Fractional differencing: (in)stability of spectral structure and risk measures of financial networks

Starlings (Sturnus vulgaris) learned 2 series of discrimination problems with complex tones. Within each problem the complex tones contained the same fundamental frequency but differed in the relational structure of spectral components. The S+ spectral structure was constant across problems within a series; the S− spectral structure and the fundamental frequency of the 2 complex tones varied across problems. The 2 series used different spectral structures as S+. For both series the starlings were able to discriminate among complex tones according to spectral structure. The starlings transferred to novel complex tone discriminations but did not transfer to complex tone discriminations in which the reinforcement contingencies associated with spectral structure were reversed

Calculating and comparing security returns is harder than you think: A comparison between logarithmic and simple returns

Infants' perception of timbre: Classification of complex tones by spectral structure

A novel method of model order selection and separation of spectral structures is proposed for practical signals which are composed of multiple structured spectra. The input time sequence signal is considered to have multiple spectral structure, i.e. a main spectral structure and a residual spectral structure. The main structure is determined by the signal dominant power spectral component of the input signal and the residual structure is defined by the residual power spectral component after the dominant power spectral component is removed from the whole spectral structure. The method first estimates the dominant spectrum of the main structure using the AR (autoregressive) model with order p(AR (p)) and then estimates the spectrum of the residual structure using the AR model with order q(AR (q)). By computer simulation, the method is proved to give a good solution to the problem of reliable spectral estimation. >

We give an intrinsic (set theoretical) method to obtain all realcompletions of a Tychonoff space X. It is based on the concept of a spectral structure. Every realcompletion of the space can be obtained as a space of terminal clusters relative to an appropriate spectral structure on X. Various applications of this concept are then given. For example we may characterize those spectral structures which yield the realcompletions between X and βX, or which yield spaces that are realcomplete (that is, realcompact), or compact or pseudocompact or Lindelof. We also determine the class K of compactifications K of X for which X will be real closed in K for every K e K.

In this paper we describe and clarify the definitions and the usage of the simple and logarithmic returns for financial assets like stocks or portfolios. It can be proven that the distributions of the simple and logarithmic returns are really close to each other. Because of this fact we investigate the question whether the calculated financial risk depends on the use of simple or log returns. To show the effect of the return-type on the calculations, we consider and compare the riskiness order of stocks and portfolios. For our purposes, in the empirical study we use seven Hungarian daily stock prices and for the risk calculation we focus on the following risk measures: standard deviation, semivariance, Value at Risk and Expected Shortfall. The results clearly show that the riskiness order can depend on the use of the return type (i.e. log or simple return). Generally, often – due to missing data or the nature of the analysis – one has to use approximations. We also examine the effect of these approximations on the riskiness order of stocks and of portfolios. We found differences in the riskiness order using exact or approximated values. Therefore, we believe, if this is possible, exact values instead of approximated ones should be used for calculations. Additionally, it is important that one uses the same type of return within one study and one has to be aware of the possible instabilities when comparing return results.
 JEL Code: C18

Although almost all auditory brainstem nuclei receive serotonergic innervation, little is known about its effects on auditory neurons. We address this question by evaluating the effects of serotonin on sound-evoked activity of neurons in the inferior colliculus (IC) of Mexican free-tailed bats. Two types of auditory stimuli were used: tone bursts at the neuron's best frequency and frequency-modulated (FM) sweeps with a variety of spectral and temporal structures. There were two main findings. First, serotonin changed tone-evoked responses in 66% of the IC neurons sampled. Second, the influence of serotonin often depended on the type of signal presented. Although serotonin depressed tone-evoked responses in most neurons, its effects on responses to FM sweeps were evenly mixed between depression and facilitation. Thus in most cells serotonin had a different effect on tone-evoked responses than it did on FM-evoked responses. In some neurons serotonin depressed responses evoked by tone bursts but left the responses to FM sweeps unchanged, whereas in others serotonin had little or no effect on responses to tone bursts but substantially facilitated responses to FM sweeps. In addition, serotonin could differentially affect responses to various FM sweeps that differed in temporal or spectral structure. Previous studies have revealed that the efficacy of the serotonergic innervation is partially modulated by sensory stimuli and by behavioral states. Thus our results suggest that the population activity evoked by a particular sound is not simply a consequence of the hard wiring that connects the IC to lower and higher regions but rather is highly dynamic because of the functional reconfigurations induced by serotonin and almost certainly other neuromodulators as well.

This paper considers the blind separation of the harmonic and percussive components of multichannel music signals. We model the contribution of each source to all mixture channels in the time-frequency domain via a spatial covariance matrix, which encodes its spatial characteristics, and a scalar spectral variance, which represents its spectral structure. We then exploit the spatial continuity and the different spectral continuity structures of harmonic and percussive components as prior information to derive maximum a posteriori (MAP) estimates of the parameters using the expectation-maximization (EM) algorithm. Experimental results over professional musical mixtures show the effectiveness of the proposed approach.

For the most recent years, risk has become one of the essential parameters in portfolio optimization problems. Today most practitioners and researchers in portfolio optimization have used variance as a standard risk measure. This approach has been found subjective. The Markowitz (1952) mean-variance model considered variance as an adequate portfolio risk measure, and asset returns are multivariate normally distributed and that investors have a quadratic utility function which is subjective too. Other risk measures have been suggested to overcome the limitations of the mean-variance model. This paper analyzes which portfolio optimization models can better explain the optimal portfolio performance (high return, low risk) for the Uganda Security Exchange (USE). We compare Mean-Variance (MV), Mean Absolute Deviation (MAD), Robust Portfolios and Covariance Estimation Models (The Shrinked Mean-Variance (SMV) Models & Alternative Covariance Estimator (ACE) Models) and Mean-Conditional Value-at-Risk (Mean-CVaR) models in terms of the risk and performance. For the computed monthly returns and price data (February 2010 to January 2021) for USE selected stocks, we considered the results to show that Mean-CVaR and ACE portfolios have the highest performance ratio compared to other models. We find that VaR is the best risk measure for portfolio optimization for the USE since it has lower values across all models than other risk measures. It is vital to consider all the available risk measures for a regulator or practitioner to make a good decision since using one can be subjective; as seen in our results, different risk measures yield different results.

The X-ray photoelectron spectral structure of CeO2 valence electrons in the binding energy range of 0 to ∼50 eV was analyzed. The core-electron spectral structure parameters and the results of relativistic discrete-variational calculations of CeO8 and Ce63O216 clusters were taken into account. Comparison of the valence and the core-electron spectral structures showed that the formation of the inner (IVMO) and the outer (OVMO) valence molecular orbitals contributes to the spectral structure more than the many-body processes. The Ce 4f electrons were established to participate directly in chemical bond formation in CeO2 losing partially their f character. They were found to be localized mostly within the outer valence band. The Ce 5p atomic orbitals were shown to participate in the formation of both the inner and the outer valence molecular orbitals (MOs). A large part in the IVMO formation is taken by the filled Ce 5p1/2, 5p3/2 and O 2s atomic shells, while the Ce 5s electrons participate weakly in the chemical bond formation. The composition and the sequent order of the molecular orbitals in the binding energy range of 0 to ∼50 eV were established. A quantitative scheme for the molecular orbitals of CeO2 was built. This scheme is fundamental for understanding the nature of chemical bonding and also for the interpretation of other X-ray spectra of CeO2. Evaluations revealed that the IVMO electrons weaken the chemical bond formed by the OVMO electrons by 37%.

Under consideration of the basic factors such as the level hyperfine structure and isotope split etc., the spectral structures of spontaneous emission of copper atom 578/511nm lines are calculated. A sealed operation CuBr laser with periodical refreshment of the buffer gas are designed and constructed. The 578/511nm spectral structures, produced by CuBr laser at lower Ne buffer gas pressure were measured under different working temperatures and exciting voltages. The spectral structure of spontaneous emission of copper atom 578/511nm lines have multi-peaks similar to their laser spectral structure, but the 578nm laser line isstrongly dependent on the working temperature and the exciting voltage. Some possible reasons are suggested to qualitatively explain the experimental results.

The far-field displacement amplitude spectra of 14 earthquakes from Iran, Fiji Islands, Tonga-Kermadec Islands, and South America were obtained using the long-period P and S-wave records of the WWSSN stations. Despite recent improvements in theories of the seismic-source mechanism there is continuing doubt concerning the interpretation of the observed seismic spectrum. Although the dislocation model has been adopted by many investigators, and in particular Brune's model is widely used, there has often not been satisfactory evidence in the observed spectral data to justify this. Generally speaking, however, the spectral structures obtained in the present work have a form closely similar to those calculated by Brune and Savage, and it is shown that anomalies in the observed amplitude spectral structure are most likely due to noise and propagation effects.

Stochastic excitation and detection of persistent spectral structures

We investigate the emergence of a structure in the correlation matrix of assets' returns as the time horizon over which returns are computed increases from the minutes to the daily scale. We analyze data from different stock markets (New York, Paris, London, Milano) and with different methods. In addition to the usual correlations, we also analyze those obtained by subtracting the dynamics of the "center of mass" (i.e., the market mode). We find that when the center of mass is not removed the structure emerges, as the time horizon increases, from splitting a single large cluster into smaller ones. By contrast, when the market mode is removed the structure of correlations observed at the daily scale is already well defined at very high frequency (5 min in the New York Stock Exchange). Moreover, this structure accounts for 80% of the classification of stocks in economic sectors. Similar results, though less sharp, are found for the other markets. We also find that the structure of correlations in the overnight returns is markedly different from that of intraday activity.

The literature provides strong evidence that stock price values can be predicted from past price data. Principal component analysis (PCA) identifies a small number of principle components that explain most of the variation in a data set. This method is often used for dimensionality reduction and analysis of the data. In this paper, we develop a general method for stock price prediction using time-varying covariance information. To address the time-varying nature of financial time series, we assign exponential weights to the price data so that recent data points are weighted more heavily. Our proposed method involves a dimension-reduction operation constructed based on principle components. Projecting the noisy observation onto a principle subspace results in a well-conditioned problem. We illustrate our results based on historical daily price data for 150 companies from different market-capitalization categories. We compare the performance of our method to two other methods: Gauss-Bayes, which is numerically demanding, and moving average, a simple method often used by technical traders and researchers. We investigate the results based on mean squared error and directional change statistic of prediction, as measures of performance, and volatility of prediction as a measure of risk.