|
on Network Economics |
| By: | Florian Huber; Gary Koop; Massimiliano Marcellino; Tobias Scheckel |
| Abstract: | Commonly used priors for Vector Autoregressions (VARs) induce shrinkage on the autoregressive coefficients. Introducing shrinkage on the error covariance matrix is sometimes done but, in the vast majority of cases, without considering the network structure of the shocks and by placing the prior on the lower Cholesky factor of the precision matrix. In this paper, we propose a prior on the VAR error precision matrix directly. Our prior, which resembles a standard spike and slab prior, models variable inclusion probabilities through a stochastic block model that clusters shocks into groups. Within groups, the probability of having relations across group members is higher (inducing less sparsity) whereas relations across groups imply a lower probability that members of each group are conditionally related. We show in simulations that our approach recovers the true network structure well. Using a US macroeconomic data set, we illustrate how our approach can be used to cluster shocks together and that this feature leads to improved density forecasts. |
| Date: | 2024–07 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2407.16349 |
| By: | Lillethun, Erik; Sharma, Rishi |
| Abstract: | This paper develops a theoretical model of cyberwarfare between nations, focusing on the factors that determine the severity and outcomes of cyber conflicts. We introduce a two-country model where nations invest in offensive or defensive cyber capabilities across networked systems. We show that resource expenditure intensifies when players' effective values are similar, which can help explain the rise of cyberwarfare. We explore the implications of network structures, showing how larger attack surfaces worsen outcomes for defenders. Additionally, we investigate the impact of private cyber defence provision, and find that centralized policies may either improve or exacerbate cyber conflict. |
| Keywords: | cyberwarfare; cyberattacks; networks |
| JEL: | C72 D74 D85 F5 |
| Date: | 2024–06–18 |
| URL: | https://d.repec.org/n?u=RePEc:pra:mprapa:121299 |
| By: | Célestin Coquidé (UTINAM - Univers, Théorie, Interfaces, Nanostructures, Atmosphère et environnement, Molécules (UMR 6213) - INSU - CNRS - Institut national des sciences de l'Univers - CNRS - Centre National de la Recherche Scientifique - UFC - Université de Franche-Comté - UBFC - Université Bourgogne Franche-Comté [COMUE]); José Lages (UTINAM - Univers, Théorie, Interfaces, Nanostructures, Atmosphère et environnement, Molécules (UMR 6213) - INSU - CNRS - Institut national des sciences de l'Univers - CNRS - Centre National de la Recherche Scientifique - UFC - Université de Franche-Comté - UBFC - Université Bourgogne Franche-Comté [COMUE]); Dima Shepelyansky (LPT - Laboratoire de Physique Théorique - UT3 - Université Toulouse III - Paul Sabatier - UT - Université de Toulouse - CNRS - Centre National de la Recherche Scientifique - FeRMI - Fédération de recherche « Matière et interactions » - INSA Toulouse - Institut National des Sciences Appliquées - Toulouse - INSA - Institut National des Sciences Appliquées - UT - Université de Toulouse - UT3 - Université Toulouse III - Paul Sabatier - UT - Université de Toulouse - CNRS - Centre National de la Recherche Scientifique) |
| Abstract: | Abstract During the April 2023 Brazil–China summit, the creation of a trade currency supported by the BRICS countries was proposed. Using the United Nations Comtrade database, providing the frame of the world trade network associated to 194 UN countries during the decade 2010–2020, we study a mathematical model of influence battle of three currencies, namely, the US dollar, the euro, and such a hypothetical BRICS currency. In this model, a country trade preference for one of the three currencies is determined by a multiplicative factor based on trade flows between countries and their relative weights in the global international trade. The three currency seed groups are formed by 9 eurozone countries for the euro, 5 Anglo-Saxon countries for the US dollar and the 5 BRICS countries for the new proposed currency. The countries belonging to these 3 currency seed groups trade only with their own associated currency whereas the other countries choose their preferred trade currency as a function of the trade relations with their commercial partners. The trade currency preferences of countries are determined on the basis of a Monte Carlo modeling of Ising type interactions in magnetic spin systems commonly used to model opinion formation in social networks. We adapt here these models to the world trade network analysis. The results obtained from our mathematical modeling of the structure of the global trade network show that as early as 2012 about 58% of countries would have preferred to trade with the BRICS currency, 23% with the euro and 19% with the US dollar. Our results announce favorable prospects for a dominance of the BRICS currency in international trade, if only trade relations are taken into account, whereas political and other aspects are neglected. |
| Keywords: | World trade network, International trade, Currency, Opinion formation model |
| Date: | 2023–09–19 |
| URL: | https://d.repec.org/n?u=RePEc:hal:journl:hal-04641803 |
| By: | Shi, Chengchun; Zhou, Yunzhe; Li, Lexin |
| Abstract: | In this article, we propose a new hypothesis testing method for directed acyclic graph (DAG). While there is a rich class of DAG estimation methods, there is a relative paucity of DAG inference solutions. Moreover, the existing methods often impose some specific model structures such as linear models or additive models, and assume independent data observations. Our proposed test instead allows the associations among the random variables to be nonlinear and the data to be time-dependent. We build the test based on some highly flexible neural networks learners. We establish the asymptotic guarantees of the test, while allowing either the number of subjects or the number of time points for each subject to diverge to infinity. We demonstrate the efficacy of the test through simulations and a brain connectivity network analysis. Supplementary materials for this article are available online. |
| Keywords: | brain connectivity networks; directed acrylic graph; hypothesis testing; generative adversarial networks; multilayer perceptron neural networks; Hypothesis testing; CIF-2102227; R01AG061303; R01AG062542; EP/W014971/1 |
| JEL: | C1 |
| Date: | 2023–07–12 |
| URL: | https://d.repec.org/n?u=RePEc:ehl:lserod:119446 |
| By: | Kamesh Korangi; Christophe Mues; Cristi\'an Bravo |
| Abstract: | Apart from assessing individual asset performance, investors in financial markets also need to consider how a set of firms performs collectively as a portfolio. Whereas traditional Markowitz-based mean-variance portfolios are widespread, network-based optimisation techniques have built upon these developments. However, most studies do not contain firms at risk of default and remove any firms that drop off indices over a certain time. This is the first study to incorporate risky firms and use all the firms in portfolio optimisation. We propose and empirically test a novel method that leverages Graph Attention networks (GATs), a subclass of Graph Neural Networks (GNNs). GNNs, as deep learning-based models, can exploit network data to uncover nonlinear relationships. Their ability to handle high-dimensional features and accommodate customised layers for specific purposes makes them particularly appealing for large-scale problems such as mid- and small-cap portfolio optimization. This study utilises 30 years of data on mid-cap firms, creating graphs of firms using distance correlation and the Triangulated Maximally Filtered Graph approach. These graphs are the inputs to a GAT model that we train using custom layers which impose weight and allocation constraints and a loss function derived from the Sharpe ratio, thus directly maximising portfolio risk-adjusted returns. This new model is benchmarked against a network characteristic-based portfolio, a mean variance-based portfolio, and an equal-weighted portfolio. The results show that the portfolio produced by the GAT-based model outperforms all benchmarks and is consistently superior to other strategies over a long period while also being informative of market dynamics. |
| Date: | 2024–07 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2407.15532 |
| By: | Wenbo Yan; Ying Tan |
| Abstract: | Recently, the incorporation of both temporal features and the correlation across time series has become an effective approach in time series prediction. Spatio-Temporal Graph Neural Networks (STGNNs) demonstrate good performance on many Temporal-correlation Forecasting Problem. However, when applied to tasks lacking periodicity, such as stock data prediction, the effectiveness and robustness of STGNNs are found to be unsatisfactory. And STGNNs are limited by memory savings so that cannot handle problems with a large number of nodes. In this paper, we propose a novel approach called the Temporal-Correlation Graph Pre-trained Network (TCGPN) to address these limitations. TCGPN utilize Temporal-correlation fusion encoder to get a mixed representation and pre-training method with carefully designed temporal and correlation pre-training tasks. Entire structure is independent of the number and order of nodes, so better results can be obtained through various data enhancements. And memory consumption during training can be significantly reduced through multiple sampling. Experiments are conducted on real stock market data sets CSI300 and CSI500 that exhibit minimal periodicity. We fine-tune a simple MLP in downstream tasks and achieve state-of-the-art results, validating the capability to capture more robust temporal correlation patterns. |
| Date: | 2024–07 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2407.18519 |
| By: | Xavier Giroud; Ernest Liu; Holger Mueller |
| Abstract: | The vast majority of U.S. inventors work for firms that also have inventors and plants in other tech clusters. Using merged USPTO–U.S. Census Bureau plant-level data, we show that larger tech clusters not only make local inventors more productive but also raise the productivity of inventors and plants in other clusters, which are connected to the focal cluster through their parent firms' networks of innovating plants. Cross-cluster innovation spillovers do not depend on the physical distance between clusters, and plants cite disproportionately more patents from other firms in connected clusters, across large physical distances. To rationalize these findings, and to inform policy, we develop a tractable model of spatial innovation that features both within- and cross-cluster innovation spillovers. Based on our model, we derive a sufficient statistic for the wedge between the social and private returns to innovation in a given location. Taking the model to the data, we rank all U.S. tech clusters according to this wedge. While larger tech clusters exhibit a greater social-private innovation wedge, this is not because of local knowledge spillovers, but because they are well-connected to other clusters through firms' networks of innovating plants. In counterfactual exercises, we show that an increase in the interconnectedness of U.S. tech clusters raises the social-private innovation wedge in (almost) all locations, but especially in tech clusters that are large and well-connected to other clusters. |
| JEL: | G30 O30 R30 |
| Date: | 2024–07 |
| URL: | https://d.repec.org/n?u=RePEc:nbr:nberwo:32677 |
| By: | Shi, Xiangyu; Liu, Yu |
| Abstract: | This paper establishes a novel argument that social networks among local politicians reduce spatial frictions of corporate investment. We leverage the replacement of city officials and the resulting exogenous variations of hometown ties among city party secretaries to examine their impact on intercity capital flows in China. The results provide strong evidence that such connections significantly enhance capital flows between cities. These social bonds appear to effectively lower entry barriers for businesses and offer sustained support to connected firms without negatively impacting unconnected ones. Our research indicates that the increase in hometown-related investments does not displace non-hometown-related investments. |
| Keywords: | hometown ties, capital flow, transaction costs, rent seeking, economic growth |
| JEL: | D2 D7 G1 O1 |
| Date: | 2024–07 |
| URL: | https://d.repec.org/n?u=RePEc:pra:mprapa:121412 |
| By: | Sidarth Erat |
| Abstract: | The Colonel Blotto game, introduced by Borel in the 1920s, is often used for modeling various real-life settings, such as elections, lobbying, etc. The game is based on the allocation of limited resources by players to a set of fields. Each field is ``won'' and a corresponding field-specific value is obtained by the player who sends the most resources. In this paper, we formulate a discrete Blotto game played on a general \textit{accessibility network} (i.e., the bipartite graph made of players and the fields they can allocate resources to). The primary goal is to find how the topology of the accessibility network controls the existence and uniqueness of equilibrium allocations, and how it affects the fraction of fields that are entered and the average payoff of players at equilibrium. We establish that, in a 2-regular topology, when the values of fields are close enough and the number of players is not a multiple of 4, then there is a unique equilbrium. We also prove that players are better off and fields are more likely to be entered in a regular topology than a random topology. We find numerically that dispersion of field weights negatively affects average player payoff. The main contribution is a framework for analyzing contests where players are permitted access to some (but not necessarily all) venues of competition. |
| Date: | 2024–07 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2407.16707 |