|
on Computational Economics |
|
Issue of 2020‒10‒19
twenty-two papers chosen by |
| By: | Lorenc Kapllani; Long Teng |
| Abstract: | We study deep learning-based schemes for solving high dimensional nonlinear backward stochastic differential equations (BSDEs). First we show how to improve the performances of the proposed scheme in [W. E and J. Han and A. Jentzen, Commun. Math. Stat., 5 (2017), pp.349-380] regarding computational time and stability of numerical convergence by using the advanced neural network architecture instead of the stacked deep neural networks. Furthermore, the proposed scheme in that work can be stuck in local minima, especially for a complex solution structure and longer terminal time. To solve this problem, we investigate to reformulate the problem by including local losses and exploit the Long Short Term Memory (LSTM) networks which are a type of recurrent neural networks (RNN). Finally, in order to study numerical convergence and thus illustrate the improved performances with the proposed methods, we provide numerical results for several 100-dimensional nonlinear BSDEs including a nonlinear pricing problem in finance. |
| Date: | 2020–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2010.01319&r=all |
| By: | Chuheng Zhang; Yuanqi Li; Xi Chen; Yifei Jin; Pingzhong Tang; Jian Li |
| Abstract: | Modern machine learning models (such as deep neural networks and boosting decision tree models) have become increasingly popular in financial market prediction, due to their superior capacity to extract complex non-linear patterns. However, since financial datasets have very low signal-to-noise ratio and are non-stationary, complex models are often very prone to overfitting and suffer from instability issues. Moreover, as various machine learning and data mining tools become more widely used in quantitative trading, many trading firms have been producing an increasing number of features (aka factors). Therefore, how to automatically select effective features becomes an imminent problem. To address these issues, we propose DoubleEnsemble, an ensemble framework leveraging learning trajectory based sample reweighting and shuffling based feature selection. Specifically, we identify the key samples based on the training dynamics on each sample and elicit key features based on the ablation impact of each feature via shuffling. Our model is applicable to a wide range of base models, capable of extracting complex patterns, while mitigating the overfitting and instability issues for financial market prediction. We conduct extensive experiments, including price prediction for cryptocurrencies and stock trading, using both DNN and gradient boosting decision tree as base models. Our experiment results demonstrate that DoubleEnsemble achieves a superior performance compared with several baseline methods. |
| Date: | 2020–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2010.01265&r=all |
| By: | Tullio Mancini; Hector Calvo-Pardo; Jose Olmo |
| Abstract: | The aim of this paper is to propose a suitable method for constructing prediction intervals for the output of neural network models. To do this, we adapt the extremely randomized trees method originally developed for random forests to construct ensembles of neural networks. The extra-randomness introduced in the ensemble reduces the variance of the predictions and yields gains in out-of-sample accuracy. An extensive Monte Carlo simulation exercise shows the good performance of this novel method for constructing prediction intervals in terms of coverage probability and mean square prediction error. This approach is superior to state-of-the-art methods extant in the literature such as the widely used MC dropout and bootstrap procedures. The out-of-sample accuracy of the novel algorithm is further evaluated using experimental settings already adopted in the literature. |
| Date: | 2020–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2010.04044&r=all |
| By: | Janusz Gajda (Faculty of Economic Sciences, University of Warsaw); Rafał Walasek (Faculty of Economic Sciences, University of Warsaw) |
| Abstract: | This article covers the implementation of fractional (non-integer order) differentiation on four datasets based on stock prices of main international stock indexes: WIG 20, S&P 500, DAX, Nikkei 225. This concept has been proposed by Lopez de Prado to find the most appropriate balance between zero differentiation and fully differentiated time series. The aim is making time series stationary while keeping its memory and predictive power. This paper makes also the comparison between fractional and classical differentiation in terms of the effectiveness of artificial neural networks. This comparison is done in two viewpoints: Root Mean Square Error (RMSE) and Mean Absolute Error (MAE). The conclusion of the study is that fractionally differentiated time series performed better in trained ANN. |
| Keywords: | fractional differentiation, financial time series, stock exchange, artificial neural networks |
| JEL: | C22 C32 G10 |
| Date: | 2020 |
| URL: | http://d.repec.org/n?u=RePEc:war:wpaper:2020-32&r=all |
| By: | Mosavi, Amir; Faghan, Yaser; Ghamisi, Pedram; Duan, Puhong; Ardabili, Sina Faizollahzadeh; Hassan, Salwana; Band, Shahab S. |
| Abstract: | The popularity of deep reinforcement learning (DRL) applications in economics has increased exponentially. DRL, through a wide range of capabilities from reinforcement learning (RL) to deep learning (DL), offers vast opportunities for handling sophisticated dynamic economics systems. DRL is characterized by scalability with the potential to be applied to high-dimensional problems in conjunction with noisy and nonlinear patterns of economic data. In this paper, we initially consider a brief review of DL, RL, and deep RL methods in diverse applications in economics, providing an in-depth insight into the state-of-the-art. Furthermore, the architecture of DRL applied to economic applications is investigated in order to highlight the complexity, robustness, accuracy, performance, computational tasks, risk constraints, and profitability. The survey results indicate that DRL can provide better performance and higher efficiency as compared to the traditional algorithms while facing real economic problems in the presence of risk parameters and the ever-increasing uncertainties. |
| Date: | 2020–09–01 |
| URL: | http://d.repec.org/n?u=RePEc:osf:osfxxx:jrc58&r=all |
| By: | Abramov, Dimitri Marques |
| Abstract: | Despite the market economy to be contemporaneously considered as a complex adaptive system, there are no collective feedback mechanism that provide long range stability and complexity to the system. In this scenario, the logical prediction is a long-term economic collapse by positive loops. In this work, I outline the fundamental idea of a floating taxation system as a feedback system to prevent market collapse by asymmetrical company overgrowth and extreme reduction of system complexity. The paradigm would promote the long-term stability of the economic system. I’ve implemented a generic computational neural network with 5000 virtual companies whose initial states (i.e. capital) and connective weights (trading network) were normally distributed. A negative feedback loop was implemented with different weights. The market complexity was measured in terms of joint entropy in an algorithm to calc neural complexity in networks. Without feedback, some companies had explosive growth annihilating all collateral ones until all system collapses. With feedback loops, the complexity was stable while many companies disappeared (negative selection) and the capital variance substantially increased (from 10 units in initial conditions to 2000 times) as well complexity (increment on order to 104). This data supports a theory about feedback dynamic mechanisms for market self-regulation based on floating taxes, maintaining homeostasis with complexity, capital growth, and competitive balance. |
| Date: | 2020–09–15 |
| URL: | http://d.repec.org/n?u=RePEc:osf:socarx:xj2gb&r=all |
| By: | Nicola Curci (Bank of Italy); Giuseppe Grasso (Luxembourg Institute of Socio-Economic Research (LISER) and University of Luxembourg); Pasquale Recchia (Bank of Italy); Marco Savegnago (Bank of Italy) |
| Abstract: | Introduced in 2019, the Reddito di cittadinanza (RdC) has replaced the Reddito di inclusione (ReI) as a universal minimum income scheme in Italy. In this paper, we use BIMic, the Bank of Italy’s static (non-behavioural) microsimulation model, to measure the effects of the RdC in terms of inequality reduction and, as a novel contribution, of absolute poverty alleviation. Our results, which do not account for behavioural responses to policy changes, show that the RdC is effective in reducing inequality, and attenuating the incidence, and even more so the intensity, of absolute poverty. We also document how certain features of the design of this benefit affect the distribution of these effects across the population. For this purpose, we simulate two hypothetical changes to the current design of the RdC: one that directs more resources to large households with minors (on average more in need than other households) and the other that takes into account the differences in the cost of living according to geographical areas and municipality size. |
| Keywords: | microsimulation model, redistribution, poverty, minimum income, progressivity |
| JEL: | C15 C63 H23 H31 I32 |
| Date: | 2020–09 |
| URL: | http://d.repec.org/n?u=RePEc:bdi:wptemi:td_1298_20&r=all |
| By: | Battula, Swathi; Tesfatsion, Leigh; McDermott, Thomas E. |
| Abstract: | The ERCOT Test System developed in this study is an open-source library of Java/Python software classes, together with a synthetic grid construction method, specifically designed to facilitate the study of ERCOT market operations over successive days. In default form, these classes permit a high-level modeling of existing ERCOT market operations. Users can conduct a broad range of computational experiments under alternative parameter settings. In addition, users can readily extend these classes to model additional existing or envisioned ERCOT market features to suit different research purposes. An 8-bus test case is used to illustrate the capabilities of the test system. Ongoing studies making use of the test system to model larger-scale transmission components for integrated transmission and distribution systems are also reported. |
| Date: | 2019–12–23 |
| URL: | http://d.repec.org/n?u=RePEc:isu:genstf:201912230800001078&r=all |
| By: | Nadja Klein; Michael Stanley Smith; David J. Nott |
| Abstract: | Recurrent neural networks (RNNs) with rich feature vectors of past values can provide accurate point forecasts for series that exhibit complex serial dependence. We propose two approaches to constructing deep time series probabilistic models based on a variant of RNN called an echo state network (ESN). The first is where the output layer of the ESN has stochastic disturbances and a shrinkage prior for additional regularization. The second approach employs the implicit copula of an ESN with Gaussian disturbances, which is a deep copula process on the feature space. Combining this copula with a non-parametrically estimated marginal distribution produces a deep distributional time series model. The resulting probabilistic forecasts are deep functions of the feature vector and also marginally calibrated. In both approaches, Bayesian Markov chain Monte Carlo methods are used to estimate the models and compute forecasts. The proposed deep time series models are suitable for the complex task of forecasting intraday electricity prices. Using data from the Australian National Electricity Market, we show that our models provide accurate probabilistic price forecasts. Moreover, the models provide a flexible framework for incorporating probabilistic forecasts of electricity demand as additional features. We demonstrate that doing so in the deep distributional time series model in particular, increases price forecast accuracy substantially. |
| Date: | 2020–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2010.01844&r=all |
| By: | Manson, Steven; An, Li; Clarke, Keith C.; Heppenstall, Alison; Koch, Jennifer; Krzyzanowski, Brittany; Morgan, Fraser; O'Sullivan, David; Runck, Bryan C.; Shook, Eric; Tesfatsion, Leigh |
| Abstract: | Agent based modeling (ABM) is a standard tool that is useful across many disciplines. Despite widespread and mounting interest in ABM, even broader adoption has been hindered by a set of methodological challenges that run from issues around basic tools to the need for a more complete conceptual foundation for the approach. After several decades of progress, ABMs remain difficult to develop and use for many students, scholars, and policy makers. This difficulty holds especially true for models designed to represent spatial patterns and processes across a broad range of human, natural, and human-environment systems. In this paper, we describe the methodological challenges facing further development and use of spatial ABM (SABM) and suggest some potential solutions from multiple disciplines. We first define SABM to narrow our object of inquiry, and then explore how spatiality is a source of both advantages and challenges. We examine how time interacts with space in models and delve into issues of model development in general and modeling frameworks and tools specifically. We draw on lessons and insights from fields with a history of ABM contributions, including economics, ecology, geography, ecology, anthropology, and spatial science with the goal of identifying promising ways forward for this powerful means of modeling. |
| Date: | 2020–01–01 |
| URL: | http://d.repec.org/n?u=RePEc:isu:genstf:202001010800001690&r=all |
| By: | Michael B\"ucker; Gero Szepannek; Alicja Gosiewska; Przemyslaw Biecek |
| Abstract: | 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. |
| Date: | 2020–09 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2009.13384&r=all |
| By: | Ferdinands, Gerbrich; Schram, Raoul; de Bruin, Jonathan; Bagheri, Ayoub; Oberski, Daniel Leonard (Tilburg University); Tummers, Lars (Utrecht University); van de Schoot, Rens |
| Abstract: | Background Conducting a systematic review requires great screening effort. Various tools have been proposed to speed up the process of screening thousands of titles and abstracts by engaging in active learning. In such tools, the reviewer interacts with machine learning software to identify relevant publications as early as possible. To gain a comprehensive understanding of active learning models for reducing workload in systematic reviews, the current study provides a methodical overview of such models. Active learning models were evaluated across four different classification techniques (naive Bayes, logistic regression, support vector machines, and random forest) and two different feature extraction strategies (TF-IDF and doc2vec). Moreover, models were evaluated across six systematic review datasets from various research areas to assess generalizability of active learning models across different research contexts. Methods Performance of the models were assessed by conducting simulations on six systematic review datasets. We defined desirable model performance as maximizing recall while minimizing the number of publications needed to screen. Model performance was evaluated by recall curves, WSS@95, RRF@10, and ATD. Results Within all datasets, the model performance exceeded screening at random order to a great degree. The models reduced the number of publications needed to screen by 91.7% to 63.9%. Conclusions Active learning models for screening prioritization show great potential in reducing the workload in systematic reviews. Overall, the Naive Bayes + TF-IDF model performed the best. |
| Date: | 2020–09–16 |
| URL: | http://d.repec.org/n?u=RePEc:osf:osfxxx:w6qbg&r=all |
| By: | Hinterlang, Natascha; Hollmayr, Josef |
| Abstract: | This paper identifies U.S. monetary and fiscal dominance regimes using machine learning techniques. The algorithms are trained and verified by employing simulated data from Markov-switching DSGE models, before they classify regimes from 1968-2017 using actual U.S. data. All machine learning methods outperform a standard logistic regression concerning the simulated data. Among those the Boosted Ensemble Trees classifier yields the best results. We find clear evidence of fiscal dominance before Volcker. Monetary dominance is detected between 1984-1988, before a fiscally led regime turns up around the stock market crash lasting until 1994. Until the beginning of the new century, monetary dominance is established, while the more recent evidence following the financial crisis is mixed with a tendency towards fiscal dominance. |
| Keywords: | Monetary-fiscal interaction,Machine Learning,Classification,Markov-switching DSGE |
| JEL: | C38 E31 E63 |
| Date: | 2020 |
| URL: | http://d.repec.org/n?u=RePEc:zbw:bubdps:512020&r=all |
| By: | Tarun Bhatia |
| Abstract: | U.S. Nonfarm employment is considered one of the key indicators for assessing the state of the labor market. Considerable deviations from the expectations can cause market moving impacts. In this paper, the total U.S. nonfarm payroll employment is predicted before the release of the BLS employment report. The content herein outlines the process for extracting predictive features from the aggregated payroll data and training machine learning models to make accurate predictions. Publically available revised employment report by BLS is used as a benchmark. Trained models show excellent behaviour with R2 of 0.9985 and 99.99% directional accuracy on out of sample periods from January 2012 to March 2020. Keywords Machine Learning; Economic Indicators; Ensembling; Regression, Total Nonfarm Payroll |
| Date: | 2020–09 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2009.14282&r=all |
| By: | J-C Gerlach (ETH Zürich - Department of Management, Technology, and Economics (D-MTEC)); Jerome L Kreuser (ETH Zurich); Didier Sornette (ETH Zürich - Department of Management, Technology, and Economics (D-MTEC); Swiss Finance Institute; Southern University of Science and Technology; Tokyo Institute of Technology) |
| Abstract: | We present a modified version of the super-exponential rational expectations “Efficient Crashes” bubble model of (Kreuser and Sornette, 2019) with a different formulation of the expected return that makes clearer the additive nature of corrective jumps. We derive a Kelly trading strategy for the new model. We combine the strategy with a simplified estimation procedure for the model parameters from price time series. We optimize the control parameters of the trading strategy by maximizing the return-weighted accuracy of trades. This enables us to predict the out-of-sample optimal investment, purely based on in-sample calibration of the model on historical data. Our approach solves the difficult problem of selecting the portfolio rebalancing time, as we endogenize it as an optimization parameter. We develop an ex-ante backtest that allows us to test our strategy on twenty equity asset indices. We find that our trading strategy achieves positive trading performance for 95% of tested assets and outperforms the Buy-and-Hold-Strategy in terms of CAGR and Sharpe Ratio in 60% of cases. In our simulations, we do not allow for any short trading or leverage. Thus, we simply simulate allocation of 0-100% of one’s capital between a risk-free and the risky asset over time. The optimal rebalancing periods are mostly of duration around a month; thus, the model does not overtrade, ensuring reasonable trading costs. Furthermore, during crashes, the model reduces the invested amount of capital sufficiently soon to reduce impact of price drawdowns. In addition to the Dotcom bubble, the great financial crisis of 2008 and other historical crashes, our study also covers the most recent crash in March 2020 that happened globally as a consequence of the economic shutdowns that were imposed as a reaction to the spread of the Coronavirus across the world. |
| Keywords: | financial bubbles, efficient crashes, positive feedback, rational expectation, Kelly criterion, optimal investment, Covid-19 crash |
| JEL: | C53 G01 G17 |
| Date: | 2020–10 |
| URL: | http://d.repec.org/n?u=RePEc:chf:rpseri:rp2085&r=all |
| By: | Raquel Almeida Ramos; Federico Bassi (Università Cattolica del Sacro Cuore; Dipartimento di Economia e Finanza, Università Cattolica del Sacro Cuore); Dany Lang |
| Abstract: | This paper intends to contribute to the theoretical literature on the determinants of exchange rate fluctuations. We build an agent-based model, based on behavioral assumptions inspired by the literature on behavioral finance and by empirical surveys about the behavior of foreign exchange professionals. In our artificial economy with two countries, traders can speculate on both exchange and interest rates, and allocate their wealth across heterogeneous assets. Fundamentalists use both fundamental and technical analysis, while chartists only employ the latter, and are either trend followers or trend contrarians. In our model, trend contrarians and cash in mechanisms provide the sufficient stability conditions, and allow explaining and replicating most stylized facts of foreign exchange markets, namely (i) the excess volatility of the exchange rate with respect to its fundamentals, (ii) booms, busts and precarious equilibria, (iii) clusters of volatility, (iv) long memory and (v) fat tails. |
| Keywords: | retirement, Foreign exchange markets, clusters of volatility, fat tails, heterogeneous beliefs, agent-based models, Stock-flow consistent models. |
| JEL: | D40 D84 G11 G12 |
| Date: | 2020–10 |
| URL: | http://d.repec.org/n?u=RePEc:ctc:serie1:def090&r=all |
| By: | Rakshit Jha; Mattijs De Paepe; Samuel Holt; James West; Shaun Ng |
| Abstract: | This paper shows that temporal CNNs accurately predict bitcoin spot price movements from limit order book data. On a 2 second prediction time horizon we achieve 71\% walk-forward accuracy on the popular cryptocurrency exchange coinbase. Our model can be trained in less than a day on commodity GPUs which could be installed into colocation centers allowing for model sync with existing faster orderbook prediction models. We provide source code and data at https://github.com/Globe-Research/deep-o rderbook. |
| Date: | 2020–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2010.01241&r=all |
| By: | Lining Yu; Wolfgang Karl H\"ardle; Lukas Borke; Thijs Benschop |
| Abstract: | 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. |
| Date: | 2020–09 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2009.13222&r=all |
| By: | Marco Avellaneda; Juan Andr\'es Serur |
| Abstract: | Modeling cross-sectional correlations between thousands of stocks, across countries and industries, can be challenging. In this paper, we demonstrate the advantages of using Hierarchical Principal Component Analysis (HPCA) over the classic PCA. We also introduce a statistical clustering algorithm for identifying of homogeneous clusters of stocks, or "synthetic sectors". We apply these methods to study cross-sectional correlations in the US, Europe, China, and Emerging Markets. |
| Date: | 2020–10 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2010.04140&r=all |
| By: | Foltas, Alexander; Pierdzioch, Christian |
| Abstract: | We use quantile random forests (QRF) to study the efficiency of the growth forecasts published by three leading German economic research institutes for the sample period from 1970 to 2017. To this end, we use a large array of predictors, including topics extracted by means of computational-linguistics tools from the business-cycle reports of the institutes, to model the information set of the institutes. We use this array of predictors to estimate the quantiles of the conditional distribution of the forecast errors made by the institutes, and then fit a skewed t-distribution to the estimated quantiles. We use the resulting density forecasts to compute the log probability score of the predicted forecast errors. Based on an extensive insample and out-of-sample analysis, we find evidence, particularly in the case of longer-term forecasts, against the null hypothesis of strongly efficient forecasts. We cannot reject weak efficiency of forecasts. |
| Keywords: | Growth forecasts,Forecast efficiency,Quantile-random forests,Density forecasts |
| JEL: | C53 E32 E37 |
| Date: | 2020 |
| URL: | http://d.repec.org/n?u=RePEc:zbw:pp1859:21&r=all |
| By: | Eric Benhamou; David Saltiel; Sandrine Ungari; Abhishek Mukhopadhyay |
| Abstract: | Can an asset manager plan the optimal timing for her/his hedging strategies given market conditions? The standard approach based on Markowitz or other more or less sophisticated financial rules aims to find the best portfolio allocation thanks to forecasted expected returns and risk but fails to fully relate market conditions to hedging strategies decision. In contrast, Deep Reinforcement Learning (DRL) can tackle this challenge by creating a dynamic dependency between market information and hedging strategies allocation decisions. In this paper, we present a realistic and augmented DRL framework that: (i) uses additional contextual information to decide an action, (ii) has a one period lag between observations and actions to account for one day lag turnover of common asset managers to rebalance their hedge, (iii) is fully tested in terms of stability and robustness thanks to a repetitive train test method called anchored walk forward training, similar in spirit to k fold cross validation for time series and (iv) allows managing leverage of our hedging strategy. Our experiment for an augmented asset manager interested in sizing and timing his hedges shows that our approach achieves superior returns and lower risk. |
| Date: | 2020–09 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2009.14136&r=all |
| By: | Masahiro Kato; Shota Yasui |
| Abstract: | We consider training a binary classifier under delayed feedback (DF Learning). In DF Learning, we first receive negative samples; subsequently, some samples turn positive. This problem is conceivable in various real-world applications such as online advertisements, where the user action takes place long after the first click. Owing to the delayed feedback, simply separating the positive and negative data causes a sample selection bias. One solution is to assume that a long time window after first observing a sample reduces the sample selection bias. However, existing studies report that only using a portion of all samples based on the time window assumption yields suboptimal performance, and the use of all samples along with the time window assumption improves empirical performance. Extending these existing studies, we propose a method with an unbiased and convex empirical risk constructed from the whole samples under the time window assumption. We provide experimental results to demonstrate the effectiveness of the proposed method using a real traffic log dataset. |
| Date: | 2020–09 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2009.13092&r=all |