FROM AUTHOR]

ERIC (Enteropathogen Resource Information Center) is one of the National Institute of Allergy and Infectious Diseases (NIAID) Bioinformatics Resource Centers for Biodefense and Emerging/Re-emerging Infectious Disease. ERIC serves as a comprehensive information resource for five related pathogens: Yersinia enterocolitica, Yersinia pestis, diarrheagenic E. coli, Shigella spp., and Salmonella spp. ERIC integrates genomics, proteomics, biochemical and microbiological information to facilitate the interpretation and understanding of ERIC pathogens and select related non-pathogens for the advancement of diagnostics, therapeutics, and vaccines. [ABSTRACT FROM AUTHOR]

In the last years, optical studies of Isolated Neutron Stars (INSs) have expanded from the more classical rotation-powered ones to other categories, like the Anomalous X-ray Pulsars (AXPs) and the Soft Gamma-ray Repeaters (SGRs), which make up the class of the magnetars, the radio-quiet INSs with X-ray thermal emission and, more recently, the enigmatic Compact Central Objects (CCOs) in supernova remnants. Apart from 10 rotation-powered pulsars, so far optical/IR counterparts have been found for 5 magnetars and for 4 INSs. In this work we present some of the latest observational results obtained from optical/IR observations of different types of INSs. [ABSTRACT FROM AUTHOR]

We describe a new method for reliably identifying conserved segments among genome sequences that have undergone rearrangement, horizontal transfer, and substantial nucleotide-level divergence. A Gibbs-like sampler explores different combinations of sequence-based markers shared by the genomes under study. The sampler assigns each marker a posterior probability based on how frequently it participates in some collinear group of markers. Markers with high p.p. values are likely members of conserved segments. The method identifies both large-scale and local trends in segmental collinearity, providing suitable input for genome alignment and rearrangement history inference tools. Applying our method to genomes of four Streptococci reveals that rearranged segments in these organisms belong in two size categories: large conserved segments that are interrupted by a staccato of single gene or operon-size small segments. The rearrangement pattern of large segments is best explained by symmetric inversions about the origin of replication while the pattern of small segments is not. [ABSTRACT FROM AUTHOR]

By now there is substantial evidence that Gamma-Ray Bursts (GRBs) originate at cosmological distances from unusually powerful explosions. The interaction between a GRB and its surrounding environment has dramatic consequences on the environment itself. At early times, the strong X-ray UV flux photoionizes the medium on distance scales on the order of 100 pc [7], destroys dust [10], creates photoionization edges [2]. These are short term effects, that occur while the afterglow is propagating in the medium. [ABSTRACT FROM AUTHOR]