Computational Economics
http://lists.repec.orgmailman/listinfo/nep-cmp
Computational Economics
2016-04-23
A Method for Agent-Based Models Validation
http://d.repec.org/n?u=RePEc:ssa:lemwps:2016/16&r=cmp
This paper proposes a new method for empirically validate simulation models that generate artificial time series data comparable with real-world data. The approach is based on comparing structures of vector autoregression models which are estimated from both artificial and real-world data by means of causal search algorithms. This relatively simple procedure is able to tackle both the problem of confronting theoretical simulation models with the data and the problem of comparing different models in terms of their empirical reliability. Moreover the paper provides an application of the validation procedure to the Dosi et al. (2015) macro-model.
Mattia Guerini
Alessio Moneta
Models validation; Agent-Based models; Causality; Structural Vector Autoregressions
2016-12-04
Comparing the market risk premia forecasts in JSE and NYSE equity markets
http://d.repec.org/n?u=RePEc:got:gotcrc:203&r=cmp
This paper examines the evidence regarding predictability in the market risk premium using artificial neural networks (ANNs), namely the Elman Network (EN) and the Higher Order Neural network (HONN), univariate ARMA and exponential smoothing techniques, such as Single Exponential Smoothing (SES) and Exponentially Weighted Moving Average (EWMA). The contribution of this paper is the inclusion of the South African market risk premium to the forecasting exercise and its direct comparison with US forecasting results. The market risk premium is defined as the expected rate of return on the market portfolio in excess of the shortterm interest rate for each market. All data are taken from January 2007 till December 2014 on a daily basis. Elman networks provide superior results among the tested models in both insample and out-of sample periods as well as among the tested markets. In general, neural networks beat the naive benchmark model and achieve to perform better than the rest of their linear tested counterparts. The forecasting models successfully capture patterns in the data that improve the forecasting accuracy of the tested models. Therefore, they can be applied to trading and investment purposes.
Leoni Eleni Oikonomikou
forecasting performance; market risk premium; South African stock market; US stock market
2016-04-14
Reducing the role of random numbers in matching algorithms for school admission
http://d.repec.org/n?u=RePEc:pra:mprapa:70374&r=cmp
New methods for solving the college admissions problem with indifference are presented and characterised with a Monte Carlo simulation in a variety of simple scenarios. Based on a qualifier defined as the average rank, it is found that these methods are more efficient than the Boston and Deferred Acceptance algorithms. The improvement in efficiency is directly related to the reduced role of random tie-breakers. The strategy-proofness of the new methods is assessed as well.
Hulsbergen, Wouter
college admission problem; deferred acceptance algorithm; Boston algorithm; Zeeburg algorithm; pairwise exchange algorithm; strategic behaviour
2016-03-12
Macroeconomic Policy in DSGE and Agent-Based Models Redux: New Developments and Challenges Ahead
http://d.repec.org/n?u=RePEc:ssa:lemwps:2016/17&r=cmp
The Great Recession seems to be a natural experiment for economic analysis, in that it has shown the inadequacy of the predominant theoretical framework -- the New Neoclassical Synthesis (NNS) -- grounded on the DSGE model. In this paper, we present a critical discussion of the theoretical, empirical and political-economy pitfalls of the DSGE-based approach to policy analysis. We suggest that a more fruitful research avenue should escape the strong theoretical requirements of NNS models (e.g., equilibrium, rationality, representative agent, etc.) and consider the economy as a complex evolving system, i.e. as an ecology populated by heterogenous agents, whose far-from-equilibrium interactions continuously change the structure of the system. This is indeed the methodological core of agent-based computational economics (ACE), which is presented in this paper. We also discuss how ACE has been applied to policy analysis issues, and we provide a survey of macroeconomic policy applications (fiscal and monetary policy, bank regulation, labor market structural reforms and climate change interventions). Finally, we conclude by discussing the methodological status of ACE, as well as the problems it raises.
Giorgio Fagiolo
Andrea Roventini
Economic Policy, New Neoclassical Synthesis, New Keynesian Models, DSGE Models, Agent-Based Computational Economics, Agent-Based Models, Complexity Theory, Great Recession, Crisis
2016-04-13
An agent-based model of dynamics in corporate bond trading
http://d.repec.org/n?u=RePEc:boe:boeewp:0592&r=cmp
We construct a heterogeneous agent-based model of the corporate bond market capturing the interaction of market maker behaviour, fund trading strategies, and cash allocation by investors in funds to study feedback effects and the impact of market changes. The model parameters are calibrated against empirical data on US corporate bond trading. Where available, inputs are taken from market data. Others are calibrated through matching statistical features of market returns such as auto-correlations, volatility and fat tails. We use the model to explore the impact of shocks. We find that the sensitivity of the market maker to demand and the degree to which momentum traders are active strongly influence the over and undershooting of yields in response to a shock. This suggests that correlation in fundsâ€™ trading strategies can exacerbate extreme price movements and contribute to the procyclicality of financial markets. While the behaviour of investors in funds based on past experience plays a comparatively smaller role in model dynamics, it represents another source of amplification which could be particularly problematic if investors respond to a shock with greater risk aversion. Simple measures to reduce the speed with which investors can redeem investments can reduce the extent of yield dislocation. We also explore the impact of the growth in passive investment, and find that it increases the tail risk of big yield dislocations after shocks, though, on average, volatility may be reduced.
Braun-Munzinger, Karen
Liu, Zijun
Turrell, Arthur
Agent-based model; corporate bond market; trading strategies
2016-04-18
Policy Conflicts and the Performance of Emissions Trading Markets: An Adaptive Agent-based Analysis
http://d.repec.org/n?u=RePEc:eep:report:rr20160339&r=cmp
Bing Zhang
Yongliang Zhang
Emissions,Agent-based Analysis,Trading Markets
2016-03
Deep neural networks, gradient-boosted trees, random forests: Statistical arbitrage on the S&P 500
http://d.repec.org/n?u=RePEc:zbw:iwqwdp:032016&r=cmp
In recent years, machine learning research has gained momentum: New developments in the field of deep learning allow for multiple levels of abstraction and are starting to supersede well-known and powerful tree-based techniques mainly operating on the original feature space. All these methods can be applied to various fields, including finance. This article implements and analyses the effectiveness of deep neural networks (DNN), gradient-boosted-trees (GBT), random forests (RAF), and a combination (ENS) of these methods in the context of statistical arbitrage. Each model is trained on lagged returns of all stocks in the S&P 500, after elimination of survivor bias. From 1992 to 2015, daily one-day-ahead trading signals are generated based on the probability forecast of a stock to outperform the general market. The highest k probabilities are converted into long and the lowest k probabilities into short positions, thus censoring the less certain middle part of the ranking. Empirical findings are promising. A simple ensemble consisting of one deep neural network, one gradient-boosted tree, and one random forest produces out-of-sample returns exceeding 0.45 percent per day for k = 10, prior to transaction costs. Irrespective of the fact that profits are declining in recent years, our findings pose a severe challenge to the semi-strong form of market efficiency.
Krauss, Christopher
Do, Xuan Anh
Huck, Nicolas
statistical arbitrage,deep learning,gradient-boosting,random forests,ensemble learning
2016
Simulated ML Estimation of Financial Agent-Based Models
http://d.repec.org/n?u=RePEc:fau:wpaper:wp2016_07&r=cmp
This paper proposes computational framework for empirical estimation of Financial Agent-Based Models (FABMs) that does not rely upon restrictive theoretical assumptions. We customise a recent methodology of the Non-Parametric Simulated Maximum Likelihood Estimator (NPSMLE) based on kernel methods by Kristensen and Shin (2012) and elaborate its capability for FABMs estimation purposes. To start with, we apply the methodology to the popular and widely analysed model of Brock and Hommes (1998). We extensively test finite sample properties of the estimator via Monte Carlo simulations and show that important theoretical features of the estimator, the consistency and asymptotic efficiency, also hold in small samples for the model. We also verify smoothness of the simulated log-likelihood function and identification of parameters. Main empirical results of our analysis are the statistical insignificance of the switching coefficient but markedly significant belief parameters defining heterogeneous trading regimes with an absolute superiority of trend-following over contrarian strategies and a slight proportional dominance of fundamentalists over trend following chartists.
Jiri Kukacka
Jozef Barunik
Heterogeneous Agent Model, Heterogeneous Expectations, Behavioural Finance, Intensity of Choice, Switching, Non-Parametric Simulated Maximum Likelihood Estimator
2016-03
Assessment of Post-merger Coordinated Effects: Characterization by Simulations
http://d.repec.org/n?u=RePEc:cpr:ceprdp:11218&r=cmp
This paper aims at evaluating the coordinated effects of horizontal mergers by simulating their impact on firms' critical discount factors. We consider a random coefficient model on the demand side and heterogeneous price-setting firms on the supply side. Results suggest that mergers strengthen the incentives to collude among merging parties, but weaken the incentives of non-merging parties, with the former effect being stronger. To assess the magnitudes of these effects, we introduce the concepts of Asymmetry in Payoffs and Change in Payoffs effects, which allow us to identify appropriate screening tools according to the relative pre-merger payoffs of merging parties.
Ivaldi, Marc
Lagos, Vicente
Collusion; Coordinated effects; Critical Discount Factor; Merger Simulation
2016-04