| SEMINAR ŞTIINŢIFIC MONEDA, FINANŢE, BĂNCI – ARHIVĂ |
Measuring, Modeling, and Forecasting Heat Discomfort, Day and Night
Marți, 9 Aprilie 2024 – ora 16:35 – sala 3M4 (etaj 1, clădirea Moxa)
Dan Gabriel Anghel
Bucharest University of Economic Studies; Institute for Economic Forecasting, Romanian Academy, dan.anghel@fin.ase.ro
Abstract:
Climate change makes heat-related human discomfort increasingly important for health outcomes, economic outcomes, and overall well-being. Heat-induced discomfort can be felt throughout the day, with both diurnal (e.g., daily maximum,
MAX) and nocturnal (e.g., daily minimum, MIN) temperatures contributing. In this paper we propose (1) a heat discomfort measurement framework that jointly incorporates MAX and MIN, and (2) a joint dynamic VARX-ACDX-Copula modeling framework for MAX and MIN. We then combine the measurement and modeling frameworks to examine past and present heat discomfort risk, and to forecast its future, using the city of Philadelphia as a detailed case study. We find that climate change has significantly impacted the mean, skewness, and kurtosis of both MAX and MIN in recent decades, in a way that feeds very strongly into increased overall heat discomfort risk, which we forecast to continue increasing through the century.
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: TBD
Stock markets co-movement during times of crisis
Marți, 26 Martie 2024 – ora 16:35 – sala 3M4 (etaj 1, clădirea Moxa)
Andreea Iordache
Bucharest University of Economic Studies, andreea.iordache@fin.ase.ro
Abstract:
This paper examines the links between capital markets from European Union. The sample includes a pre-pandemic period (January 2016 – December 2019) and a COVID-19 pandemic period (January 2020 – December 2023). Based on correlation coefficients and non-linear Granger causality test, our results show an increase in the financial market integration after the outbreak of the COVID-19 pandemic. Our findings provide a better understanding of the dynamics of European capital markets and offer valuable insight for investors and policymakers.
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: La Radu
Exploring Bitcoin Trends via AI Analysis of Facebook Emotions
Marți, 12 Martie 2024 – ora 16:35 – sala 3M4 (etaj 1, clădirea Moxa)
Daniel Traian Pele
Bucharest University of Economic Studies; Institute for Economic Forecasting, Romanian Academy, danpele@ase.ro
Abstract:
In the dynamic world of cryptocurrency trading, the influence of social media sentiment on market trends has become a critical research focus. This investigation delves into the complex relationship between the sentiments expressed on Facebook regarding Bitcoin and the fluctuations in Bitcoin’s trading volumes. By analyzing a rich dataset comprising 120,000 Facebook posts tagged with Bitcoin-centric keywords and detailed financial data from CoinGecko, we adopt a sophisticated analytical approach that combines the strengths of Long Short-Term Memory (LSTM) networks, sentiment analysis powered by FinBERT, and topic modeling strategies. The unique capability of LSTM models to process and understand time-related data nuances is especially valuable in the unpredictable realm of cryptocurrency markets. This attribute enables these models to surpass traditional time-series models in forecasting accuracy. Our research outcomes demonstrate a pronounced link between social media sentiments and Bitcoin market movements. These connections not only affirm the premise that social media sentiment is a viable predictor for market shifts but also highlight the enhanced accuracy of volume forecasts achieved through LSTM networks.
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: Începând cu ora 18:00: Enso Cafe
Dividend policy and democracy
Marți, 27 Februarie 2024 – ora 16:30 – sala 3M4 (etaj 1, clădirea Moxa)
Victor Dragotă
Bucharest University of Economic Studies, victor.dragota@fin.ase.ro
Abstract:
My presentation is based on a joint work, with Ingrid-Mihaela Dragotă, Cosmin-Octavian Cepoi and Hanaan Yaseen. Dividend policy is a constant concern in financial literature. We use data from 120,690 firm-year observations from 12,069 companies in 59 countries, in the period 2013-2022. Our sample covers both a pre-COVID-19 period, but also a COVID-19 one. According to financial literature, dividend policy is explained by many indicators. We add democracy to this list of determinants, using different proxies. We test for the robustness of the results.
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: TBD
Rolul Trezoreriei Statului in peisajul financiar romanesc
Marți, 16 Ianuarie 2024 – ora 16:30 – sala 3M4 (etaj 1, clădirea Moxa)
Gyorgy Attila
Bucharest University of Economic Studies, attila.gyorgy@fin.ase.ro
Abstract:
Trezoreria Statului a aniversat recent 20 de ani de functionare. In acesta perioada au avut loc numeroase schimbari socio-politice, unii punand la indoiala necesitatea existentei Trezoreriei Statului in detrimentul externalizarii operatiunilor financiare ale sectorului public catre sectorul bancar. Prezentarea vine sa prezinte cu ajutorul datelor statistice volumul si tipologia operatiunilor derulate prin trezoreria statului in ultimii ani, precum si provocarile in fata carora se afla institutia.
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: TBD
Estimating an optimal capital target for banks
Marți, 12 Decembrie 2023 – ora 16:30 – sala 3M4 (etaj 1, clădirea Moxa)
Bogdan Moinescu
Bucharest University of Economic Studies and National Bank of Romania, bogdan.moinescu@fin.ase.ro
Abstract:
Although it is generally accepted that higher capital ratios mean safer banking, a key issue of the ongoing regulatory review that still raises debates in the academic and prudential policy spheres is what robust capital levels actually mean. Our work revisits the role of solvency ratios, both risk-weighted and simple gearing, in banking stability assessment, from both the macro- and micro-prudential perspectives, by strictly aligning the operationalization of the bank soundness concept with the likelihood of stress. We use classification algorithms to provide risk ladders for both capital metrics, based on both the incidence of banking crises identified by the ECB/ESRB for the systemic layer, as well as on the EBA’s inventory of commercial bank failures in the EU for institutional-level insights. The early warning systems are then run to compute the capital shortfall against a presumptive probability standard of the supervisory authority, given macrofinancial developments, banks’ exposure level, and their financial control capabilities. Our findings are particularly relevant for policymakers, supervisors, and bankers looking to enhance the adequacy process of prudential buffers.
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: TBD
Investigating the impact of interbank interest rate on information asymmetry. Evidence from Bucharest Stock Exchange
Marți, 5 Decembrie 2023 – ora 16:30 – sala 3M4 (etaj 1, clădirea Moxa)
Mihai Danuț
Bucharest University of Economic Studies, danut.mihai@fin.ase.ro
Abstract:
The aim of this article is to examine the relationship between several economic and financial variables and the 3-month Romanian Interbank Offered Rate (ROBOR3M) in Romania. Using a database that covers August 2019 to October 2020 period, our results show that information asymmetry is influencing the interest rate evolution in such way that during episodes of increased uncertainty, a higher level of informed traders are executing transactions on Bucharest Stock Exchange which leads to higher interbank rates. Moreover, our study uncovers statistically significant associations between ROBOR3M and various factors, including prior ROBOR performance, shifts in currency exchange rates, movements in the stock market, fluctuations in cryptocurrency values, alterations in government bond yields, and the presence of information asymmetry among major banks.
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: TBD
Data-rich model for the Romanian economy – a hybrid DSGE-DFM approach
Marți, 28 Noiembrie 2023 – ora 16:30 – sala 3M4 (etaj 1, clădirea Moxa)
Georgiana Pleșa
Bucharest University of Economic Studies and National Bank of Romania, georgiana.plesa@fin.ase.ro
Abstract:
Dynamic Stochastic General Equilibrium models (DSGE) are widely used to assess the dynamics of macroeconomic variables in response to monetary and fiscal shocks. Standard methods assume that each economic variable is solely related to a single indicator and all the relevant information is contained by a small data set. Recently, when the economy faced heightened uncertainty and a multitude of adverse and persistent shocks, the behavior of macroeconomic indicators became difficult to be anticipated. Recent empirical research proved that the estimation of DSGE model that exploits information from a large panel of data possibly outperforms classical DSGE. Thus, we propose in this paper an analysis for Romanian economy, based on the DSGE-DFM method. This involves a canonical Smets and Wouters (2003; 2007) model to which add the financial accelerator into a data-rich environment and which is estimated with dynamic factor analysis. The results covering the period 2008Q1-2022Q4 point out that if we investigate only a limited amount of information, there might be the possibility to overestimate the effects of several negative shocks.
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: TBD
Metaheuristics for Portfolio Selection
Marți, 21 Noiembrie 2023 – ora 16:30 – sala 3M4 (etaj 1, clădirea Moxa)
Giacomo di Tollo
University of Sannio, giacomo.ditollo@unisannio.it
Abstract:
Portfolio Selection (PSP) is amongst the most studied issues in finance: it is concerned with selecting, out of a given set of assets, what assets to invest in and by how much. PSP becomes computationally intractable even for small sized instances, and many methods have been proposed to solve real-world formulation of the problem. Amongst them, metaheuristics are gaining more and more importance, due to their capability to produce good-quality solutions in a reasonable time. In our seminar, we are outlining metaheuristic approaches for the Portfolio Selection problem, along with some recent extensions made to tackle formulations that include short sellings.
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: TBD
A robust investment-consumption optimization problem in a switching regime interest rate setting
Marți, 14 Noiembrie 2023 – ora 16:30 – sala 3M4 (etaj 1, clădirea Moxa)
Bogdan Iftimie
Bucharest University of Economic Studies, iftimieb@csie.ase.ro
Abstract:
A classical utility optimization problem is concerned with the maximiza-tion of the expected utility of final wealth under a certain reference probabil-ity measure. However, in reality an agent doesn’t know the ”right” probabil-ity measure which models the distributions of traded assets, and in order toface the risks induced by considering a ”misleading” probability it is naturalto consider the worst possible situation the agent may be confronted with.Therefore, in this situation we are lead to maximize the expected utility ofthe final wealth over all admissible strategies, and minimize over some set ofprobability measures (called also priors) which we assume that do not vary”too much” from some reference probability P0.This presentation is based on a paper dealing with a robust investment-consumption optimization problem in an incomplete market with a switchingregime stochastic interest rate. The methodology combines duality approachwith stochastic control techniques (applied to the dual problem) specific to anon-Markovian setting, such as dynamic programming principle and Back-ward Stochastic Differential Equations (BSDEs) theory. An auxiliary dualproblem is established by means of infinite-dimensional convex duality. Ex-plicit formulas are derived for the optimal trading strategy and consumptionrate in terms of the solution of some nonstandard BSDE with jumps. Linksto other significant results in the domain are also provided.
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: TBD
Banks’ stock market reaction to fiscal and financial stability policy announcements during Covid-19 crisis. The role of central bank independence
Marți, 7 Noiembrie 2023 – ora 16:30 – sala 3M4 (etaj 1, clădirea Moxa)
Maura Andreea Bobiceanu
Babes-Bolyai University, Faculty of Economics and Business Administration, andreea.bobiceanu@econ.ubbcluj.ro
Abstract:
Employing an event study of bank stock prices, we examine investors’ responses to the fiscal and financial stability policy announcements across 127 European banks. Overall, we find that fiscal and financial stability measures lead to negative abnormal returns in bank stocks. This effect is intensified in countries with a higher political independence of central bank and supervisors, while greater operational independence of central bank and supervisory forbearance moderate stock market reaction..
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: TBD
Credit risk management in a low default portfolio setting using Machine Learning
Marți, 31 Octombrie 2023 – ora 16:30 – sala 3M4 (etaj 1, clădirea Moxa)
Codrut Ivascu
Bucharest University of Economic Studies, codrut.ivascu@fin.ase.ro
Abstract:
Credit risk is the most important source of risk for any lending institution. In order to quantify it and mitigate it, various and sophisticated models have been developed. Among those, machine learning algorithms showed the most promising results and have been studied intensively in the last 2 decades. However, one of the biggest concerns is the quality of public available datasets that have different distributions of default/non-default observations than the real-life cases. This paper uses an original dataset with a low default ratio (less than 2.3%) in order to evaluate the performance of the most popular ML algorithms. Numerous resampling techniques have been applied in order to rebalance the dataset. The results suggest that resampling approaches greatly improve the overall results with Neighbourhood Cleaning Rule and Borderline SMOTE SVM performing the best. Among ML algorithms, decision tree models outperform by a great margin Neural Networks or Support Vector Machines. The statistical significance of the results has been ensured by a two-way ANOVA approach.
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: TBD
Volatility Analysis And Cryptocurrency Interconnections
Marți, 24 Octombrie 2023 – ora 16:30 – sala 3M4 (etaj 1, clădirea Moxa)
Maria-Cristina Zwak-Cantoriu
Bucharest University of Economic Studies, maria.cantoriu@fin.ase.ro
Abstract:
The main purpose of this paper is to investigate the volatility in the cryptocurrency market and the relationships between them, using the cryptocurrencies: Bitcoin, Cardano and Stellar. The results obtained suggest that volatility in cryptocurrency prices is influenced by previous events and the level of past volatility. The propagation of volatility and shocks between the three analyzed cryptocurrencies, and the observation of the interconnection between them was also followed in this paper. By determining the degree of correlation between the analyzed cryptocurrencies, we observed that in the selected period there are quite strong positive correlations. Also, we noticed that the volatility of cryptocurrencies was strongly influenced by certain economically uncertain periods, a fact that caused their prices to have a strong fluctuation, especially in the period 2021-2023. The results obtained show an interdependence between the three cryptocurrencies, which is significant in the decision-making of investors. News and events play a significant role in the cryptocurrency market, especially those in the financial, technological, and political worlds that can have a considerable impact on cryptocurrency prices and volatility. In the case of Bitcoin, there was an increase in interest in this cryptocurrency when there were investments from large companies and financial institutions.
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: TBD
Bank Financial Strength Ratings vs. Credit Ratings: Evidence from the CDS Market
Marți, 17 Octombrie 2023 – ora 16:30 – sala 3M4 (etaj 1, clădirea Moxa)
Iuliana Ismailescu
Pace University, iismailescu@pace.edu
Abstract:
This paper analyzes the reaction of the credit default swap (CDS) market to bank financial strength rating (BFSR) and credit rating announcements made by Kroll Bond Rating Agency (KBRA) and Moody’s Investors Service (Moody’s). Our findings show an inconsistent response of CDS spreads to KBRA and Moody’s BFSR changes. They exhibit a strong but short-lived reaction to KBRA BFSR downgrades and a significant post-announcement response to its upgrades. In contrast, Moody’s BFSR negative changes have a longer-lasting effect, but its BFSR upgrades leave only sporadic marks on CDS spreads. Finally, we find that the CDS market cannot separate the informational content of the BFSRs and credit ratings issued by the same rating agency, raising questions about the value of BFSRs to credit markets.
Zoom:
https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: TBD
Does the Level of Financial Development Matter for the Fiscal Response? A PSTR Approach for the EU and Selected OECD Countries
Marți, 10 Octombrie 2023 – ora 16:30 – sala 3M4 (etaj 1, clădirea Moxa)
Andreea Stoian
Bucharest University of Economic Studies, andreea.stoian@fin.ase.ro
Abstract:
This paper empirically studies the role of financial development in the fiscal response function by distinguishing high and low financial development regimes based on a data driven selection mechanism and investigating two groups of economies, European Union (EU) members and OECD countries from 2000 to 2019. Applying the panel smooth transition regression to separate two regimes based on the status of financial development we find a threshold of about 0.60 of financial development (FD) on average in the EU, a little higher in the OECD. Our results indicate that the stance of financial development matters for fiscal policy design and business cycle behavior. The response effectively differs for low financial development regime situations compared to high financial development regimes, both, with regard to debt sustainability as well as the output gap in the EU and the OECD, too. While in the low FD regime debt indicates to be sustainable and the business cycle performs pro-cyclically, in high FD regimes the picture changes to a more mixed behavior including debt non-sustainability and counter-cyclical output manner. This holds true for both, the EU as well as the OECD. Thus, our results indicate that financial markets do fuel (fiscal) policy behavior. Too much financial development can influence particularly the debt reaction and the business cycle.
Zoom: https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: Începând cu ora18:00: J’ai bistro
The Volatility Network of US Industries and Investor Sentiment
Marți, 3 Octombrie 2023 – ora 16:30 – sala 3M4 (etaj 1, clădirea Moxa)
Petre Caraiani
Bucharest University of Economic Studies and IPE, Academia Română, petre.caraiani@fabiz.ase.ro
Abstract:
We investigate the influence of social and news sentiment on the high-frequency weighted volatility network of industry portfolios in the United States. We find that positive sentiment has no impact on the network, but negative sentiment associated to stocks in the ID, CS-C, FN and IT economic sectors have. The results imply that negative social and news sentiment has a significant impact on systemic risk.
Zoom: https://ase.zoom.us/j/88975840175?pwd=Q0ZnbHl0TjhqcEtzVnhMUWRmdWdPQT09
Activitate de socializare: Începând cu ora 18:00: J’ai Bistrot