|
on Payment Systems and Financial Technology |
|
Issue of 2022‒08‒29
twenty-two papers chosen by |
| By: | Jones, Sam (UNU-WIDER); Sen, Kunal (University of Manchester) |
| Abstract: | Can digital labour market platforms reduce search frictions in formal or informal labour markets? We study this question using a randomized experiment embedded in a tracer study of the work transitions of graduates from technical and vocational colleges in Mozambique. We implement an encouragement design, inviting graduates by SMS to join established digital platforms: Biscate, a site to find freelancers for informal manual tasks; and Emprego, a conventional formal jobs website. In contrast to positive estimates of the contribution of both platforms to job outcomes from naïve (per-treatment) estimates, both intent-to-treat and complier average treatment effects are consistently zero in the full sample, while the impact on life satisfaction is negative. However, use of the informal jobs platform leads to better work outcomes for women, especially those with manual qualifications, for whom earnings rise by over 50%. |
| Keywords: | digital labour platforms, search frictions, technical and vocational education, unemployment, Mozambique |
| JEL: | J64 J68 O15 |
| Date: | 2022–06 |
| URL: | http://d.repec.org/n?u=RePEc:iza:izadps:dp15409&r= |
| By: | Daisuke Ikeda (Director and Senior Economist, Institute for Monetary and Economic Studies (currently, Financial System and Bank Examination Department), Bank of Japan (E-mail: daisuke.ikeda@boj.or.jp)) |
| Abstract: | The rise of digital money may bring about privately issued money that circulates across borders and coexists with public money. This paper uses an open-economy search model with multiple currencies to study the impact of such global money on monetary policy autonomy -- the capacity of central banks to set a policy instrument. I show that the circulation of global money can entail a loss of monetary policy autonomy, but it can be preserved if government policy that limits the amount or use of global money for transactions is introduced or if the global currency is subject to counterfeiting. The result suggests that global digital money and monetary policy autonomy can be compatible. |
| Keywords: | Cryptocurrency, Monetary policy autonomy, Currency counterfeiting, Government transaction policy |
| JEL: | D82 E4 E5 F31 |
| Date: | 2022–07 |
| URL: | http://d.repec.org/n?u=RePEc:ime:imedps:22-e-10&r= |
| By: | Carolina E S Mattsson; Teodoro Criscione; Frank W Takes |
| Abstract: | Circulation is the characteristic feature of successful currency systems, from community currencies to cryptocurrencies to national currencies. In this paper, we propose a network analysis methodology for studying circulation given a system's digital transaction records. This is applied to Sarafu, a digital community currency active in Kenya over a period that saw considerable economic disruption due to the COVID-19 pandemic. Representing Sarafu as a network of monetary flow among the 40,000 users reveals meaningful patterns at multiple scales. Circulation was highly modular, geographically localized, and occurring among users with diverse livelihoods. Network centrality highlights women's participation, early adopters, and the especially prominent role of community-based financial institutions. These findings have concrete implications for humanitarian and development policy, helping articulate when community currencies might best support interventions in marginalized areas. Overall, networks of monetary flow allow for studying circulation within digital currency systems at a striking level of detail. |
| Date: | 2022–07 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2207.08941&r= |
| By: | Samuel Senyo Okae (Bank of Ghana); Eugene Yarboi Mensah (Ghana Deposit Protection Corporation) |
| Abstract: | The usage of electronic money (e-money) for transactions has grown across Ghana and has the potential to revolutionise the cash-dominant economy to become cashless. Propelling this growth are mobile money operators (MMOs), which have developed to offer a specific type of e-money, termed mobile money (MM). With the increased use of mobile-money services and growth in the payment systems sector each day, it is imperative for Ghana to design a holistic approach to the use of e-money as well as consider its operationalisation of the coverage by the deposit insurer. The Ghana Payment Systems Act, 2019 (Act 987) sets out the rules for the issuance of e-money within Ghana and the supervision of the business of e-money institutions (EMIs), which includes MMOs. There are growing concerns about safeguarding client funds held by EMIs worldwide. In Ghana, client funds held by EMIs must be placed in custodial accounts at banks. As a result, it has become necessary for Ghana's deposit insurance system to consider how to protect these funds. Funds backing the electronic value belonging to customers of MMOs are kept in a custodian account which resides with banks and hence the need for these funds to be protected in case of a bank failure is being discussed. This brief describes the distinctions between deposits and e-money and provides a description of the key features of e-money in the Ghanaian context. It discusses the factors influencing the protection of e-money wallets and the float (defined as the cash equivalent of outstanding electronic money liabilities of an electronic money issuer with partner banks) kept with commercial banks. Finally, options to be considered for the possible protection of these wallets in case of bank liquidation are presented. |
| Keywords: | deposit insurance, bank resolution |
| JEL: | G21 G33 |
| Date: | 2022–08 |
| URL: | http://d.repec.org/n?u=RePEc:awl:finbri:9&r= |
| By: | Yassine Lefouili (TSE - Toulouse School of Economics - UT1 - Université Toulouse 1 Capitole - Université Fédérale Toulouse Midi-Pyrénées - EHESS - École des hautes études en sciences sociales - CNRS - Centre National de la Recherche Scientifique - INRAE - Institut National de Recherche pour l’Agriculture, l’Alimentation et l’Environnement); Leonardo Madio (Universita degli Studi di Padova) |
| Abstract: | Public authorities in many jurisdictions are concerned about the proliferation of illegal content and products on online platforms. One often discussed solution is to make the platform liable for third parties' misconduct. In this paper, we first identify platform incentives to stop online misconduct in the absence of liability. Then, we provide an economic appraisal of platform liability that highlights the intended and unintended effects of a more stringent liability rule on several key variables such as prices, terms and conditions, business models, and investments. Specifically, we discuss the impact of the liability regime applying to online platforms on competition between them and the incentives of third parties relying on them. Finally, we analyze the potential costs and benefits of measures that have received much attention in recent policy discussions. |
| Keywords: | Liability rules,Online platform,Illegal content and products,Intellectual property |
| Date: | 2022–06 |
| URL: | http://d.repec.org/n?u=RePEc:hal:journl:hal-03711652&r= |
| By: | Fantazzini, Dean |
| Abstract: | This paper examined a set of over two thousand crypto-coins observed between 2015 and 2020 to estimate their credit risk by computing their probability of death. We employed different definitions of dead coins, ranging from academic literature to professional practice, alternative forecasting models, ranging from credit scoring models to machine learning and time series-based models, and different forecasting horizons. We found that the choice of the coin death definition affected the set of the best forecasting models to compute the probability of death. However, this choice was not critical, and the best models turned out to be the same in most cases. In general, we found that the \textit{cauchit} and the zero-price-probability (ZPP) based on the random walk or the Markov Switching-GARCH(1,1) were the best models for newly established coins, whereas credit scoring models and machine learning methods using lagged trading volumes and online searches were better choices for older coins. These results also held after a set of robustness checks that considered different time samples and the coins' market capitalization. |
| Keywords: | Bitcoin, Crypto-assets, Crypto-currencies, Credit risk, Default Probability, Probability of Death, ZPP, Cauchit, Logit, Probit, Random Forests, Google Trends. |
| JEL: | C32 C35 C51 C53 C58 G12 G17 G32 G33 |
| Date: | 2022 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:113744&r= |
| By: | Koo, J.; Kramer, B.; Langan, Simon; Ghosh, A.; Monsalue, A. G.; Lunt, T. |
| Keywords: | Digital technology; Innovation; Data; Agrifood systems; Sustainability; Climate change; Risk; Weather forecasting; Digital divide; Access to information; Policies; Women |
| Date: | 2022 |
| URL: | http://d.repec.org/n?u=RePEc:iwt:bosers:h051155&r= |
| By: | Ramit Sawhney; Shivam Agarwal; Vivek Mittal; Paolo Rosso; Vikram Nanda; Sudheer Chava |
| Abstract: | The rapid spread of information over social media influences quantitative trading and investments. The growing popularity of speculative trading of highly volatile assets such as cryptocurrencies and meme stocks presents a fresh challenge in the financial realm. Investigating such "bubbles" - periods of sudden anomalous behavior of markets are critical in better understanding investor behavior and market dynamics. However, high volatility coupled with massive volumes of chaotic social media texts, especially for underexplored assets like cryptocoins pose a challenge to existing methods. Taking the first step towards NLP for cryptocoins, we present and publicly release CryptoBubbles, a novel multi-span identification task for bubble detection, and a dataset of more than 400 cryptocoins from 9 exchanges over five years spanning over two million tweets. Further, we develop a set of sequence-to-sequence hyperbolic models suited to this multi-span identification task based on the power-law dynamics of cryptocurrencies and user behavior on social media. We further test the effectiveness of our models under zero-shot settings on a test set of Reddit posts pertaining to 29 "meme stocks", which see an increase in trade volume due to social media hype. Through quantitative, qualitative, and zero-shot analyses on Reddit and Twitter spanning cryptocoins and meme-stocks, we show the practical applicability of CryptoBubbles and hyperbolic models. |
| Date: | 2022–05 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2206.06320&r= |
| By: | Jordan Hill |
| Abstract: | The digital environment offers opportunities that can enrich children’s physical and mental well-being. Yet, false and misleading digital content, including disinformation and misinformation, is a risk. It can deepen political polarisation, erode public trust in democratic institutions and threaten public health. Media literacy is part of a suite of policies countries are using to maximise digital opportunities and minimise digital risks. This paper has four parts. First, it outlines current research and definitions relating to false and misleading digital content and looks at children's behaviour in the digital environment. Second, the concepts of media literacy, digital literacy and other relevant competencies are discussed. Third, research on children’s experiences of false and misleading digital content and their perceived levels of digital media literacy is analysed. Finally, policies and practices which deliver media literacy are discussed. Research limitations and other barriers, such as teacher training, are described. |
| Date: | 2022–08–02 |
| URL: | http://d.repec.org/n?u=RePEc:oec:eduaab:275-en&r= |
| By: | Claire Greene; Joanna Stavins |
| Abstract: | Using transaction data from a US consumer payments diary, we revisit the credit card debt puzzle—a scenario in which consumers revolve credit card debt while also keeping liquid assets as bank account deposits. This scenario is very common: 42 percent of consumers in our sample were borrower-savers in 2019 (those who carry $100 or more in credit card debt and $100 or more in liquid assets). We explain the puzzle by showing that consumers need their liquid assets to pay monthly bills and other necessary expenses, including mortgage or rent. More than 80 percent of bills by value were paid out of bank accounts and could not be charged to credit cards, so bank account balances were needed to cover those basic expenses. On average, borrower-savers’ credit card debt exceeded their liquid assets. The average borrower-saver carried almost $6,400 in unpaid credit card debt and had $5,400 in liquid assets, including checking and savings accounts, cash, and general-purpose prepaid cards. Only 40 percent of borrower-savers had liquid assets greater than their unpaid credit card balance. In addition, borrower-savers’ monthly expenses (bills and purchases) averaged 77 percent of their liquid assets, not leaving enough to repay their credit card debt. On average, the value of their liquid assets could cover only about 60 percent of their unpaid debt plus monthly bills. In almost every category of assets or debts, both housing and non-housing related, borrower-savers were significantly worse off financially than savers. Thus, the differences between borrower-savers and savers are much broader than just their credit card debt and bank account balances; they extend to mortgage debt and home equity. Even when we control for income and demographics in a regression, we find that carrying a mortgage or other debt (such as auto or educational loans) is associated with a higher probability of revolving on a credit card, suggesting that various types of household debt are complements rather than substitutes. During the COVID-19 pandemic in 2020, consumers’ unpaid credit card debt decreased and their liquid assets increased, so the fraction of borrower-savers dropped to 35 percent of the sample. |
| Keywords: | credit card debt; consumer payments; liquid assets; COVID-19 |
| JEL: | D12 D14 E42 |
| Date: | 2022–06–01 |
| URL: | http://d.repec.org/n?u=RePEc:fip:fedbwp:94585&r= |
| By: | Nisrine Essanoussi (E.N.C.G - Ecole nationale de commerce et de gestion - USMBA - Université Sidi Mohamed Ben Abdellah); Zineb Bennis Nechba (USMBA - Université Sidi Mohamed Ben Abdellah) |
| Abstract: | Online payment represents one of the innovative ways of technology allowing consumers to benefit from financial services to pay for their purchases in less time and often with lower costs. This method of payment, which is carried out through digitalized applications, was implemented in response to the requirements and expectations of the modern consumer. In Morocco, banks have followed this evolution and are constantly developing their online payment services. Our work, which is based on the technology acceptance model, uses a questionnaire survey, administered online, on the acceptance of online payment by the effective customers of Moroccan banks. We study the following variables: perceived usefulness, perceived ease of use, security, privacy and use of the payment during the COVID 19. We use IBM® SPSS® 21 software and a logistic regression model to test our research hypotheses and answer our problem. Over a period of May 2020 to March 2021, our data collection resulted in 267 responses. After eliminating individuals who never opted for online payment, as well as responses with missing values, we retained a total of 201 valid and accepted questionnaires. Our study indicated that the context of the health crisis has reconciled many Moroccan consumers with online payment. Nevertheless, it should also be noted that the majority of consumers for whom the frequency of online payment increased during the health crisis and lockdown period already had a good acceptance of online payment. |
| Abstract: | Le paiement en ligne représente l'un des moyens innovants de la technologie permettant aux consommateurs de bénéficier des services financiers pour régler leurs achats en moins de temps et souvent avec des coûts plus bas. Ce mode de paiement qui s'effectue à travers des applications digitalisées a été mis en place en réponse aux exigences et attentes du consommateur moderne. Au Maroc, les banques ont abouti à cette évolution et ne cessent de développer leurs prestations du paiement en ligne. Notre travail, qui s'appuie sur le modèle d'acceptation de la technologie, fait recours à une enquête par questionnaire, administré en ligne, et portant sur l'acceptation du paiement en ligne par la clientèle effective des banques marocaines. Nous étudions, dans ce sens, les variables suivantes : l'utilité perçue, la facilité d'utilisation perçue, la sécurité, la confidentialité ainsi que l'utilisation du paiement lors de la COVID 19. Nous utilisons le logiciel IBM® SPSS® 21 et empruntons un modèle de régression logistique, afin de tester nos hypothèses de recherche et de répondre à notre problématique. Sur une période allant de Mai 2020 à Mars 2021, notre collecte des données a abouti à 267 réponses. Après avoir éliminé les individus n'ayant jamais opté pour le paiement en ligne, ainsi que les réponses avec des valeurs manquantes, nous avons retenu un total de 201 questionnaires valides et acceptés. Notre étude a indiqué que le contexte de la crise sanitaire a réconcilié d'innombrables consommateurs marocains avec le paiement en ligne. Néanmoins, il convient de souligner également que la majorité des consommateurs pour lesquels la fréquence du paiement en ligne a augmenté, pendant la période de la crise sanitaire et du confinement, avait déjà une bonne acceptabilité du paiement en ligne. |
| Keywords: | TAM,e-payment,Moroccan consumers,COVID 19,logistic regression,e-paiement,consommateurs marocains,régression logistique |
| Date: | 2022 |
| URL: | http://d.repec.org/n?u=RePEc:hal:journl:halshs-03705130&r= |
| By: | Ozili, Peterson Kitakogelu |
| Abstract: | This paper analyse the level of financial inclusion in Nigeria using data from the global findex indicators. The findings reveal that Nigeria witnessed growth in several financial inclusion indicators in the early years of financial inclusion in 2014 but the benefits were not sustained in the later years especially in 2017. Nigeria’s level of financial inclusion is very low compared to the World average. In the population group analysis, it was observed that the female, poorest, male, older and uneducated population were worse-off in all indicators of financial inclusion in 2017. The implication of the observed decline in the level of financial inclusion in 2017 suggest that there are barriers to financial inclusion in the post-2014 years. |
| Keywords: | formal account, borrowing, sustainable development, Nigeria, financial inclusion, access to finance, financial institutions, credit cards, debit cards, account ownership, female, savings. |
| JEL: | G21 G23 G28 G29 |
| Date: | 2022 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:113572&r= |
| By: | Dante Donati |
| Abstract: | Asymmetric information can distort market outcomes. I study how the online disclosure of information affects consumers’ behavior and firms’ incentives to upgrade product quality in markets where information is traditionally limited. I first build a model of consumer search with firms’ endogenous quality decisions. In this model, lower search costs reallocate demand toward higher-quality producers, raising firms’ incentives to upgrade quality, and more so for firms selling ex-ante lower-quality products. I then use the access to online reviews to proxy for reductions in consumers’ search costs and estimate its impact on the restaurant industry in Rome, exploiting the abolition of mobile roaming charges in the EU in 2017 for identification. Based on a unique dataset combining monthly information from Tripadvisor with administrative social-security records, I find that, after the policy, revenues and total employment in mid- and high-rating restaurants grow by 3-10%. In turn, the probability for low-rating restaurants to exit the market doubles compared to the pre-policy period, while surviving low- and mid-rating establishments hire workers with higher wages and better curricula, eventually improving their Tripadvisor ratings. Overall, the share of low-rating restaurants in the most tourist areas decreases by 2.5 pp. My findings have implications for the role of review platforms in the performance of offline industries under asymmetric information. |
| Keywords: | review platforms, asymmetric information, search costs, service industry, quality |
| JEL: | D82 D83 L15 L80 |
| Date: | 2022 |
| URL: | http://d.repec.org/n?u=RePEc:ces:ceswps:_9834&r= |
| By: | Glowka, Marc; Müller, Alexander; Polo Friz, Livia; Testi, Sara; Valentini, Massimo; Vespucci, Stefano |
| Abstract: | As the operator of a systemically important payment system (SIPS), the Eurosystem has the responsibility of regularly assessing the resilience of the Trans-European Automated Real-time Gross Settlement Express Transfer System (TARGET2) to various types of risks, as set out in the Principles for Financial Market Infrastructures (PFMIs) drawn up by the Committee on Payments and Market Infrastructures (CPMI) and International Organization of Securities Commissions (IOSCO). To identify, measure, monitor and mitigate these risks over time, the TARGET2 operator has developed specific approaches that include both qualitative and quantitative elements. JEL Classification: G20, E42, E58, C10, C63 |
| Keywords: | FMIs, payment systems, PFMIs, TARGET2 |
| Date: | 2022–08 |
| URL: | http://d.repec.org/n?u=RePEc:ecb:ecbops:2022300&r= |
| By: | Erica M. Field; Natalia Rigol; Charity M. Troyer Moore; Rohini Pande; Simone G. Schaner |
| Abstract: | Do information frictions limit the benefits of financial inclusion drives for the rural poor? We evaluate an experimental intervention among recently banked poor Indian women receiving government cash transfers via direct deposit. Treated women were provided automated voice calls confirming details of transactions posted to their accounts. The intervention increased women's knowledge of account balances and trust in their local banking agent. Indicative of improved consumption-smoothing by income-constrained women, administrative data show that treated women accessed government transfers faster when the service was active, with treatment effects dissipating after the notifications were discontinued. On average, other aspects of account use remained unchanged. However, consistent with account information benefiting those with high transaction costs more, the intervention increased account use among women who lived more than an hour from the kiosk. |
| JEL: | G21 O12 |
| Date: | 2022–07 |
| URL: | http://d.repec.org/n?u=RePEc:nbr:nberwo:30289&r= |
| By: | Buda, G.; Carvalho, V. M.; Hansen, S.; Mora, J. V. R.; Ortiz, Ã .; Rodrigo, T. |
| Abstract: | This paper provides the first proof of concept that naturally occurring transaction data, arising from the decentralized activity of millions of economic agents, can be harnessed to produce national accounts-like objects. We deploy comprehensive transaction-level data and its associated metadata arising from the universe of Spanish retail accounts of Banco Bilbao Vizcaya Argentaria (BBVA). We organize the resulting 3 billion individual transactions by 1.8 million bank customers in a large and highly detailed representative consumption panel to show (i) that the aggregation of such data, once organized according to national accounting principles, can reproduce current official statistics on aggregate consumption in the national accounts with a high degree of precision and, as a result of the richness of transaction data, (ii) produce novel, highly detailed distributional accounts for consumption. Finally, exploiting the panel nature of our data, we (iii) offer a non-parametric analysis of individual consumption dynamics across the consumption distribution. |
| Date: | 2022–07–20 |
| URL: | http://d.repec.org/n?u=RePEc:cam:camjip:2220&r= |
| By: | International Monetary Fund |
| Abstract: | Cybersecurity risk continues to grow both in complexity and severity and is a function of an increasingly open and interconnected cyber and financial ecosystem. The South African financial system has a long history of incorporating technology and as for many financial systems across the globe, digitalization has become a strategic priority. For risk management to keep pace with the dynamic nature of cyber threats and threat agents, systemically important financial institutions (SIFIs) have made substantial investments in cyber resilience programs (e.g., establishing cyber strategies, frameworks, and governance structures). Consistent with many jurisdictions, and partly a result of widespread remote working arrangements implemented in response to the global pandemic, cybersecurity threats to financial stability increased. However, high standards of risk management meant threats did not materialize into significant losses and/or disruptions. |
| Date: | 2022–06–17 |
| URL: | http://d.repec.org/n?u=RePEc:imf:imfscr:2022/181&r= |
| By: | Harald Uhlig |
| Abstract: | After remaining close to 1 US Dollar since its inception in November 2020, the algorithmic stablecoin UST crashed in the two weeks of May 9th to May 15th, 2022, leading to a price collapse of the underlying LUNA token and the erasure of more than 50 Billion U.S. Dollar or 90% in market value. I provide a novel theory to account for these phenomena and use it to shed light on the data. I break new ground methodologically by showing how crashes unfold gradually, and by introducing the method of quantitative interpretation. To obtain a gradual unfolding of the crash, I allow for the possibility that the market might return to normal at any moment. Suspension of convertibility happens, once the price has fallen sufficiently far. Agents price LUNA, taking into account these probabilities as well as the ongoing inflow from burning UST coins. Agents sell their UST coins when the probability of an eventual suspension of convertibility exceeds some agent-specific threshold. The implications of the theory are highlighted in an analytically tractable example. The theory is then used as a guide to examine and interpret the data, using bi-hourly observations. I use the observed data to quantify theory variables and use them in turn to interpret the data. I find that the majority of the UST coin holders waited until the probability of suspension was rather high, before deciding to burn their holdings. |
| JEL: | E41 F31 F32 G01 G12 G23 |
| Date: | 2022–07 |
| URL: | http://d.repec.org/n?u=RePEc:nbr:nberwo:30256&r= |
| By: | Hannu Laurila |
| Abstract: | Cash money can be a rational devise of saving as an insurance against external uncertainty. Liquid money, controlled by a stable and trustworthy central bank, offers an insurance against stock market crashes, bankrupts and other economic turmoils that endanger the yield of illiquid saving modes. In turbulent times, the value-carrying property of money is accentuated, and the recent dark episodes including the last financial crisis, the pandemic and the war in Ukraine have made the public in Europe respond to uncertainty by increasing their cash holdings. The paper constructs a simple life cycle framework for the analysis of rational and irrational motives to save money, answers to questions about the effects of saving liquid money on illiquid saving and education and examines the inherent cost of the use of cash as a saving mode. The main findings of the paper are the following. The insurance motive to save money increases total savings by replacing deposit saving more than one-to-one. The share of deposit savings depends positively on the expected interest rate, while the share of cash savings is the higher the less there is inflation. Deposit saving correlates positively and education negatively with the expected market interest rate thus affecting their relative proportion, but education does not affect the implicit price paid for cash insurance. Incorporating money illusion adds an internal bias to life-time optimization. Misjudgment of the inflation rate makes consumers save excessively in cash at the cost of market deposits and increases the cost of using cash as rational insurance against external uncertainty. |
| Date: | 2022–08 |
| URL: | http://d.repec.org/n?u=RePEc:tam:wpaper:2235&r= |
| By: | Léa Fabre (LAET - Laboratoire Aménagement Économie Transports - UL2 - Université Lumière - Lyon 2 - ENTPE - École Nationale des Travaux Publics de l'État - CNRS - Centre National de la Recherche Scientifique, SAF - Laboratoire de Sciences Actuarielle et Financière - UCBL - Université Claude Bernard Lyon 1 - Université de Lyon, Explain); Caroline Bayart (SAF - Laboratoire de Sciences Actuarielle et Financière - UCBL - Université Claude Bernard Lyon 1 - Université de Lyon); Patrick Bonnel (LAET - Laboratoire Aménagement Économie Transports - UL2 - Université Lumière - Lyon 2 - ENTPE - École Nationale des Travaux Publics de l'État - CNRS - Centre National de la Recherche Scientifique); Nicolas Mony (Explain) |
| Abstract: | Using technologies such as Wi-Fi and Bluetooth allows to gather passive mobility data, useful for ensuring the sustainable development of transport infrastructures. The challenge of passive data collection is to be able to identify relevant data. Our research presents interesting solutions for sorting the transmitted signals and reconstructing quality Origin-Destination matrices. Its originality consists not only in comparing the results with those of other data sources, but also in proposing a methodology that can be reproduced. Thanks to a partitioning algorithm, it is possible to automatically distinguish passengers from non-passengers to get transit ridership flow and O-D matrices. The findings show that this algorithm provides concrete and replicable solutions to transport operators for understanding travel demand. |
| Keywords: | Travel behavior,passive tracking,data clustering,Wi-Fi/Bluetooth sensors,trajectory reconstruction,mobile devices data,data quality,Working Papers du LAET |
| Date: | 2022 |
| URL: | http://d.repec.org/n?u=RePEc:hal:wpaper:halshs-03721297&r= |
| By: | Muriuki, James M.; Badruddoza, Syed; Fuad, Syed M. |
| Keywords: | Risk and Uncertainty, Institutional and Behavioral Economics, Agribusiness |
| Date: | 2022–08 |
| URL: | http://d.repec.org/n?u=RePEc:ags:aaea22:322533&r= |
| By: | Andrés Alonso (Banco de España); José Manuel Carbó (Banco de España) |
| Abstract: | One of the biggest challenges for the application of machine learning (ML) models in finance is how to explain their results. In recent years, innovative interpretability techniques have appeared to assist in this task, although their usefulness is still a matter of debate within the industry. In this article we propose a novel framework to assess how accurate these techniques are. Our work is based on the generation of synthetic datasets. This allows us to define the importance of the variables, so we can calculate to what extent the explanations given by these techniques match the ground truth of our data. We perform an empirical exercise in which we apply two non-interpretable ML models (XGBoost and Deep Learning) to the synthetic datasets, , and then we explain their results using two popular interpretability techniques, SHAP and permutation Feature Importance (FI). We conclude that generating synthetic datasets shows potential as a useful approach for supervisors and practitioners who wish to assess interpretability techniques. |
| Keywords: | synthetic datasets, artificial intelligence, interpretability, machine learning, credit assessment |
| JEL: | C55 C63 G17 |
| Date: | 2022–06 |
| URL: | http://d.repec.org/n?u=RePEc:bde:wpaper:2222&r= |