# Research articles for the 2020-09-28

An AI approach to measuring financial risk
Lining Yu,Wolfgang Karl Härdle,Lukas Borke,Thijs Benschop
arXiv

AI artificial intelligence brings about new quantitative techniques to assess the state of an economy. Here we describe a new measure for systemic risk: the Financial Risk Meter (FRM). This measure is based on the penalization parameter (lambda) of a linear quantile lasso regression. The FRM is calculated by taking the average of the penalization parameters over the 100 largest US publicly traded financial institutions. We demonstrate the suitability of this AI based risk measure by comparing the proposed FRM to other measures for systemic risk, such as VIX, SRISK and Google Trends. We find that mutual Granger causality exists between the FRM and these measures, which indicates the validity of the FRM as a systemic risk measure. The implementation of this project is carried out using parallel computing, the codes are published on www.quantlet.de with keyword FRM. The R package RiskAnalytics is another tool with the purpose of integrating and facilitating the research, calculation and analysis methods around the FRM project. The visualization and the up-to-date FRM can be found on hu.berlin/frm.

An Adaptive and Explicit Fourth Order Runge-Kutta-Fehlberg Method Coupled with Compact Finite Differencing for Pricing American Put Options
Chinonso Nwankwo,Weizhong Dai
arXiv

We propose an adaptive and explicit fourth-order Runge-Kutta-Fehlberg method coupled with a fourth-order compact scheme to solve the American put options problem. First, the free boundary problem is converted into a system of partial differential equations with a fixed domain by using logarithm transformation and taking additional derivatives. With the addition of an intermediate function with a fixed free boundary, a quadratic formula is derived to compute the velocity of the optimal exercise boundary analytically. Furthermore, we implement an extrapolation method to ensure that at least, a third-order accuracy in space is maintained at the boundary point when computing the optimal exercise boundary from its derivative. As such, it enables us to employ fourth-order spatial and temporal discretization with Dirichlet boundary conditions for obtaining the numerical solution of the asset option, option Greeks, and the optimal exercise boundary. The advantage of the Runge-Kutta-Fehlberg method is based on error control and the adjustment of the time step to maintain the error at a certain threshold. By comparing with some existing methods in the numerical experiment, it shows that the present method has a better performance in terms of computational speed and provides a more accurate solution.

An Agent-Based Model of Delegation Relationships With Hidden-Action: On the Effects of Heterogeneous Memory on Performance
Patrick Reinwald,Stephan Leitner,Friederike Wall
arXiv

We introduce an agent-based model of delegation relationships between a principal and an agent, which is based on the standard-hidden action model introduced by Holmstr\"om and, by doing so, provide a model which can be used to further explore theoretical topics in managerial economics, such as the efficiency of incentive mechanisms. We employ the concept of agentization, i.e., we systematically transform the standard hidden-action model into an agent-based model. Our modeling approach allows for a relaxation of some of the rather "heroic" assumptions included in the standard hidden-action model, whereby we particularly focus on assumptions related to the (i) availability of information about the environment and the (ii) principal's and agent's cognitive capabilities (with a particular focus on their learning capabilities and their memory). Our analysis focuses on how close and how fast the incentive scheme, which endogenously emerges from the agent-based model, converges to the solution proposed by the standard hidden-action model. Also, we investigate whether a stable solution can emerge from the agent-based model variant. The results show that in stable environments the emergent result can nearly reach the solution proposed by the standard hidden-action model. Surprisingly, the results indicate that turbulence in the environment leads to stability in earlier time periods.

An analysis of network filtering methods to sovereign bond yields during COVID-19
Raymond Ka-Kay Pang,Oscar Granados,Harsh Chhajer,Erika Fille Legara
arXiv

In this work, we investigate the impact of the COVID-19 pandemic on sovereign bond yields amongst European countries. We consider the temporal changes from financial correlations using network filtering methods. These methods consider a subset of links within the correlation matrix, which gives rise to a network structure. We use sovereign bond yield data from 17 European countries between the 2010 and 2020 period as an indicator of the economic health of countries. We find that the average correlation between sovereign bonds within the COVID-19 period decreases, from the peak observed in the 2019-2020 period, where this trend is also reflected in all network filtering methods. We also find variations between the movements of different network filtering methods under various network measures.

Banksâ€™ Complexity and Risk: Agency Problems and Diversification Benefits
SSRN
Bank complexity is often associated with risk, due to moral hazard and agency problems. At the same time, complexity may be linked to diversification and scale economies, thus leading to less risk. In this paper, we provide empirical evidence on the relationship between bank complexity and risk-taking. We find a positive relationship between geographical complexity and bank risk. Banks that operate in more countries, both through banks and non-banks, have riskier balance sheets and more non-performing loans. Further, banks that operate in Africa have higher risk levels due to larger volatility of returns. The link between structural complexity and bank risk is weaker, but generally negative. Our results suggest that moral hazard and agency problems may be more acute when banks operate in many geographies and in emerging market economies. In contrast, the results are consistent with diversification and scale benefits arising from operating in more business areas.

C-Sharpe Optimal Portfolio
Marinescu, Mircea
SSRN
The C-Sharpe ratio is defined as the amount of expected excess return per unit of risk, where the risk is given by the CVaR dispersion measure (not to be confused with the CVaR risk measure). Then, C-Sharpe optimal portfolio is the portfolio with the largest C-Sharpe ratio. It can be identified as the tangency CVaR efficient portfolio. We have shown that the C-Sharpe optimal portfolio weights can be evaluated by solving a straightforward linear programming (LP) problem. Numerical examples were carried out for a quarterly re-balanced portfolio of ETF's with long-only positions. The C-Sharpe optimal strategies tend to outperform, {\it i.e.} shallower draw-downs and higher rates of return, a similar strategy based on the mean-variance optimal portfolio with the largest Sharpe ratio. C-Sharpe optimal portfolio based strategies may be the preferred choice for an investor seeking some protection during draw-down events and reasonably high rates of return.

Calibration of exponential Hawkes processes using a Modified Bionomic Algorithm
Chen, Jing,Pierre, Sebastien
SSRN
The aim of this research is to develop a fast and robust variant of the evolutionary heuristic Bionomic algorithm and assess its contribution to solving complex parametric estimation problems, in conjunction with other traditional optimization techniques. We introduce a modified version of the Bionomic Algorithm (MB), designed to efficiently compute the MLE of self-exciting exponential Hawkes processes with increasing dimensionality of the solution space. Performance tests, performed on simulated and historical S&P 500 financial data, show that the MB algorithm, with its solutions locally improved by either the standard Nelder Mead (NM) or Expectation Maximization (EM) algorithm, converges significantly faster and more frequently to a near-global solution than the NM or EM algorithms operating alone. These test results illustrate the robustness and computational efficiency of the MB algorithm, combined with traditional optimization methods, in the optimization of complex objective functions of high dimensionality.

Client engineering of XVA in crisis and normality: Restructuring, Mandatory Breaks and Resets
Chris Kenyon
arXiv

Crises challenge client XVA management when continuous collateralization is not possible because a derivative locks in the client credit level and the provider's funding level, on the trade date, for the life of the trade. We price XVA reduction strategies from the client point of view comparing multiple trade strategies using Mandatory Breaks or Restructuring, to modifications of a single trade using a Reset. We analyse previous crises and recovery of CDS to inform our numerical examples. In our numerical examples Resets can be twice as effective as Mandatory Break/Restructuring if there is no credit recovery. When recovery is at least 1/3 of the credit shock then Mandatory Break/Restructuring can be more effective.

Could the Poor Bank on Stablecoins? Discussion Prompts for Innovators, Regulators, and Consumers
Calabia, F. Christopher
SSRN
Emerging technology is driving optimism about building a more inclusive digital economy. Will one innovation â€" a digital asset known as stable-coins â€" expand access to financial services among the poor? To encourage further study and dialogue, this note explores five open questions about whether stable-coins could promote financial inclusion in lower- and middle-income countries. The questions cover two practical issues concerning processing speed and technology available to the poor; costs to users; regulatory issues, especially compliance with existing customer funds protection rules; and implications for financial systems with limited foreign exchange reserves.The note identifies relevant considerations regarding the gender gap, "global stable-coins," and central bank digital currency. Dialogue between innovators, regulators, and consumer advocates will be crucial in determining whether stable-coins or other "crypto-assets" could help to build a more inclusive digital financial system and an economy that serves the needs of all, including the poor.

Detecting Changes in Asset Co-Movement Using the Autoencoder Reconstruction Ratio
Bryan Lim,Stefan Zohren,Stephen Roberts
arXiv

Detecting changes in asset co-movements is of much importance to financial practitioners, with numerous risk management benefits arising from the timely detection of breakdowns in historical correlations. In this article, we propose a real-time indicator to detect temporary increases in asset co-movements, the Autoencoder Reconstruction Ratio, which measures how well a basket of asset returns can be modelled using a lower-dimensional set of latent variables. The ARR uses a deep sparse denoising autoencoder to perform the dimensionality reduction on the returns vector, which replaces the PCA approach of the standard Absorption Ratio, and provides a better model for non-Gaussian returns. Through a systemic risk application on forecasting on the CRSP US Total Market Index, we show that lower ARR values coincide with higher volatility and larger drawdowns, indicating that increased asset co-movement does correspond with periods of market weakness. We also demonstrate that short-term (i.e. 5-min and 1-hour) predictors for realised volatility and market crashes can be improved by including additional ARR inputs.

Enhancing Time Series Momentum Strategies Using Deep Neural Networks
Bryan Lim,Stefan Zohren,Stephen Roberts
arXiv

While time series momentum is a well-studied phenomenon in finance, common strategies require the explicit definition of both a trend estimator and a position sizing rule. In this paper, we introduce Deep Momentum Networks -- a hybrid approach which injects deep learning based trading rules into the volatility scaling framework of time series momentum. The model also simultaneously learns both trend estimation and position sizing in a data-driven manner, with networks directly trained by optimising the Sharpe ratio of the signal. Backtesting on a portfolio of 88 continuous futures contracts, we demonstrate that the Sharpe-optimised LSTM improved traditional methods by more than two times in the absence of transactions costs, and continue outperforming when considering transaction costs up to 2-3 basis points. To account for more illiquid assets, we also propose a turnover regularisation term which trains the network to factor in costs at run-time.

Expected Market Returns in the International Cross-Section
Berkman, Henk,Malloch, Hamish
SSRN
Two recent articles, Martin (2017) and Chabi-Yo and Loudis (2019), derive a lower bound for the expected market risk premium that does not require parameter estimation and can be computed in real time. Based on evidence from 15 international markets, we cannot reject the hypothesis that these expected return lower bounds are tight. Furthermore, asset pricing factors that are insignificant in asset pricing tests if realized returns are used as expected return proxy become significant when expected market risk premium lower bounds are used instead; for example, the lower bound is higher for countries with greater exposure to global and regional stock markets and for countries with high exposure to the dollar and carry factors. Finally, we find that a strategy that takes a long position in markets with relatively high expected returns and a short position in markets with relatively low expected returns, yields statistically significant positive returns that are not explained by traditional risk factors.

Fifty Shades of QE: Conflicts of Interest in Economic Research
Fabo, Brian,Jancokova, Martina,Kempf, Elisabeth,Pastor, Lubos
SSRN
Central banks sometimes evaluate their own policies. To assess the inherent conflict of interest, we compare the research findings of central bank researchers and academic economists regarding the macroeconomic effects of quantitative easing (QE). We find that central bank papers report larger effects of QE on output and inflation. Central bankers are also more likely to report significant effects of QE on output and to use more positive language in the abstract. Central bankers who report larger QE effects on output experience more favorable career outcomes. A survey of central banks reveals substantial involvement of bank management in research production.

High-Resolution Poverty Maps in Sub-Saharan Africa
Kamwoo Lee,Jeanine Braithwaite
arXiv

Up-to-date poverty maps are an important tool for policy makers, but until now, have been prohibitively expensive to produce. We propose a generalizable prediction methodology to produce poverty maps at the village level using geospatial data and machine learning algorithms. We tested the proposed method for 25 Sub-Saharan African countries and validated them against survey data. The proposed method can increase the validity of both single country and cross-country estimations leading to higher precision in poverty maps of 44 Sub-Saharan African countries than previously available. More importantly, our cross-country estimation enables the creation of poverty maps when it is not practical or cost-effective to field new national household surveys, as is the case with many low- and middle-income countries.

Integrated Reporting, Disclosure Processing Costs Frictions and Capital Market Effects
Ng, Anthony C.,Low, Steven Yik-Pui,Gul, Ferdinand A
SSRN
Barth et al. (2017) (BCCV) show that firms with higher levels of compliance with Integrated Reporting () principles are associated with higher levels of firm value. However, according to disclosure processing costs theory, the positive association may not hold for high levels of disclosures. We use observations from large Australian firms to show that at low to moderate levels of there is a positive relationship between scores and (1) firm values, and (2) liquidity (a proxy for the capital market channel), but at progressively higher levels, the relationships become negative mainly because investors face frictions induced by disclosure processing costs. In additional tests, we show a similar non-linear relationship between and earnings response coefficients (ERCs), and and earnings management. The higher earnings management for high is consistent with the idea that managers may use high as an obfuscation tool.

Integration Among US Banks
Anand, Abhinav,Cotter, John
SSRN
We define and measure integration among a sample of 357 US banks over 25 years from 1993 to 2017 and show that the median US bankâ€™s integration has increased significantly post-2005. During the great recession and the Eurozone crisis, integration levels among US banks display a significant rise over and above their trend. We find that bank size is the most economically and statistically significant characteristic in explaining integration levels. Size and the equity ratio show positive association with bank integration while the net interest margin and combined tier 1 and tier 2 capital ratio influence bank integration negatively. For regulators, abnormally high in- tegration levels indicate warning signs of potential distress in the banking sector.

Intermediary Asset Pricing During the National Banking Era
Weiss, Colin
SSRN
Financial intermediary balance sheets matter for asset returns even when these intermediaries do not directly participate in the relevant asset markets. During the National Banking Era, liquidity conditions for the New York Clearinghouse (NYCH) banks forecast excess returns for stocks, bonds, and currencies. The NYCH banks had little to no direct participation in these markets; their main link to these markets was through securities financing. Liquidity conditions affect asset prices through the credit growth of the NYCH banks, which shapes marginal investors' discount rates. I use institutional features of this era to provide evidence in favor of this mechanism.

Internet Appendix for 'The Oligopoly Lucas Tree: Consumption Risk and Industry-Level Risk Exposure'
Dou, Winston,Ji, Yan,Wu, Wei
SSRN
This is the supplemental material to the paper titled "The Oligopoly Lucas Tree: Consumption Risk and Industry-Level Risk Exposure." It includes additional empirical, theoretical, and quantitative results. It also includes illustration for the numerical algorithm for our model solution.

Investor Demands for Safety, Bank Capital, and Liquidity Measurement
Passmore, S. Wayne,Temesvary, Judit
SSRN
We construct a model of a bank's optimal funding choice, where the bank negotiates with both safety-driven short-term bondholders and (mostly) risk-taking long-term bondholders. We establish that investor demands for safety create a negative relationship between the bank's capital choices and short-term funding, as well as negative relationships between capital and common measures of bank liquidity. Consistent with our model, our bank-level empirical analysis of these capital-liquidity tradeoffs show (1) that bank liquidity measures have a strong and negative relationship to its capital ratio for both large and small banks, and (2) that this relationship has weakened with the advent of stronger liquidity regulation. Our results suggest that the safety concerns of bank debt investors may underlie capital-liquidity tradeoffs and that a bank's share of collateralized short-term debt may be a more robust measure of bank liquidity.

Investor Sentiment and the (Discretionary) Accrual-Return Relation
Jiang, Jiajun,Liu, Qi,Sun, Bo
SSRN
Discretionary accruals are positively associated with stock returns at the aggregate level but negatively so in the cross section. Using Baker-Wurgler investor sentiment index, we find that a significant presence of sentiment-driven investors is important in accounting for both patterns. We document that the aggregate relation is only prominent during periods of high investor sentiment. Similarly, the cross-section relation is considerably stronger in high-sentiment periods in both economic magnitude and statistical significance. We then embed investor sentiment into a stylized model of earnings management, and illustrate that a positive (negative) relationship between stock returns and earnings management can endogenously emerge in the aggregate (cross section). Our analysis suggests that the (discretionary) accrual-return relation at both the aggregate and firm levels at least partially reflects mispricing that is related to market-wide investor sentiment.

Late to Recessions: Stocks and the Business Cycle
Gomez Cram, Roberto
SSRN
I find that returns are predictably negative for several months after the onset of recessions, and only become high thereafter. I identify business-cycle turning points by estimating a state-space model using macroeconomic data. Conditioning on the business cycle further reveals that returns exhibit momentum in recessions, whereas in expansions they display the mild reversals expected from discount rate changes. A market timing strategy that optimally exploits this business-cycle pattern produces a 60% increase in the buy-and-hold Sharpe ratio. I find that a subset of hedge funds add value for their clients in part by avoiding stock market crashes during recessions.

Learning in a Small/Big World
Benson Tsz Kin Leung
arXiv

Savage (1972) lays down the foundation of Bayesian decision theory, but asserts that it is not applicable in big worlds where the environment is complex. Using the theory of finite automaton to model belief formation, this paper studies the characteristics of optimal learning behavior in small and big worlds, where the complexity of the environment is low and high, respectively, relative to the cognitive ability of the decision maker. Confirming Savage's claim, optimal learning behavior is closed to Bayesian in small worlds but significantly different in big worlds. In addition, I show that in big worlds, the optimal learning behavior could exhibit a wide range of well-documented non-Bayesian learning behavior, including the use of heuristic, correlation neglect, persistent over-confidence, inattentive learning, and other behaviors of model simplification or misspecification. These results establish a clear and testable relationship between the prominence of non-Bayesian learning behavior, complexity and cognitive ability.

Liquidations: DeFi on a Knife-edge
Daniel Perez,Sam M. Werner,Jiahua Xu,Benjamin Livshits
arXiv

The trustless nature of permissionless blockchains renders overcollateralization a key safety component relied upon by decentralized finance (DeFi) protocols. Nonetheless, factors such as price volatility may undermine this mechanism. In order to protect protocols from suffering losses, undercollateralized positions can be \textit{liquidated}. In this paper, we present the first in-depth empirical analysis of liquidations on protocols for loanable funds (PLFs). We examine Compound, one of the most widely used PLFs, for a period starting from its conception to September 2020. We analyze participants' behavior and risk-appetite in particular, to elucidate recent developments in the dynamics of the protocol. Furthermore, we assess how this has changed with a modification in Compound's incentive structure and show that variations of only 3% in an asset's price can result in over 10m USD becoming liquidable. To further understand the implications of this, we investigate the efficiency of liquidators. We find that liquidators' efficiency has improved significantly over time, with currently over 70% of liquidable positions being immediately liquidated. Lastly, we provide a discussion on how a false sense of security fostered by a misconception of the stability of non-custodial stablecoins, increases the overall liquidation risk faced by Compound participants.

Media Coverage and Debt Financing
Cheng, C.S. Agnes,Jiang, Liangliang,Song, Wei-Ling
SSRN
We examine whether media coverage influences firmsâ€™ debt financing and, more importantly, the channels through which it occurs. We find that more media coverage significantly reduces firmsâ€™ reliance on bank loans while increasing their nonbank debt financing. This effect is concentrated among firms with more information available from other sources, suggesting that the media complements other information sources, and among firms with a higher demand for external monitoring, suggesting that the media substitutes for other governance mechanisms. News sentiment predicts subsequent credit ratings and firm value changes and affects access to new debt financing. Our findings validate the mediaâ€™s role as one of providing supplemental information: it supplies additional information content after controlling for other information sources. We also find that, as negative news coverage increases, firms opt for the private placement of debt over pubic bonds and bank debt and report significantly negative discretionary accruals. These additional findings lend further support to the media having a governance effect.

Modeling and analysis of the effect of COVID-19 on the stock price: V and L-shape recovery
Ajit Mahata,Anish rai,Om Prakash,Md Nurujjaman
arXiv

The emergence of the COVID-19 pandemic, a new and novel risk factors, leads to the stock price crash due to a rapid and synchronous sell-off by the investors. However, within a short period, the quality sectors start recovering from the bottom. A model of the stock price movement has been developed to explain such phenomena based on the Institutional fund flow and financial antifragility, which represents the financial indicator of a company. The assumes that during the crash, the stock does not depend on the financial antifragility of a company. We study the effects of shock lengths and antifragility parameter on the stock price during the crises period using the synthetic and real fund flow data. We observed that the possibility of recovery of a quality stock decreases with an increase in shock-length beyond a specific period. On the other hand, a quality stock with higher antifragility shows V-shape recovery and outperform others. The shock lengths and recovery length of quality stock are almost equal that is seen in the Indian market. Financially stressed stocks, i.e., the stock with negative antifragility, showed L-shape recovery during the pandemic. The results show that the investors, in the uncertainty like COVID-19, restructure their portfolio to de-risk the investment towards quality stocks. The study may help the investors to make the right investment decision during a crisis.

Mortgage Loss Severities: What Keeps Them so High?
An, Xudong,Cordell, Larry
SSRN
Mortgage loss-given-default (LGD) increased significantly when house prices plummeted during the financial crisis, but it has remained over 40 percent in recent years, despite a strong housing recovery. Our results indicate that the sustained high LGDs post-crisis is due to a combination of an overhang of crisis-era foreclosures and prolonged liquidation timelines, which have offset higher sales recoveries. Simulations show that cutting foreclosure timelines by one year would cause LGD to decrease by 5 to 8 percentage points, depending on the tradeoff between lower liquidation expenses and lower sales recoveries. Using difference-in-differences tests, we also find that recent consumer protection programs have extended foreclosure timelines and increased loss severities despite their potential benefits of increasing loan modifications and enhancing consumer protections.

On Track for Retirement?
Matthew Olckers
arXiv

Over sixty percent of employees at a large South African company contribute the minimum rate of 7.5 percent to a retirement fund, far below the rate of 15 percent recommended by financial advisers. I use a field experiment to investigate whether providing employees with a retirement calculator, which shows projections of retirement income, leads to increases in contributions. The impact is negligible. The lack of response to the calculator suggests many employees may wish to save less than the minimum. I use a model of asymmetric information to explain why the employer sets a binding minimum.

On the Continuity of the Feasible Set Mapping in Optimal Transport
Mario Ghossoub,David Saunders
arXiv

Consider the set of probability measures with given marginal distributions on the product of two complete, separable metric spaces, seen as a correspondence when the marginal distributions vary. In problems of optimal transport, continuity of this correspondence from marginal to joint distributions is often desired, in light of Berge's Maximum Theorem, to establish continuity of the value function in the marginal distributions, as well as stability of the set of optimal transport plans. Bergin (1999) established the continuity of this correspondence, and in this note, we present a novel and considerably shorter proof of this important result. We then examine an application to an assignment game (transferable utility matching problem) with unknown type distributions.

On the Importance of Household versus Firm Credit Frictions in the Great Recession
Kehoe, Patrick J.,Lopez, Pierlauro,Midrigan, Virgiliu,Pastorino, Elena
SSRN
Although a credit tightening is commonly recognized as a key determinant of the Great Recession, to date, it is unclear whether a worsening of credit conditions faced by households or by firms was most responsible for the downturn. Some studies have suggested that the household-side credit channel is quantitatively the most important one. Many others contend that the firm-side channel played a crucial role. We propose a model in which both channels are present and explicitly formalized. Our analysis indicates that the household-side credit channel is quantitatively more relevant than the firm-side credit channel. We then evaluate the relative benefits of a fixed-sized transfer to households and to firms that improves each groupâ€™s access to credit. We find that the effects of such a transfer on employment are substantially larger when the transfer targets households rather than firms. Hence, we provide theoretical and quantitative support to the view that the employment decline during the Great Recession would have been less severe if instead of focusing on easing firmsâ€™ access to credit, the government had expended an equal amount of resources to alleviate householdsâ€™ credit constraints.

Out of Sight No More? The Effect of Fee Disclosures on 401(K) Investment Allocations
Kronlund, Mathias,Pool, Veronika K.,Sialm, Clemens ,Stefanescu, Irina
SSRN
We examine the effects of a 2012 regulatory reform that mandated fee and performance disclosures for the investment options in 401(k) plans. We show that participants became significantly more attentive to expense ratios and short-term performance after the reform. The disclosure effects are stronger among plans with large average contributions per participant and weaker for plans with many investment options. Additionally, these results are not driven by secular changes in investor behavior or sponsor-initiated changes to the investment menus. Our findings suggest that providing salient fee and performance information can mitigate participants' inertia in retirement plans.

Policy Uncertainty and Textual Disclosure
Jiang, Liangliang,Pittman, Jeffrey,Saffar, Walid
SSRN
We analyze the importance of policy uncertainty to textual disclosure in the U.S. over the 1996 to 2015 period. Consistent with the information-acceleration view, we find that policy uncertainty increases textual disclosure quantity evident in disclosure length. We also document that policy uncertainty lowers textual readability, and increases the tone of uncertainty and negativity, suggesting that firms do not strategically disclose countervailing information to mislead the market. We also provide evidence implying that such firms enjoy cheaper equity financing costs afterward. Additionally, we find that the length effect is exacerbated by post-SOX filing, while the readability effect is mitigated by tough external monitoring evident in the presence of Big 4 auditors. Finally, tone disclosure becomes more uncertain and negative if firms have high government policy sensitivity amidst policy uncertainty.

Predictors of Social Distancing and Mask-Wearing Behavior: Panel Survey in Seven U.S. States
Plamen Nikolov,Andreas Pape,Ozlem Tonguc,Charlotte Williams
arXiv

This paper presents preliminary summary results from a longitudinal study of participants in seven U.S. states during the COVID-19 pandemic. In addition to standard socio-economic characteristics, we collect data on various economic preference parameters: time, risk, and social preferences, and risk perception biases. We pay special attention to predictors that are both important drivers of social distancing and are potentially malleable and susceptible to policy levers. We note three important findings: (1) demographic characteristics exert the largest influence on social distancing measures and mask-wearing, (2) we show that individual risk perception and cognitive biases exert a critical role in influencing the decision to adopt social distancing measures, (3) we identify important demographic groups that are most susceptible to changing their social distancing behaviors. These findings can help inform the design of policy interventions regarding targeting specific demographic groups, which can help reduce the transmission speed of the COVID-19 virus.

Price Negotiation with Merchant Heterogeneity in the Payment Card Industry
Ho, Chun-Yu,Xu, Li,Zhang, Daiqiang
SSRN
We examine price negotiation in the payment card industry by exploiting a unique merchant-, industry-, and city-level dataset. Motivated by the substantial variation in acquirer fees paid by merchants to the acquirer and the heterogeneous merchant card transactions, we use Nash bargaining to model the negotiation over the acquirer fee between an acquirer and a merchant.The empirical results show that the average bargaining power is weaker for the acquirer than for the merchants, which results in a larger incremental surplus secured by the merchants than the surplus secured by the acquirer on average. Our counterfactuals show that merchants might face upward pressure on acquirer fees as the card penetration rate rises over time and that policies that weaken the acquirer's bargaining power could relieve the upward fee pressure.

Risk Management Committee and Bank Performance: Evidence from the Adoption of Dodd-Frank Act
Jiang, Liangliang
SSRN
This paper tests the effect of the establishment of risk management committee on bank risk, bank loan performance and bank value. The Dodd Frank Act of 2010 provides us with quasi-experimental variation on risk management committee establishment that facilitates identification. I present two identification methods: (1) identifying the risk management committee effect using an instrumental variable that is based on the difference-in-differences; and (2) testing the risk management committee effect using the $10 billion assets as a cutoff and employing the fuzzy regression discontinuity design. I find that the establishment of risk committee has effectively reduced bank risks, including total risk, tail risk, residual risk, and asset risk. The risk committee is also beneficial to firm value increment and non-performing loan reduction. In addition, I find that the risk reduction effect from the risk management committee is more pronounced among asset diversified banks. Strategic Forbearance Applications by Borrowers & Selective Verification of Unemployment Status by Servicers: Evidence from CARES Act during COVID-19 Bandyopadhyay, Arka SSRN I utilize a novel data on proprietary servicer comments to investigate strategic borrower responses to the mortgage forbearance program contained in the Coronavirus Aid, Relief, and Economic Security Act. The unique text data allows me to corroborate the selective verification of unemployment status (financial hardship) by the servicer. I also discern unintended distributional implications for African American and Hispanic borrowers with performing loans via this selective verification with respect to ex-post risk, although the servicer does not have the race information about the borrowers. The soft information obtained from servicer call transcripts helps me identify the reasons for these communications and the incentive compatibility between the borrower and the servicer. My finding sheds light on the policy failure on Government programs, like FHA, VA, USDA, etc., during exacerbated income shocks, such as, COVID-19. Teacher turnover in Rwanda Andrew Zeitlin arXiv Despite widely documented shortfalls of teacher skills and effort, there is little systematic evidence of rates of teacher turnover in low-income countries. I investigate the incidence and consequences of teacher turnover in Rwandan public primary schools over the period from 2016-2019. To do so, I combine the universe of teacher placement records with student enrollment figures and school-average Primary Leaving Exam scores in a nationally representative sample of 259 schools. Results highlight five features of teacher turnover. First, rates of teacher turnover are high: annually, 20 percent of teachers separate from their jobs, of which 11 percent exit from the public-sector teaching workforce. Second, the burden of teacher churn is higher in schools with low learning levels and, perhaps surprisingly, in low pupil-teacher-ratio schools. Third, teacher turnover is concentrated among early-career teachers, male teachers, and those assigned to teach Math. Fourth, replacing teachers quickly after they exit is a challenge; 23 percent of exiting teachers are not replaced the following year. And fifth, teacher turnover is associated with subsequent declines in learning outcomes. On average, the loss of a teacher is associated with a reduction in learning levels of 0.05 standard deviations. In addition to class-size increases, a possible mechanism for these learning outcomes is the prevalence of teachers teaching outside of their areas of subject expertise: in any given year, at least 21 percent of teachers teach in subjects in which they have not been trained. Taken together, these results suggest that the problem of teacher turnover is substantial in magnitude and consequential for learning outcomes in schools. The Global Financial Centres Index 24 Yeandle, Mark,Wardle, Mike,Mainelli, Michael SSRN September 2018 sees publication of the twenty-fourth edition of the Global Financial Centres Index (GFCI 24). GFCI 24 provides evaluations of future competitiveness and rankings for 100 major financial centres around the world. The GFCI serves as a valuable reference for policy and investment decision-makers.China Development Institute (CDI) in Shenzhen and Z/Yen Partners in London collaborate in producing the GFCI. The GFCI is updated and published every March and September, and receives considerable attention from the global financial community.110 financial centres were researched for GFCI 24 of which 100 are now in the main index. The GFCI is compiled using 137 instrumental factors. These quantitative measures are provided by third parties including the World Bank, the Economist Intelligence Unit, the OECD and the United Nations.The instrumental factors are combined with financial centre assessments provided by respondents to the GFCI online questionnaire (www.globalfinancialcentres.net). GFCI 24 uses 31,326 assessments from 2,453 respondents.The Headlines Of GFCI 24 Include:Not for the first time, New York took first place in the index, just two points head of London, although both centres fell slightly in the ratings. Hong Kong is now only three points (on a scale of 1,000) behind London. Shanghai overtook Tokyo to move into fifth place, gaining 25 points in the ratings. Beijing, Zurich, and Frankfurt moved into the top ten centres, replacing Toronto, Boston, and San Francisco.In Western Europe, Zurich, Frankfurt, Amsterdam, Vienna, and Milan moved up the rankings significantly. These centres may be the main beneficiaries of the uncertainty caused by Brexit. Surprisingly, despite some evident success in attracting new business, Dublin, Munich, Hamburg, Copenhagen, and Stockholm fell in the rankings, reflecting respondentsâ€™ views of their future prospects.In Asia/Pacific, the leading centres performed well, closing the gap on London and New York at the top of the rankings. There were steady increases for Shanghai, Sydney, Beijing, and Guangzhou and GIFT City (Gujarat) and Hangzhou entered the index for the first time.The leading North American centres fell back in the rankings and ratings overall, although Los Angeles and Washington DC gained places.In Eastern Europe and Central Asia, there were significant gains for Astana, Budapest, St Petersburg, and Tallinn. Astana only officially launched their financial centre in July, and it is unusual for such a new centre to perform so strongly.In the Middle East and Africa, Dubai, Abu Dhabi, and Doha all rose significantly reversing the trend from GFCI 23. Cape Town is the highest new entrant to the index, ranking 38th in its first entry.In Latin America and the Caribbean there were mixed results. Bermuda, Sao Paulo, Mexico City, and Rio de Janeiro performed strongly, while other centres fell in the rankings.Island centres fell in the index, with the exception of Bermuda, which rose six places. The British Crown dependencies of Jersey, Guernsey, and the Isle of Man all fell significantly in the rankings, with the Isle of Man dropping 27 places. The Impact of Automated Information Acquisition on the Stock Market JÃ¤ger, Ivika SSRN Does collecting regulatory financial information with the help of automated computer algorithms (robots) affect the stock market? Using the EDGAR Server Log data set, I construct firm-level measures of information acquisition by robots and non-robots and show that robots are extensively used for information acquisition when new information becomes available. The SEC's mandateregarding interactive data leads to a notable increase in information demand, consistent with decreased information acquisition costs for standardised regulatory financial information in XBRL-format. A higher relative importance of robots acquiring information about a firm combined with the XBRL adoption is associated with a consequent increase in trading volume, smaller bid-ask spreads, lower volatility, positive cumulative abnormal return and increased volume coefficient of variation. The findings are consistent with the idea that automation and standardisation benefits informed investors disproportionately more than uninformed traders. The Importance of Deposit Insurance Credibility Bonfim, Diana,Santos, JoÃ£o A. C. SSRN The success of deposit insurance arrangements at eliminating bank runs is likely closely tied to their credibility. We investigate this hypothesis building on two episodes which tested the insurance protection offered by the Portuguese arrangement in the midst of the country's sovereign debt crisis. Our results show that Portuguese depositors responded to foreign banks' decision to convert their subsidiaries into branches by relocating their deposits into the latter. We find a similar response following the announcement that insured depositors in Cyprus would lose part of their savings. On both instances responses are concentrated on household deposits. Given that foreign banks' branches offer the insurance protection of these banks' home countries, rather than that granted by their host country arrangement, our findings confirm that the credibility of the deposit insurance arrangement is critical for the protection it offers banks against the risk of depositor runs. These results show that sovereign-bank links can be detrimental to financial stability through a novel channel: the credibility of deposit insurance. The Relevance of Public Law to Private Ordering: The Consequences of Uncertain Judicial Review for Stock Exchange Self-Regulation Chan, Jonathan SSRN Self-regulation relies on private ordering, whereby private actors make and enforce rules governing their conduct. Private ordering is not outside the reach of public law principles, making the certainty of private ordering in financial regulation and other contexts dependent on the predictability of whether public law principles apply. This article conducts a detailed study of the London Stock Exchangeâ€™s self-regulation of AIM (Alternative Investment Market) to illustrate the relevance of public law to private ordering. This article argues that it is difficult to ascertain whether public power is being exercised in seemingly private self-regulatory arrangements due to doctrinal uncertainty in the tests for publicness under English law. It suggests that regulation on AIM does not likely have a sufficiently public element to be amenable to judicial review. This position is desirable because it increases the certainty of private ordering in UK financial regulation, instead of leaving market participants with differing expectations of their procedural fairness rights. AIM demonstrates how unpredictable public law leads to uncertain private ordering, and how uncertainty over amenability to judicial review undermines the very logic of self-regulation by hindering informed ex ante bargaining and contracting. The Smart Centres Index 1 Wardle, Mike,Mainelli, Michael SSRN The first edition of the Smart Centres Index (SCI) was published on 14 July 2020. SCI 1 rates the innovation and technology offerings of leading commercial and financial centres across the world.The SCI focuses on centres in relation to their approach to and delivery of innovation and technology, including Science, Energy Systems, Machine Learning, Distributed Ledgers, and Fintech, along with other applications. We look at cities rather than countries in developing the index as we consider that it is in cities and other commercial clusters that the development of business is driven forward.The Smart Centres Index is based on evaluations of three dimensions:Innovation support - the approach taken to regulation and support for the innovation and technology industry provided by the commercial ecosystem.Creative Intensity - the extent to which technology and innovative industries are embedded in the economy of the centre.Delivery Capability - the quality of the work being undertaken in the field in the centre. future competitiveness and rankings for financial and commercial centres around the world.128 commercial and financial centres were researched for SCI 1 of which 48 are included in the index. SCI 1 was compiled using 127 instrumental factors. These quantitative measures are provided by third parties including the World Bank, The Economist Intelligence Unit, the OECD, and the United Nations.The instrumental factors are combined with financial centre assessments provided by respondents to the GFCI online questionnaire. SCI 1 uses 965 assessments from 92 respondents.Index Results - London takes first place in the index, with new York second, and Singapore in third place.Five of the top ten places in the ranking are taken by US centres.The leading centres are strong across all three of the SCI dimensions.Chinese centres do not feature as strongly as we might have expected, and score on average lower for Innovation Support than their overall rating.The great majority of centres featured in SCI 1 are located in North America, Asia/Pacific, and Western Europe.North America - Ten North American centres feature in the SCI, dominated by US centres, who hold five of the top ten places globally.Along with ranking second overall, New York also ranks second in each of the three dimensions which make up the SCI.Vancouver scores high in creative intensity, ranking 9th against its overall SCI ranking of 17th.Asia/Pacific - Three of the 13 Asia/Pacific centres in the index - Singapore, Hong Kong, and Tokyo - feature in the world top ten.The majority of Asia/Pacific centres scored lower for innovation support, including regulation, than their overall ranking in the SCI.Chinese centres such as Shenzhen, which have strong technology ecosystems, do not feature as highly in the index as we might have expected. This may be because those commenting on Chinese centres know more Hong Kong, Beijing and Shanghai better than other centres.Western Europe - Twenty centres in Western Europe feature in the index, with London and Zurich in the global top ten.The majority of Western European centres score higher for innovation support, including regulation, than their overall rank. This suggests that systems of public support for, and regulation of innovation and technology are a strength in the region.Stockholm scores significantly higher than its overall rating in creative intensity, while Oxford and Cambridge in the UK score significantly higher for delivery capability.Other Regions - Only five of the centres in SCI 1 are from other regions of the world - Middle East & Africa, Eastern Europe & Central Asia, and Latin America & The Caribbean.Of these centres, Dubai is rated highest, ranked at 34th in the world. The Stock Market Response to a &Quot;Regulatory Sine Curve&Quot; Sun, Bo,Tam, Xuan S.,Young, Eric R. SSRN We construct new indicators of financial regulatory intensity and find evidence that a "regulatory sine curve" generally exists: regulatory oversight increases following a recession and wanes as the economy returns to normalcy. We then build an asset pricing model, based on the idea that regulatory oversight both deters incentives to commit fraud ex ante and reveals hidden negative information ex post. Our calibration suggests that these mechanisms can be quantitatively important for stock price dynamics. The value of power-related options under spectrally negative L\'evy processes Jean-Philippe Aguilar arXiv We provide analytical tools for pricing power options with exotic features (capped or log payoffs, gap options ...) in the framework of exponential L\'evy models driven by one-sided stable or tempered stable processes. Pricing formulas take the form of fast converging series of powers of the log-forward moneyness and of the time-to-maturity; these series are obtained via a factorized integral representation in the Mellin space evaluated by means of residues in$\mathbb{C}$or$\mathbb{C}^2$. Comparisons with numerical methods and efficiency tests are also discussed. Towards a Unified Approach to Economic Assessment in International Commercial Law Reform Varzaly, Jenifer SSRN Economic assessment is critical to international commercial law reform in order to ensure law reform efforts are evidence-based and result in economic benefit. Yet, such assessment is seldom undertaken and, when done so, is not uniform in its scope or methodology, nor is it subject to systematic review. This article proposes a set of guidelines in order to address this challenge. While there are both political and technical challenges to law reform efforts and the implementation of a clear framework for assessment, this article aims to provide a first step forward towards creating a unified approach. It does so by proposing a series of guidelines which traverse five key components of any economic assessment undertaking, drawing on international and within country data regarding assessment processes. There is a particular focus on quantitative analysis, replicability, transparency, and continuous improvement, in relation to unifying private international law within the commercial sphere. Trade Credit, Markups, and Relationships Garcia-Marin, Alvaro,Justel, Santiago,Schmidt-Eisenlohr, Tim SSRN Trade credit is the most important form of short-term finance for firms. In 2019, U.S. non-financial firms had about$4.5 trillion in trade credit outstanding equaling 21 percent of U.S. GDP. This paper documents two striking facts about trade credit use. First, firms with higher markups supply more trade credit. Second, trade credit use increases in relationship length, as firms often switch from cash in advance to trade credit but rarely away from trade credit. These two facts can be rationalized in a model where firms learn about their trading partners, sellers charge markups over production costs, and financial intermediation is costly. The model also shows that saving on financial intermediation costs provides a strong rationale for the dominance of trade credit. Using Chilean data at the firm-product-level and the trade-transaction level, we find support for all predictions of the model.

Trading from Home: The Impact of COVID-19 on Trading Volume around the World
Chiah, Mardy,Zhong, Angel
SSRN
This paper examines the impact of COVID-19 on trading volume in stock markets around the world. We document a large spike in trading volume in 37 international equity markets. The surge in trading volume is found to be associated with the national culture and institutional environment of individual countries. In particular, investors tend to trade more heavily in societies characterized by a higher level of trust and individualism, as well as a lower level of uncertainty avoidance. Investors are also more willing to trade in wealthier nations, as well as those with stronger protection of legal rights, better governance systems, and greater gambling opportunities.

Transparency, Auditability and eXplainability of Machine Learning Models in Credit Scoring
Michael Bücker,Gero Szepannek,Alicja Gosiewska,Przemyslaw Biecek
arXiv

A major requirement for credit scoring models is to provide a maximally accurate risk prediction. Additionally, regulators demand these models to be transparent and auditable. Thus, in credit scoring, very simple predictive models such as logistic regression or decision trees are still widely used and the superior predictive power of modern machine learning algorithms cannot be fully leveraged. Significant potential is therefore missed, leading to higher reserves or more credit defaults. This paper works out different dimensions that have to be considered for making credit scoring models understandable and presents a framework for making black box'' machine learning models transparent, auditable and explainable. Following this framework, we present an overview of techniques, demonstrate how they can be applied in credit scoring and how results compare to the interpretability of score cards. A real world case study shows that a comparable degree of interpretability can be achieved while machine learning techniques keep their ability to improve predictive power.

Volatility Forecasting Accuracy for Bitcoin
KÃ¶chling, Gerrit,Schmidtke, Philipp,Posch, Peter N.
SSRN
We analyze the quality of Bitcoin volatility forecasting of GARCH-type models applying different volatility proxies and loss functions. We construct model confidence sets and find them to be systematically smaller for asymmetric loss functions and a jump robust proxy.

lCARE -- localizing Conditional AutoRegressive Expectiles
Xiu Xu,Andrija Mihoci,Wolfgang Karl Härdle
arXiv

We account for time-varying parameters in the conditional expectile-based value at risk (EVaR) model. The EVaR downside risk is more sensitive to the magnitude of portfolio losses compared to the quantile-based value at risk (QVaR). Rather than fitting the expectile models over ad-hoc fixed data windows, this study focuses on parameter instability of tail risk dynamics by utilising a local parametric approach. Our framework yields a data-driven optimal interval length at each time point by a sequential test. Empirical evidence at three stock markets from 2005-2016 shows that the selected lengths account for approximately 3-6 months of daily observations. This method performs favorable compared to the models with one-year fixed intervals, as well as quantile based candidates while employing a time invariant portfolio protection (TIPP) strategy for the DAX, FTSE 100 and S&P 500 portfolios. The tail risk measure implied by our model finally provides valuable insights for asset allocation and portfolio insurance.