Essays on Financial Analytics

This paper examines foreign exchange risk hedging determinants for a sample of 82 French non-financial firms. Starting from the observation that firms, often, use both currency derivatives and foreign debt, we find evidence that foreign debt can be considered as hedging tool in addition to currency derivatives. Our results show that currency derivatives’ hedging depends from firm size, financial distress risk, liquidity level, foreign sales and future growth opportunities. Foreign debt level depends from firm size, debt level, foreign sales and its future growth opportunities.

We demonstrate, further, that foreign debt and currency derivatives are quite different hedging tools. Our results show that the level of operational hedging with foreign debt seems to be loosely correlated with that of currency derivatives.

This paper’s content is devoted to the study of the monetary utility functions and their use in optimal portfolio choice and optimal risk allocation. In most of the relative papers, the domain of a monetary utility function is a dual space. This approach implies that closed and convex sets are weak-star compact. The main contribution of the present paper is the definition of such a function on any Riesz space, which is not necessarily a dual space, but it formulates a symmetric Riesz dual pair together with its topological dual. This way of definition implies the weak compactness of the sets usually needed for the solution of the above optimization problems.

We apply the Koopman operator theory and Extended Dynamic Mode Decomposition (EDMD) in a non-linear dynamical system. This system describes the capital accumulation, and it is similar to the Solow-Swan model and the Ramsey-Cass-Koopmans model. However, the usual derivative is replaced with a fractional derivative. This dynamical system is approximated by a finite-dimensional linear system which is defined in some augmented state space. However, because of the presence of the fractional derivative, one expects that the dimension of the linear system will be quite bigger.

This paper provides new insights into the efficiency of European firms using accounting and financial ratios. In particular, we discuss how the data envelopment analysis (DEA) method can be used with accounting and financial data to highlight the importance of firm profitability as a counterbalance to crisis-induced weakness in demand. We consider several DEA models for studying the technical, financial, and financing efficiency of firms, including a unique set of variables (inputs/outputs) for productivity analysis. Our results provide recommendations for financial managers and analysts dealing with European firms, especially from the southern parts of Europe, i.e., Greece, Italy, and Spain.

Using the Household Finance Survey (EFH), this work shows that the use of financial instruments—whether financial assets or insurance contracts—among Chilean households increased substantially since 2007. Complementing this analysis with the Family Expenditures Survey (EPF) between the years of 1987 and 2017, I show that the share of financial goods in expenditures dropped significantly, while the share of insurance products in consumption roughly doubled in this period. This indicates that financial goods are now much less expensive and the number of its users increased significantly. The use of the different insurance contracts (life and health, vehicles, home, and loan insurance) increased across all income levels. Overall, the widespread use of financial goods, insurance contracts, and purchase of durable goods among the Chilean population across all the income levels shows that the financial access to goods and services increased significantly over the last 35 years.

The rise of cryptocurrencies and social media platforms has given us unique insight on the impact of investor attention on investor trading behavior. In this paper, we focus specifically on the impact of news and social media attention on Bitcoin across five major global exchanges: Bitfinex, Bitstamp, BTC-e, Coinbase, and Kraken. We break attention into three categories: social media attention by existing investors proxied through Reddit posts (seasoned attention), social media attention by new investors proxied through Reddit subscribers (novice attention), and traditional online media attention proxied through the number of Bloomberg news articles. We find that new entrants have a greater impact on Bitcoin than discussions and posts by existing Bitcoin holders. This suggests that rise in Bitcoin prices is driven by new investors entering into the market rather than by existing investors adjusting their valuations and beliefs. In short, the increase in attention by new investors has pushed Bitcoin prices and induced extra noise in the market. We also document some asymmetries in the transmission of investor attention to Bitcoin trades depending on exogenous news shocks.

This study examines the detection of information related to small and large equity styles. Using a novel database of magazines targeting institutional investors, the institutional media, we compare the performance of dictionary-based and supervised machine learning algorithms (Naïve Bayes and support vector machine). Our three main findings are (1) restricted word lists are the most efficient approach, (2) bigram term frequency matrices are the best weighting scheme for algorithms, and (3) Naïve Bayes exhibits overfitting while support vector machine delivers encouraging results. Overall, our results provide material to construct small-cap and large-cap coverage indexes from specialized financial media.

Financial analytics involves the analysis of financial data by using statistical and quantitative methods to make decisions that improve businesses’ results. Specifically, this system includes data mining, predictive analytics, and applied analytics and statistics and is delivered as a custom application to a business user. The integration of financial analytics by companies has already started changing their operational process while giving them the ability to leverage data from different sources, create easy-to-use dashboards, and visualize and predict future performance tools. This means that financial analytics offers the business a competitive edge and facilitates more the decision-making process. This chapter presents the importance of financial analytics and highlights the trends and prospects of it in the subject area of the decision-making process. To approach this issue, a Bibliometrix was applied based on R package. Data were retrieved from Scopus database and analyzed with the use of Biblioshiny and VOSviewer software.

This study examines the effects, in financial statements, from different allocations of bonds, a characteristic type of debt instrument according to business models introduced by IFRS 9. Manager discretion in allocating bonds to their investment portfolios, and specifically bank managers, who invest significant amounts in those types of assets, can lead to significant differences in figures, for the same bonds, especially in periods of relative financial stability. The findings of this study suggest that excess “freedom” allowed by the new standard can lead to distortions for each period banks report under IFRS, in accordance with managers’ decision for initial classification and subsequent measurement.

This study provides empirical evidence that underpricing is larger for more geographically dispersed firms when using a measure that captures the number of states in which firms have economic interests. The findings show that the average underpricing for local firms is 4.85% less than for dispersed firms (firms that have economic interests in more than three states in the USA). The hypothesis that underpricing is larger for more geographically dispersed firms is confirmed, and the evidence is robust for alternative measures of geographic dispersion. Results reveal that the likelihood of a firm committing accounting fraud increases the more geographically dispersed a firm’s economic interests become.

Algorithm output profit profiles from the Nixon algorithm (RGZ Ltd.) are used to analyse the benefits of diversification within many commodity and asset class sectors in order to generate a superior portfolio profile. The metrics developed are the algorithm optimisation metric (AOM) and the parameter sensitivity index (PSI). The former accounts for noise and stability in profit profiles and optimises algorithms and portfolios, yielding superior return-risk characteristics. The latter measures the stability of a given algorithm’s parameters and proportional changes in profits with respect to each parameter. Comparing these portfolio profits with those of more standard portfolios, we demonstrate the superiority of the developed metrics. The alignment of data is found to be a significant factor. Optimising a portfolio with unaligned data outputs leads to incorrect portfolio weightings and an erroneous profit profile on back-tested data. Correlations of prices and algorithmic returns are analysed showing the resultant dilution of correlation due to the effect of the strategy and the trading of security spreads.

This study examines the level acquisition companies (SPACs). The results demonstrate low underpricing for both types of SPAC, with unit and share prices of around $10 from 2010 to 2021. Leverage, market capitalisation, the size measured by total assets, and management teams with finance experience have a statistically significant impact on underpricing. Interestingly, the management team affected the share price (closing price) when the SPACs merged with the target companies on the first trading day. SPACs appear to be an alternative in comparison with IPOs. Furthermore, the relevance of agency theory, information asymmetry theory, signalling theory, and the winner’s curse is confirmed. The results provide practical implications for private target companies and investors that are interested in SPACs.

We implement a new framework to mitigate the errors-in-variables (EIV) problem in the estimation of asset pricing models. Considering an international data of portfolio stock returns from 1990 to 2021 widely used in empirical studies, we highlight the importance of the estimation method in time-series regressions. We compare the traditional ordinary-least squares (OLS) method to an alternative estimator based on a compact genetic algorithm (CGA) in the case of the CAPM. Based on intercepts, betas, adjusted R2, and the Gibbons et al. (1989) test, we find that the CGA-based method outperforms overall the OLS method. In particular, we obtain less statistically significant intercepts, smoother R2 across different portfolios, and lower GRS test statistics.

Specifically, in line with Roll’s critique (1977) on the unobservability of the market portfolio, we reduce the attenuation bias in market risk premium estimates. Moreover, our results are robust to alternative methods such as instrumental variables estimated with generalized-method of moments (GMM). Our findings have several empirical and managerial implications related to the estimation of asset pricing models as well as their interpretation as a popular tool in terms of corporate financial decision-making.