Research articles for the 2021-03-28

A Genetic Algorithm approach to Asymmetrical Blotto Games with Heterogeneous Valuations
Aymeric Vie

Blotto Games are a popular model of multi-dimensional strategic resource allocation. Two players allocate resources in different battlefields in an auction setting. While competition with equal budgets is well understood, little is known about strategic behavior under asymmetry of resources. We introduce a genetic algorithm, a search heuristic inspired from biological evolution, interpreted as social learning, to solve this problem. Most performant strategies are combined to create more performant strategies. Mutations allow the algorithm to efficiently scan the space of possible strategies, and consider a wide diversity of deviations. We show that our genetic algorithm converges to the analytical Nash equilibrium of the symmetric Blotto game. We present the solution concept it provides for asymmetrical Blotto games. It notably sees the emergence of "guerilla warfare" strategies, consistent with empirical and experimental findings. The player with less resources learns to concentrate its resources to compensate for the asymmetry of competition. When players value battlefields heterogeneously, counter strategies and bidding focus is obtained in equilibrium. These features are consistent with empirical and experimental findings, and provide a learning foundation for their existence.

Amnesty Policy and Elite Persistence in the Postbellum South: Evidence from a Regression Discontinuity Design
Jason Poulos

This paper investigates the impact of Reconstruction-era amnesty policy on the officeholding and wealth of planter elites in the postbellum U.S. South. Amnesty policy restricted the political and economic rights of the planter class for nearly three years during Reconstruction. The paper estimates the effects of being excepted from amnesty on elites' future wealth and political power using a regression discontinuity design. Results on a sample of delegates to Reconstruction conventions show that exclusion from amnesty substantially decreases the likelihood of holding political office. I find no evidence that exclusion from amnesty impacted later census wealth for Reconstruction delegates or for a larger sample of known slaveholders who lived in the South in 1860. These findings are in line with previous studies evidencing changes to the identity of the political elite and continuity of economic mobility for the planter elite across the Civil War and Reconstruction.

An Empirical View of Peer-to-Peer (P2P) Lending Platforms
Ofir, Moran,Tzang, Ido
Peer-to-Peer (P2P) lending emerged over a decade ago and quickly evolved into a global industry. Since then, the P2P lending industry has become more complex, with increasingly diverse types of business models, each involving different risks and challenges. The Article analyzes the current state of the P2P market by exploring the different business models, the platforms' characteristics, the global market trends, and the different regulatory regimes around the world. As the COVID-19 pandemic bears an unprecedented effect on the global economy, we analyze its impact on P2P markets, especially focusing on small and medium entities (SMEs) as borrowers. While the COVID-19 crisis has had a profound impact on SMEs’ access to funding, alternative finance solutions, especially digital solutions, such as P2P lending, have come to play a crucial role in reducing the risk of bankruptcy for SMEs. In light of this developing situation, we empirically analyze a rich and detailed data set on loans given by a large P2P platform to SMEs between the years 2014 and 2020, focusing on the interest rate set by the platform for both borrowers and lenders. Our main findings regarding the borrowers indicate that the interest rate decreases as the size of the loan increases; however, the rate goes up with the duration of the loan and from year to year. Significant differences in loan interest rates were found across loan statuses, corporation types, industries, and the genders of the SME owners. As for lenders, we show that the average interest rate increases with the size of the loan and decreases with the number of loans into which the investment is divided. The empirical findings highlight the significant variables affecting the interest rate, which is the most important feature of a loan, and the conclusions drawn in this study can thus serve both regulators and policy makers in designing their future responses to the evolving and growing market challenges, especially in these times of global health and economic crisis.

Analyst Optimism and Buy-Side Institutions: Evidence from Analyst Transition from Sell-Side to Buy-Side
Zhang, Biwen
Using career information collected from professional networking sites, I identify sell-side analysts who transition to buy-side institutions, and examine whether these transitioning analysts cater to their future buy-side employers before the transition. I find that, prior to the transition, transitioning analysts produce more favorable research toward stocks that are significant in the portfolios of their future employers. Importantly, I find the favoritism effects are stronger for stocks for which a single analyst's impact is likely to be large, namely, small-cap stocks and stocks with low analyst coverage. I also find such favoritism is present only during the year immediately prior to the transition. Moreover, the favoritism is present only among analysts who immediately transition to the buy-side, and not among those who move to the buy-side after a prolonged transition gap (i.e., where strategic behavior is less likely). These findings are consistent with the career concern hypothesis that sell-side analysts cater to their future buy-side employers to advance their careers.

Asset Selection via Correlation Blockmodel Clustering
Wenpin Tang,Xiao Xu,Xun Yu Zhou

We aim to cluster financial assets in order to identify a small set of stocks to approximate the level of diversification of the whole universe of stocks. We develop a data-driven approach to clustering based on a correlation blockmodel in which assets in the same cluster have the same correlations with all other assets. We devise an algorithm to detect the clusters, with a theoretical analysis and a practical guidance. Finally, we conduct an empirical analysis to attest the performance of the algorithm.

Bank Income Smoothing during the COVID-19 Pandemic: Evidence from UK Banks
Ozili, Peterson K
This paper investigates the relationship between loan loss provisions and pre-provisions earnings during the COVID-19 pandemic. A positive co-movement between the two variables indicates income smoothing. The pandemic period was compared with the pre-pandemic period using quarterly data of four systemic banks in the UK from Q1 2019 to Q4 2020. The findings show that loan loss provisions exhibit a V-shaped property during the pandemic. Provisions reached its highest level at the peak of the pandemic in Q2 2020 and declined in the subsequent quarters. The regression results show a positive relationship between loan loss provisions and pre-provisions earnings both in the pre-pandemic and pandemic quarters. The relationship is stronger in the pandemic period than in the pre-pandemic period which indicates greater income smoothing in the pandemic period. The correlation results also show a strong positive correlation between loan loss provisions and pre-provisions earnings in the pandemic period. In the individual bank analysis, three of the four systemic banks exhibit greater income smoothing in the pandemic period than in the pre-pandemic period.

CEO Endowed Trait and Corporate Tax Avoidance: Evidence from Pilot CEOs
Baghdadi, Ghasan,Podolski, Edward,Veeraraghavan, Madhu
We find evidence that CEOs’ hobby of flying airplanes is associated with significantly lower effective tax rates and a greater propensity to engage in the most aggressive forms of tax avoidance such as tax sheltering. The effect is economically as well as statistically significant. Cross-sectional tests reveal that the baseline results are not sensitive to managerial remuneration incentives, suggesting that intrinsic incentives derived from endowed traits are not easily moderated by extrinsic motivation from compensation contracts. However, our analysis shows that the baseline result only holds in settings where managers are subject to high levels of monitoring by institutional shareholders, suggesting that strong managerial oversight helps direct managerial thrill-seeking tendencies towards value creating endeavors. Taken together, our paper highlights the significant role that CEOs thrill-seeking tendencies play in driving corporate tax planning activities.

Confronting Machine Learning With Financial Research
Kristof Lommers,Ouns El Harzli,Jack Kim

This study aims to examine the challenges and applications of machine learning for financial research. Machine learning algorithms have been developed for certain data environments which substantially differ from the one we encounter in finance. Not only do difficulties arise due to some of the idiosyncrasies of financial markets, there is a fundamental tension between the underlying paradigm of machine learning and the research philosophy in financial economics. Given the peculiar features of financial markets and the empirical framework within social science, various adjustments have to be made to the conventional machine learning methodology. We discuss some of the main challenges of machine learning in finance and examine how these could be accounted for. Despite some of the challenges, we argue that machine learning could be unified with financial research to become a robust complement to the econometrician's toolbox. Moreover, we discuss the various applications of machine learning in the research process such as estimation, empirical discovery, testing, causal inference and prediction.

Contractual Evolution
Jennejohn, Matthew,Nyarko, Julian,Talley, Eric L.
Conventional wisdom portrays contracts as static distillations of parties’ shared intent at some discrete point in time. In reality, however, contract terms evolve in response to their environments, including new laws, legal interpretations, and economic shocks. While several legal scholars have offered stylized accounts of this evolutionary process, we still lack a coherent, general theory that broadly captures the dynamics of real-world contracting practice. This paper advances such a theory, in which the evolution of contract terms is a byproduct of several key features, including efficiency concerns, information, and sequential learning by attorneys who negotiate several deals over time. Each of these factors contributes to the underlying evolutionary process, and their relative prominence bears directly on the speed, direction, and desirability of how contractual innovations diffuse. Using a formal model of bargaining in a sequence of similar transactions, we demonstrate how different evolutionary patterns can manifest over time, in both desirable and undesirable directions. We then take these insights to real-world dataset of over 2,000 merger agreements negotiated over the last two decades, tracking the adoption of several contractual clauses, including pandemic-related terms, #MeToo provisions, CFIUS conditions, and reverse termination fees. Our analysis suggests that there is not a “one size fits all” paradigm for contractual evolution; rather, the constituent forces affecting term evolution appear manifest in varying strengths across differing circumstances. We highlight several constructive applications of our framework, including the study of contract negotiation unfolds when price cannot easily be adjusted, and how to incorporate other forms of cognitive and behavioral biases into our general framework.

Domain Specific Concept Drift Detectors for Predicting Financial Time Series
Filippo Neri

Concept drift detectors allow learning systems to maintain good accuracy on non-stationary data streams. Financial time series are an instance of non-stationary data streams whose concept drifts (market phases) are so important to affect investment decisions worldwide. This paper studies how concept drift detectors behave when applied to financial time series. General results are: a) concept drift detectors usually improve the runtime over continuous learning, b) their computational cost is usually a fraction of the learning and prediction steps of even basic learners, c) it is important to study concept drift detectors in combination with the learning systems they will operate with, and d) concept drift detectors can be directly applied to the time series of raw financial data and not only to the model's accuracy one. Moreover, the study introduces three simple concept drift detectors, tailored to financial time series, and shows that two of them can be at least as effective as the most sophisticated ones from the state of the art when applied to financial time series.

Evolutionary Strategies with Analogy Partitions in p-guessing Games
Aymeric Vie

In Keynesian Beauty Contests notably modeled by p-guessing games, players try to guess the average of guesses multiplied by p. Convergence of plays to Nash equilibrium has often been justified by agents' learning. However, interrogations remain on the origin of reasoning types and equilibrium behavior when learning takes place in unstable environments. When successive values of p can take values above and below 1, bounded rational agents may learn about their environment through simplified representations of the game, reasoning with analogies and constructing expectations about the behavior of other players. We introduce an evolutionary process of learning to investigate the dynamics of learning and the resulting optimal strategies in unstable p-guessing games environments with analogy partitions. As a validation of the approach, we first show that our genetic algorithm behaves consistently with previous results in persistent environments, converging to the Nash equilibrium. We characterize strategic behavior in mixed regimes with unstable values of p. Varying the number of iterations given to the genetic algorithm to learn about the game replicates the behavior of agents with different levels of reasoning of the level k approach. This evolutionary process hence proposes a learning foundation for endogenizing existence and transitions between levels of reasoning in cognitive hierarchy models.

Forecasting with Deep Learning: S&P 500 index
Firuz Kamalov,Linda Smail,Ikhlaas Gurrib

Stock price prediction has been the focus of a large amount of research but an acceptable solution has so far escaped academics. Recent advances in deep learning have motivated researchers to apply neural networks to stock prediction. In this paper, we propose a convolution-based neural network model for predicting the future value of the S&P 500 index. The proposed model is capable of predicting the next-day direction of the index based on the previous values of the index. Experiments show that our model outperforms a number of benchmarks achieving an accuracy rate of over 55%.

From the 2008-2009 Financial Crisis to the COVID-19 Pandemic: Challenges and the Way Forward
Milas, Costas
I discuss some of the (ongoing) challenges and (possible) responses of the economics and finance profession following the 2008-2009 financial crisis and the COVID-19 pandemic. I do this in the context of three examples. The first example draws on the Bank of England’s forward guidance on interest rates by flagging differences between real-time and revised data. Noting that financial factors have become more prominent in the empirical modelling of the economy since the financial crisis, the second example considers their ability to explain UK GDP movements conditional on COVID-19 related information. In this example I estimate that social distancing and lockdown restrictions reduced, on average, annual UK growth by 10.1 percentage points compared to the scenario of no government action. At the other extreme, had government stringency remained at its April 2020 ‘lockdown level’ throughout the pandemic, annual UK growth would have been lower by (a further) 3 percentage points on average compared to the impact of the imposed restrictions. I also estimate that social distancing and lockdown restrictions reduced, on average, annual CPI inflation by 0.57 percentage points compared to the scenario of no government action. At the other extreme, had government stringency remained at its April 2020 ‘lockdown level’ throughout the pandemic, annual CPI inflation would have been lower by (a further) 0.68 percentage points on average compared to the impact of the imposed restrictions. The third example draws on the potential usefulness of Google mobility data to explain UK GDP.

Growth Risks, Asset Prices, and Welfare
Croce, Mariano (Max) Massimiliano
I connect interest rates, risk premia and welfare costs of long-run consumption uncertainty in a setting with Epstein and Zin (1989) preferences. I find that long-run uncertainty can create significant welfare costs even when risk aversion is moderate and the short-run consumption volatility low. I document that the risk-free rate puzzle is a key determinant of the welfare costs.

How Do Institutional Investors React to Geographically Dispersed Information Shocks? A Test Using the COVID-19 Pandemic
Ling, David C.,Wang, Chongyu,Zhou, Tingyu
We test how institutional investors respond to geographically dispersed information shocks during periods of market turmoil. Specifically, using a sample of Real Estate Investment Trusts (REITs) that links the locations of investors, REIT firms, and the assets held by REITs, we find that institutional investors reduced their holdings in response to local information shocks in the early stage of the COVID-19 pandemic. Local ownership declined more in more heavily-affected property markets. In addition, the response to local information shocks was larger in markets with larger portfolio allocations by REITs and in markets that are home to the investors. The decline in local ownership was also more pronounced among motivated and nonpassive investors. Our study sheds light on the effects of local information shocks on the formation of investors’ expectations during market crisis.

How Much Does Your Banker’s Target-Specific Experience Matter? Evidence from Target IPO Underwriters that Advise Acquirers
Herron, Richard
In a sample of 1,507 US all-public acquisitions from 1985â€"2014, 5% of acquirers use the same advisor that underwrote the target’s initial public offering. Acquirers who use these informed advisors have acquisition announcement three-day cumulative abnormal returns (CARs) that are 2.048 percentage points higher, all else equal. Same-advisor acquisition announcements have higher combined CARs but not lower target CARs, suggesting higher synergies instead of lower deal premia. Same-advisor acquisition announcement outperformance decays as the target ages and grows. These findings show the value and timeliness of investment bank information production, retention, and transfer.

Inequality, Identity, and Partisanship: How redistribution can stem the tide of mass polarization
Alexander J. Stewart,Joshua B. Plotkin,Nolan McCarty

The form of political polarization where citizens develop strongly negative attitudes towards out-party policies and members has become increasingly prominent across many democracies. Economic hardship and social inequality, as well as inter-group and racial conflict, have been identified as important contributing factors to this phenomenon known as "affective polarization." Such partisan animosities are exacerbated when these interests and identities become aligned with existing party cleavages. In this paper we use a model of cultural evolution to study how these forces combine to generate and maintain affective political polarization. We show that economic events can drive both affective polarization and sorting of group identities along party lines, which in turn can magnify the effects of underlying inequality between those groups. But on a more optimistic note, we show that sufficiently high levels of wealth redistribution through the provision of public goods can counteract this feedback and limit the rise of polarization. We test some of our key theoretical predictions using survey data on inter-group polarization, sorting of racial groups and affective polarization in the United States over the past 50 years.

Isolating the impact of trading on grid frequency fluctuations
Benjamin Schäfer,Marc Timme,Dirk Witthaut

To ensure reliable operation of power grids, their frequency shall stay within strict bounds. Multiple sources of disturbances cause fluctuations of the grid frequency, ranging from changing demand over volatile feed-in to energy trading. Here, we analyze frequency time series from the continental European grid in 2011 and 2017 as a case study to isolate the impact of trading. We find that trading at typical trading intervals such as full hours modifies the frequency fluctuation statistics. While particularly large frequency deviations in 2017 are not as frequent as in 2011, large deviations are more likely to occur shortly after the trading instances. A comparison between the two years indicates that trading at shorter intervals might be beneficial for frequency quality and grid stability, because particularly large fluctuations are substantially diminished. Furthermore, we observe that the statistics of the frequency fluctuations do not follow Gaussian distributions but are better described using heavy-tailed and asymmetric distributions, for example L\'evy-stable distributions. Comparing intervals without trading to those with trading instances indicates that frequency deviations near the trading times are distributed more widely and thus extreme deviations are orders of magnitude more likely. Finally, we briefly review a stochastic analysis that allows a quantitative description of power grid frequency fluctuations.

On the Competition and Transparency in Public Procurement During COVID-19 Pandemic in European Union
Kubak, Matus,Nemec, Peter,Vološin, Marcel
Public procurement accounts in developed countries for about 20% of gross domestic product, thus is seen as a policy implementation tool. During COVD-19 outbreak, public procurement served as a basic tool for equipping institutions and citizens with medical supplies, ventilators, and personal protective equipment. We use data on public procurement in European countries during first wave of COVID-19 pandemic, with aim to study transparency and competition in public procurement process during ongoing state of emergency. Quantitative analysis of the data indicates, that less transparent procurement procedures were primary used during acute outbreak of COVID-19 and that the level of transparency widely varies among countries regardless the extent of the epidemic. Data suggests that the use of less transparent procurement procedures and reduced competition is more suitable for small and medium enterprises, where share of tenders submitted by small and medium enterprises represents up to 87% of all less transparent procedures.

Opportunities, Constraints and Critical Supports for Achieving Sustainable Local Pharmaceutical Manufacturing in Africa: With a Focus on the Role of Finance, Final Report
Abbott, Frederick M.,Abbott, Ryan,Fortunak, Joseph,Gehl Sampath, Padmashree,Walwyn, David
The Open Society Foundations Public Health Program commissioned an interdisciplinary team to examine the extent to which gaps in financing are an obstacle to more robust local production of pharmaceutical products (including diagnostics, vaccines and treatments (DVT)) and personal protective equipment (PPE) with focus on Africa. The Report is posted here. The Executive Summary and an Addenda compilation (which includes 3 supplemental research reports) are posted along with the Report at the website of Nova Worldwide Consulting.The overall findings of the study are that gaps in financing pose a constraint on the localization of pharmaceutical production in Africa. The gaps are not mainly due to a lack of financial capital in global financial markets that might be deployed for this purpose. The main problems are associated with the market environment in the sense that sustainable business operations require adequate demand, and market demand for pharmaceutical products in Africa is limited by various factors. In addition, comparatively weak infrastructure (recognizing variation among countries) makes it difficult to compete with large efficient foreign suppliers that are bolstered by foreign government support. Potential investors appear to perceive relatively high risks associated with investing in pharmaceutical manufacturing in Africa.Transforming political engagement:The COVID-19 pandemic may serve as adequate impetus to transform local production of pharmaceuticals into a governmental priority. Government commitment at a high level is required to engage the financial levers that will support localization of production.Sustainable business models:Particularly outside the vaccine sector, successfully operating a pharmaceutical manufacturing facility means addressing a market with sufficient demand to generate revenue and profits. Alternatively, or as a supplement, governments may provide direct subsidies, guaranteed offtake agreements, tax credits, local production pricing premiums, and other measures to substitute for market demand. These forms of support are commonly used for “infant industries”, and there should be plans to withdraw such support once a business has achieved sustainability.The social impact investor market:African governments should consider a program to encourage sovereign wealth funds and other financial asset managers to invest in local production on the African continent as a way to accomplish important social goals. To facilitate this objective, there should be some type of backstop or guarantee of the social impact investments within reasonable parameters. The African Development Bank may be helpful in establishing mechanisms for this purpose.Opportunities for advocacy:There is substantial room for advocacy by civil society to move Africa toward greater self-sufficiency in the production of pharmaceutical products. At the high level of political commitment, government authorities should be persuaded to prioritize local production of pharmaceuticals as a matter of public health security, engaging the financial levers to support such a commitment. At the level of industrial policy, the African Union should be encouraged to engage in concrete planning for regional pharmaceutical production hubs, and associated infrastructure and centers of technical expertise. Whether in conjunction with that, or separately, procurement authorities should be encouraged to form regional pooled procurement mechanisms to aggregate demand, allow for more effective bargaining with suppliers, and support regional hub manufacturers.Support for effective implementation of the African Continental Free Trade Area in terms of reducing barriers to intra-Africa trade in pharmaceutical products would improve the market situation. Similarly, continuing support for efforts to integrate the African regional regulatory structure for pharmaceutical products would accelerate access to medicines. Establishment of a library of available drug master files for reference by manufacturers would significantly lower barriers to manufacturer market entry.Foundations should be encouraged to develop a transparent platform which could provide information to African manufacturers with respect to opportunities for financing and expertise for pursuing their objectives.

Putting a price on tenure
Thiago Marzagao

Government employees in Brazil are granted tenure after three years of taking their entrance exams. Firing a tenured government employee is all but impossible, so tenure is a big perquisite. But exactly how big is it? No one has ever attempted to estimate the monetary equivalent of tenure for Brazilian government workers. We do that in this paper. We use a modified version of the Sharpe ratio to estimate what the risk-adjusted salaries of government workers should be. The difference between actual salary and risk-adjusted salary gives us an estimate of how much tenure is worth for each employee. We find that the median value of tenure is 4517 reais for federal government employees, 2560 reais for state government employees, and 672 reais for municipal government employees.

Reliability of MST identification in correlation-based market networks
V.A. Kalyagin,A.P. Koldanov,P.A. Koldanov

Maximum spanning tree (MST) is a popular tool in market network analysis. Large number of publications are devoted to the MST calculation and it's interpretation for particular stock markets. However, much less attention is payed in the literature to the analysis of uncertainty of obtained results. In the present paper we suggest a general framework to measure uncertainty of MST identification. We study uncertainty in the framework of the concept of random variable network (RVN). We consider different correlation based networks in the large class of elliptical distributions. We show that true MST is the same in three networks: Pearson correlation network, Fechner correlation network, and Kendall correlation network. We argue that among different measures of uncertainty the FDR (False Discovery Rate) is the most appropriated for MST identification. We investigate FDR of Kruskal algorithm for MST identification and show that reliability of MST identification is different in these three networks. In particular, for Pearson correlation network the FDR essentially depends on distribution of stock returns. We prove that for market network with Fechner correlation the FDR is non sensitive to the assumption on stock's return distribution. Some interesting phenomena are discovered for Kendall correlation network. Our experiments show that FDR of Kruskal algorithm for MST identification in Kendall correlation network weakly depend on distribution and at the same time the value of FDR is almost the best in comparison with MST identification in other networks. These facts are important in practical applications.

Stock price forecast with deep learning
Firuz Kamalov,Linda Smail,Ikhlaas Gurrib

In this paper, we compare various approaches to stock price prediction using neural networks. We analyze the performance fully connected, convolutional, and recurrent architectures in predicting the next day value of S&P 500 index based on its previous values. We further expand our analysis by including three different optimization techniques: Stochastic Gradient Descent, Root Mean Square Propagation, and Adaptive Moment Estimation. The numerical experiments reveal that a single layer recurrent neural network with RMSprop optimizer produces optimal results with validation and test Mean Absolute Error of 0.0150 and 0.0148 respectively.

Superiority of mild interventions against COVID-19 on public health and economic measures
Makoto Niwa,Yasushi Hara,Yusuke Matsuo,Hodaka Narita,Lim Yeongjoo,Shintaro Sengoku,Kota Kodama

During the global spread of COVID-19, Japan has been among the top countries to maintain a relatively low number of infections, despite implementing limited institutional interventions. Using a Tokyo Metropolitan dataset, this study investigated how these limited intervention policies have affected public health and economic conditions in the COVID-19 context. A causal loop analysis suggested that there were risks to prematurely terminating such interventions. On the basis of this result and subsequent quantitative modelling, we found that the short-term effectiveness of a short-term pre-emptive stay-at-home request caused a resurgence in the number of positive cases, whereas an additional request provided a limited negative add-on effect for economic measures (e.g. the number of electronic word-of-mouth (eWOM) communications and restaurant visits). These findings suggest the superiority of a mild and continuous intervention as a long-term countermeasure under epidemic pressures when compared to strong intermittent interventions.