Research articles for the 2019-11-20
arXiv
We develop an agent-based simulation of the catastrophe insurance and reinsurance industry and use it to study the problem of risk model homogeneity. The model simulates the balance sheets of insurance firms, who collect premiums from clients in return for ensuring them against intermittent, heavy-tailed risks. Firms manage their capital and pay dividends to their investors, and use either reinsurance contracts or cat bonds to hedge their tail risk. The model generates plausible time series of profits and losses and recovers stylized facts, such as the insurance cycle and the emergence of asymmetric, long tailed firm size distributions. We use the model to investigate the problem of risk model homogeneity. Under Solvency II, insurance companies are required to use only certified risk models. This has led to a situation in which only a few firms provide risk models, creating a systemic fragility to the errors in these models. We demonstrate that using too few models increases the risk of nonpayment and default while lowering profits for the industry as a whole. The presence of the reinsurance industry ameliorates the problem but does not remove it. Our results suggest that it would be valuable for regulators to incentivize model diversity. The framework we develop here provides a first step toward a simulation model of the insurance industry for testing policies and strategies for better capital management.
SSRN
The findings of this paper suggest another reason for capital structure adjustments besides the Trade-Off and Pecking Order theories propositions because asymmetric information impacts capital structure changes and deviations only for a quarter whilst stationarity impacts them for 4 quarters, even when controlled. Asymmetric information has been measured by Corwin Schultz bid-ask spread estimator and capital structure target as the mean of debt to equity ratio of 262 Nyse non-financial and non-regulated companies and their industries during 91 quarters. The data were analyzed with Johansen Fisher panel cointegration. The capital structure deviations last from 2 to 4 quarters and move toward a target
SSRN
Climate policies to keep global warming below 2â„Æ' might render some of the world’s fossil fuels and related infrastructure worthless prior to the end of their economic life time. Therefore, some energy-sector assets are at risk of becoming stranded. This paper investigates whether and how investors price in this risk of asset stranding. We exploit the gradual development of a German climate policy proposal aimed at reducing electricity production from coal and analyze its effect on the valuation of energy utilities. We find that investors take stranded asset risk into consideration, but that they also expect a financial compensation for their stranded assets.
arXiv
We introduce a constrained priority mechanism that combines outcome-based matching from machine-learning with preference-based allocation schemes common in market design. Using real-world data, we illustrate how our mechanism could be applied to the assignment of refugee families to host country locations, and kindergarteners to schools. Our mechanism allows a planner to first specify a threshold g for the minimum average outcome score that should be achieved by the assignment. In the refugee matching application, this score corresponds to the predicted probability of employment, while in the student assignment application it corresponds to standardized test scores. The mechanism is a priority mechanism that considers both outcomes and preferences by assigning agents (refugee families, students) based on their preferences, but subject to meeting the planner's specified threshold. The mechanism is both strategy-proof and constrained efficient in that it always generates a matching that is not Pareto dominated by any other matching that respects the planner's threshold.
arXiv
Cross-correlations in fluctuations of the daily exchange rates within the basket of the 100 highest-capitalization cryptocurrencies over the period October 1, 2015, through March 31, 2019, are studied. The corresponding dynamics predominantly involve one leading eigenvalue of the correlation matrix, while the others largely coincide with those of Wishart random matrices. However, the magnitude of the principal eigenvalue, and thus the degree of collectivity, strongly depends on which cryptocurrency is used as a base. It is largest when the base is the most peripheral cryptocurrency; when more significant ones are taken into consideration, its magnitude systematically decreases, nevertheless preserving a sizable gap with respect to the random bulk, which in turn indicates that the organization of correlations becomes more heterogeneous. This finding provides a criterion for recognizing which currencies or cryptocurrencies play a dominant role in the global crypto-market. The present study shows that over the period under consideration, the Bitcoin (BTC) predominates, hallmarking exchange rate dynamics at least as influential as the US dollar. The BTC started dominating around the year 2017, while further cryptocurrencies, like the Ethereum (ETH) and even Ripple (XRP), assumed similar trends. At the same time, the USD, an original value determinant for the cryptocurrency market, became increasingly disconnected, its related characteristics eventually approaching those of a fictitious currency. These results are strong indicators of incipient independence of the global cryptocurrency market, delineating a self-contained trade resembling the Forex.
arXiv
This paper sets forth a framework for deep reinforcement learning as applied to market making (DRLMM) for cryptocurrencies. Two advanced policy gradient-based algorithms were selected as agents to interact with an environment that represents the observation space through limit order book data, and order flow arrival statistics. Within the experiment, a forward-feed neural network is used as the function approximator and two reward functions are compared. The performance of each combination of agent and reward function is evaluated by daily and average trade returns. Using this DRLMM framework, this paper demonstrates the effectiveness of deep reinforcement learning in solving stochastic inventory control challenges market makers face.
SSRN
This study examines the effect of agglomeration and localized knowledge spillover on reward-based crowdfunding outcomes in the United States. By examining art projects specifically, we can measure how creative projects benefit from agglomeration or clustering, measured by the proportion of individuals employed in creative class jobs. We find that crowdfunding success among metro and non-metro counties is not homogenous, and that clustering among Creative Class or Bohemia Class workers leads to greater fundraising amounts. Our findings also suggest a mitigating effect of knowledge spillover, as projects originating from high natural amenity counties without high concentrations of Creative Class or Bohemia Class workers raise less money, supporting the Localized Knowledge Spillover Theory of Entrepreneurship.
SSRN
European banks have been criticized for holding excessive domestic government debt during the recent Eurozone crisis, which may have intensified the diabolic loop between sovereign and bank credit risks. By using a novel bank-level dataset covering the entire timeline of the Eurozone crisis, I first re-confirm that the crisis led to the reallocation of sovereign debt from foreign to domestic banks. In contrast to the recent literature focusing only on sovereign debt, I show that the banks’ private sector exposures were (at least) equally affected by the rise in home bias. Consistent with this pattern, I propose a new debt reallocation channel based on informational frictions and show that the informationally closer foreign banks increase their relative exposures when the sovereign risk rises. The effect of informational closeness is economically meaningful and robust to the use of different information measures and controls for alternative channels of sovereign debt reallocation.
arXiv
The efficiency of a modern economy depends on what we call the Value-Tracking Hypothesis: that market prices of key assets broadly track some underlying value. This can be expected if a sufficient weight of market participants are valuation-based traders, buying and selling an asset when its price is, respectively, below and above their well-informed private valuations. Such tracking will never be perfect, and we propose a natural unit of tracking error, the 'deciblack'. We then use a simple discrete-time model to show how large tracking errors can arise if enough market participants are not valuation-based traders, regardless of how much information the valuation-based traders have. We find a threshold above which value-tracking breaks down without any changes in the underlying value of the asset. Because financial markets are increasingly dominated by non-valuation-based traders, assessing how much valuation-based investing is required for reasonable value tracking is of urgent practical interest.
SSRN
Using news-based government economic policy uncertainty (EPU) index of Baker et al. (2016) and bank-level data from 17 countries over the period 1998-2012, we find that government economic policy uncertainty has significant positive association with interest rates on bank gross loans. Specifically, a one standard deviation increase in EPU leads to 21.84 basis points increase in average interest rates on bank gross loans. We conjecture the economic policy uncertainty boosts banks’ loan prices by increasing the borrowers’ default risk. The impact of EPU on banks’ loan pricing remains persistent after controlling for banks’ own idiosyncratic default risk and the political risk variables from ICRG database. Results remain robust when we use general elections as an alternative proxy of government economic policy uncertainty. We also confirm main results with syndicated loan deals data and observe a significant positive association between loan spreads and EPU index. Together, our results suggest that government economic policy uncertainty is an economically important risk factor for banks’ loan pricing.
SSRN
Correlation matrices are ubiquitous in finance. Some key applications include portfolio construction, risk management, and factor/style analysis. Correlation matrices are usually estimated from historical empirical observations or derived from historically estimated factors. It is widely acknowledged that empirical correlation matrices: (a) have poor numerical properties that lead to unreliable estimators; and (b) have poor predictive power. Additionally, factor-based correlation matrices have their own caveats. In particular, estimated factors are typically non-hierarchical and do not allow for interactions at different levels. This contravenes the fact that financial instruments typically exhibit a nested cluster structure (e.g., MSCI’s GICS levels 1-4).This paper introduces a machine learning (ML) algorithm to estimate forward-looking correlation matrices implied by economic theory. Given a particular theoretical representation of the hierarchical structure that governs a universe of securities, the method fits the correlation matrix that complies with that theoretical representation of the future. This particular use case demonstrates how, contrary to popular perception, ML solutions are not black-boxes, and can be applied effectively to develop and test economic theories.
SSRN
Factor investing has received much attention from academics and practitioners, as well as from individual and institutional investors. It has become usual for investors that aim to enhance returns to add to the core of their portfolios a factor satellite, thus tilting their portfolios toward factors that have produced a long-term risk premium. However, in most cases, investors behaving this way are not fully invested in stocks, which begs an interesting question: Should an investor with a two-asset portfolio of broadly-diversified stocks and bonds tilt the stocks slice of his portfolio toward (small-cap and value) factors, or would he be better off by simply increasing the allocation to broadly-diversified stocks in his two-asset portfolio? The results discussed here, based on different samples and sample periods, support the notion of factor-tilting portfolios.
SSRN
We examine the impact of sector-based reform on income inequality, concentrating on state banking deregulation in the US, for which we employ annual balanced panel data from all 50 states and the District of Columbia, covering the period from 1970 to 2000, for our baseline analysis. The estimation strategy exploits the variation across states and years in the enactment of laws that remove restrictions on in-state bank branch geographical expansion and cross-state bank business operational expansion to compute the effects of developments in the financial sector on income inequality. We find evidence that inequality on average decreases with within-state branching reform, whereas it on average increases with between-state banking deregulation. Utilising five different measures of inequality (top decile income share, Atkinson index, the Gini coefficient, relative mean deviation, and Theil entropy index), we determine that our finding materially depends on which measure of income inequality is being considered. We argue that this has not been stressed in the previous literature.
SSRN
The geopolitical risk found to be a relevant factor for investors to choose Islamic stock investment. This study focuses on the short-run and long-run asymmetric significance of geopolitical risk on Islamic stock market returns. Nonlinear Autoregressive Distributive Lag (NARDL) model has been used for the major Islamic countries i.e. Saudi Arabia, Malaysia, Turkey, and Indonesia. The short-run dynamics suggest the presence of short-run asymmetry in the case of Saudi Arabia and Indonesia. However, in the long-run, the positive and negative shock of geopolitical risk is having an asymmetric impact on the Islamic stock market return in the case of Indonesia. The Islamic stock market getting an asymmetric response to geopolitical shock can be because of multiple determinants related to the Islamic stock market and different environments of the Islamic stock market in each economy. The study recommends the Islamic investors and other participants to consider the asymmetric nature of Islamic stock market returns to minimize risk related to negative geopolitical events.
SSRN
We estimate the evolution of the conditional joint distribution of economic and financial conditions in the United States, documenting a novel empirical fact: while the joint distribution is approximately Gaussian during normal periods, sharp tightenings of financial conditions lead to the emergence of additional modes â€" that is, multiple economic equilibria. Although the U.S. economy has historically reverted quickly to a “good†equilibrium after a tightening of financial conditions, we conjecture that poor policy choices under these circumstances could also open a pathway to a “bad†equilibrium for a prolonged period. We argue that such multimodality arises naturally in a macro-financial intermediary model with occasionally binding intermediary constraints.
arXiv
We show a concise extension of the monotone stability approach to backward stochastic differential equations (BSDEs) that are jointly driven by a Brownian motion and a random measure for jumps, which could be of infinite activity with a non-deterministic and time inhomogeneous compensator. The BSDE generator function can be non convex and needs not to satisfy global Lipschitz conditions in the jump integrand. We contribute concrete criteria, that are easy to verify, for results on existence and uniqueness of bounded solutions to BSDEs with jumps, and on comparison and a-priori $L^{\infty}$-bounds. Several examples and counter examples are discussed to shed light on the scope and applicability of different assumptions, and we provide an overview of major applications in finance and optimal control.
SSRN
Financial analysts assume that the reliability of predictions derived from regression analysis improves with sample size. This is generally true because larger samples tend to produce less noisy results than smaller samples. But this is not always the case. Some observations are more relevant than others, and it is often the case that one can obtain more reliable predictions by censoring observations that are not sufficiently relevant. The authors introduce a methodology for identifying relevant observations by recasting the prediction of a regression equation as a weighted average of the historical values of the dependent variable in which the weights are the relevance of the independent variables. This equivalence allows them to use a subset of more relevant observations to forecast the dependent variable. The authors apply their methodology to forecast factor returns from economic variables.
SSRN
When acting as an intermediary, corporate venture capital (CVC) units must balance two different institutional settings: the rigid corporate world and the advancing startup ecosystem. As a result, CVC units are faced with multiple voids that influence their organizational orientation toward one environment. This study employs text analysis on a unique sample of 22 CVC dyads to introduce a novel empirical way of measuring isomorphic variation over time. Following a mixed-method approach, the quantitative results are used to shed light on potential drivers of isomorphism, compiled by semi-structured interviews. The findings demonstrate that the degree of isomorphism is not only determined by decisions made during the initial phase of a CVC unit, but also from mimetic processes that occur within the lifespan of such investment vehicles. The study thereby contributes to the ongoing academic discussion by elucidating potential drivers of isomorphism and provides researchers with a novel way to measure isomorphic tendencies based on organizational text excerpts.
SSRN
The Jumpstart Our Business Startups (JOBS) Act seeks to improve access to capital by deregulating both public and private equity markets in the US. We find that the Act incentivizes cash-starved firms to go public rather than raise capital privately. Proceeds raised are used to repay debt and pay executives, not to increase investments. Moreover, firms going public under the JOBS Act pay more to raise equity and often rely on further public issues to avert financial distress. These results suggest that the Act is mainly encouraging lower-quality firms to go public rather than stay private.
SSRN
This paper examines the benefits and costs associated with rookie independent directors (RIDs) in Chinese public companies from 2008 to 2014. We find that RIDs attend more board meetings. Boards with more RIDs tunnel less to controlling shareholders, suggesting that RIDs are efficient monitors. However, in state-owned firms, the presence of RIDs is negatively associated with investment efficiency, suggesting a potential cost of appointing RIDs. Overall, firms with more RIDs have higher operating performance, especially when tunneling is a more common issue, when board experience is less important and when monitoring costs are relatively low.
SSRN
The correlation between stock characteristics and the cross-section of stock returns plays a central role in empirical implementations of modern asset pricing models and, as such, has important implications for investment management. This remains true whether the correlation is due to investor preferences regarding the characteristics or whether the characteristics are proxies for state variables the risk of which investors are attempting to hedge. This paper asks what do we know about the relation between these characteristics and the cross-section of returns? The skeptic’s answer is not much. A combination of lack of persistence in the characteristics along with problems caused by model uncertainty, data snooping and nonstationarity means that our knowledge is sketchy at best. Investors should be forewarned.
SSRN
The extended Friday the 13th Effect is a calendar anomaly consisting in abnormal stock returns that occur in a time interval that starts some trading days before the supposed unlucky day of Friday the 13th and it ends some trading days after. This paper approaches the presence of such patterns in the evolution of the closed values of five indexes from the London Stock Exchange: FTSE 100, FTSE 250, FTSE 350, FTSE SmallCap and FTSE All-Share. This investigation is performed for two periods: the first from January 1998 to December 2006 and the second from January 2007 to July 2019. While the first one could be considered as relative quiet, the second one was more turbulent. In the case of first period the results revealed, for four indexes, that in the trading day that follows Friday the 13th the returns were significant higher than the average. Instead, in the case of second period, we found, for the same four indexes, that two trading days before the Friday the 13th the returns were significant lower than the average. We conclude that, like many other calendar anomalies, extended Friday the 13th Effect is not persistent in time.
SSRN
We examine the impact of economist forecasts on over-the-counter treasury bond trading and how participants with varied access to these forecasts respond. We find overnight interest rate forecasts are associated with increased treasury market trade volume and volatility, but not bond prices. Dealers at the forecasting economist’s institution sell but do not buy in response to forecasts that imply a fall and rise in bond price respectively. Dealers trade consistent with the forecasts of their own institution’s economists to reduce downside exposure but do not increase upside exposure in a uniform manner. Our results are consistent with expectations derived from theoretical models where agents heterogeneously interpret public information.
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.
SSRN
There is now an abundance of literature showing that the equal weighted portfolio outperforms the value weighted portfolio. However, this has led to claims that the act of rebalancing itself within an equal weight strategy is what leads to outperformance, whether or not individual security returns are mean-reverting. In this paper we show that you can achieve the same, if not better, returns than that of the equal weighted portfolio rebalanced monthly by rebalancing less than half of the time. This is achieved by considering only whether the diversification benefit is increasing or decreasing over some period of time.
arXiv
Increased data gathering capacity, together with the spread of data analytics techniques, has prompted an unprecedented concentration of information related to the individuals' preferences in the hands of a few gatekeepers. In the present paper, we show how platforms' performances still appear astonishing in relation to some unexplored data and networks properties, capable to enhance the platforms' capacity to implement steering practices by means of an increased ability to estimate individuals' preferences. To this end, we rely on network science whose analytical tools allow data representations capable of highlighting relationships between subjects and/or items, extracting a great amount of information. We therefore propose a measure called Network Information Patrimony, considering the amount of information available within the system and we look into how platforms could exploit data stemming from connected profiles within a network, with a view to obtaining competitive advantages. Our measure takes into account the quality of the connections among nodes as the one of a hypothetical user in relation to its neighbourhood, detecting how users with a good neighbourhood -- hence of a superior connections set -- obtain better information. We tested our measures on Amazons' instances, obtaining evidence which confirm the relevance of information extracted from nodes' neighbourhood in order to steer targeted users.
SSRN
This study provides formal theoretical evidence that value maximization is a rational behavioral, as opposed to rational expectations valuation rubric. Rational behavioral character of the value maximization rubric is evident in the axiomatic finding that, absent arrival of any unanticipated perturbations to pre-existing equilibriums, the value maximization rubric is significantly more likely to generate price drifts that are negative. In absence of any unanticipated perturbations then, investments in stock markets have character of preferences over lotteries. On the contrary, stock valuations that are arrived at in context of Pareto optimality of firms' investment allocations (`Pareto optimality') are shown to conform with rational expectations equilibriums. Under assumption of non-arrival of any unanticipated perturbations to pre-existing equilibriums, stock valuations arrived at in context of Pareto optimality always generate positive price drift. Totality of study findings provide evidence that stock markets function less as preferences over lotteries if parameterization of stock prices is transitioned from the value maximization rubric to the Pareto optimality rubric. In this respect, it is noteworthy that the workhorse Gordon Growth Model can produce stock valuations that are adapted to either of the Pareto optimality, or value maximization rubrics. It is normative then that there exists demand for an addendum to the Gordon Growth model, an addendum, which mitigates arrival at stock valuations that conform with rational behavioral, as opposed rational expectations equilibriums.
SSRN
Our study investigates the efficiency and the accuracy of collecting informetric data via Web crawling in accounting and finance research. The investigation includes a critical examination of the practice of Web crawling architecture through two logics, namely combinational logic (CL) and sequential logic (SL). To contrast the results, we specifically examine price informativeness and corporate social responsibility (CSR) public information release in the context of unicorns, where informetric data is collected via Google News Search engine based on the two logics of Web crawling architecture. We find that the CL approach is more optimal than it through SL approach by means of time-consumption, storage efficiency, and data accuracy. Based on the informetric data collected by Web crawling, our results suggest CSR media exposures have substantial influence on price informativeness. Optimistic and positive tone CSR information improves significantly price informativeness, whereas pessimistic and negative CSR exposures lowers price informativeness. Comparing to the market-based measures of price informativeness, we did not find significance between CSR media exposures and historical accounting performance. We also find the Loughran and McDonald wordlist shows higher classification accuracy than the Diction wordlist when analysing media release in financial context.
