nep-cmp New Economics Papers
on Computational Economics
Issue of 2022‒06‒13
eight papers chosen by
Stan Miles
Thompson Rivers University

  1. Machine Learning based Framework for Robust Price-Sensitivity Estimation with Application to Airline Pricing By Ravi Kumar; Shahin Boluki; Karl Isler; Jonas Rauch; Darius Walczak
  2. NFT Appraisal Prediction: Utilizing Search Trends, Public Market Data, Linear Regression and Recurrent Neural Networks By Shrey Jain; Camille Bruckmann; Chase McDougall
  3. Acquisition of Costly Information in Data-Driven Decision Making By Lukas Janasek
  4. AI Adoption in a Competitive Market By Joshua S. Gans
  5. Creating USAGE-OCC: a CGE model of the U.S. with a disaggregated occupational dimension By Peter B. Dixon; Maureen T. Rimmer
  6. Preference Restrictions in Computational Social Choice: A Survey By Edith Elkind; Martin Lackner; Dominik Peters
  7. Hot off the press: News-implied sovereign default risk By Dim, Chukwuma; Koerner, Kevin; Wolski, Marcin; Zwart, Sanne
  8. AI Adoption in a Monopoly Market By Joshua S. Gans

  1. By: Ravi Kumar; Shahin Boluki; Karl Isler; Jonas Rauch; Darius Walczak
    Abstract: We consider the problem of dynamic pricing of a product in the presence of feature-dependent price sensitivity. Based on the Poisson semi-parametric approach, we construct a flexible yet interpretable demand model where the price related part is parametric while the remaining (nuisance) part of the model is non-parametric and can be modeled via sophisticated ML techniques. The estimation of price-sensitivity parameters of this model via direct one-stage regression techniques may lead to biased estimates. We propose a two-stage estimation methodology which makes the estimation of the price-sensitivity parameters robust to biases in the nuisance parameters of the model. In the first-stage we construct the estimators of observed purchases and price given the feature vector using sophisticated ML estimators like deep neural networks. Utilizing the estimators from the first-stage, in the second-stage we leverage a Bayesian dynamic generalized linear model to estimate the price-sensitivity parameters. We test the performance of the proposed estimation schemes on simulated and real sales transaction data from Airline industry. Our numerical studies demonstrate that the two-stage approach provides more accurate estimates of price-sensitivity parameters as compared to direct one-stage approach.
    Date: 2022–05
  2. By: Shrey Jain; Camille Bruckmann; Chase McDougall
    Abstract: In this paper we investigate the correlation between NFT valuations and various features from three primary categories: public market data, NFT metadata, and social trends data.
    Date: 2022–04
  3. By: Lukas Janasek (Institute of Economic Studies, Charles University & Institute of Information Theory and Automation, Czech Academy of Sciences, Prague, Czech Republic)
    Abstract: This paper formulates and solves an economic decision problem of the acquisition of costly information in data-driven decision making. The paper assumes an agent predicting a random variable utilizing several costly explanatory variables. Prior to the decision making, the agent learns about the relationship between the random variables utilizing its past realizations. During the decision making, the agent decides what costly variables to acquire and predicts using the acquired variables. The agent´s utility consists of the correctness of the prediction and the costs of the acquired variables. To solve the decision problem, we split the decision process into two parts: acquisition of variables and prediction using the acquired variables. For the prediction, we propose an approach for training a single predictive model accepting any combination of acquired variables. For the acquisition, we propose two methods using supervised machine learning models: a backward estimation of the expected utility of each variable and a greedy acquisition of variables based on a myopic estimate of the expected utility. We evaluate the methods on two medical datasets. The results show that the methods acquire the costly variables efficiently.
    Keywords: costly information, data-driven decision-making, machine learning
    JEL: C44 C45 C52 C73 D81 D83
    Date: 2022–05
  4. By: Joshua S. Gans
    Abstract: Economists have often viewed the adoption of artificial intelligence (AI) as a standard process innovation where we expect that efficiency will drive adoption in competitive markets. This paper models AI based on recent advances in machine learning that allow firms to engage in better prediction. Using prediction of demand, it is demonstrated that AI adoption is a complement to variable inputs whose levels are directly altered by predictions and use is economised by them (that is, labour). It is shown that, in a competitive market, this increases the short-run elasticity of supply and may or may not increase average equilibrium prices. There are generically externalities in adoption with this reducing the profits of non-adoptees when variable inputs are important and increasing them otherwise. Thus, AI does not operate as a standard process innovation and its adoption may confer positive externalities on non-adopting firms. In the long-run, AI adoption is shown to generally lower prices and raise consumer surplus in competitive markets.
    JEL: D21 D41 D81 O31
    Date: 2022–04
  5. By: Peter B. Dixon; Maureen T. Rimmer
    Abstract: We created the USAGE-OCC model of the U.S. by adding to USAGE occupation-industry matrices for 2019 that identify numbers of people employed and wagebills in 233 occupations (aggregated from 789 6-digit BLS occupations) and 392 industries (BEA input-output). The aggregation from 789 to 233 occupations was performed in a way that minimized the loss of skill/experience/education detail. In specifying occupational mobility, we took account of: wage differences between occupations; physical requirements of occupations; and education/training/experience requirements. As well as providing detailed occupational projections, USAGE-OCC can generate results for employment by wage band, educational requirements and experience. In an illustrative application, we simulated the effects of a mandated 10 per cent increase in real wage rates in low-wage occupations. The results point to the idea that rectifying inequitable wage disparities without adverse employment effects requires policies such as negative tax rates that raise incomes for low-wage workers without increasing costs to employers.
    Keywords: Employment by occupation and industry, Occupational mobility, Labor-market modeling, Wage increases in low-wage occupations
    JEL: J62 J24 J31 C68
    Date: 2022–05
  6. By: Edith Elkind; Martin Lackner; Dominik Peters
    Abstract: Social choice becomes easier on restricted preference domains such as single-peaked, single-crossing, and Euclidean preferences. Many impossibility theorems disappear, the structure makes it easier to reason about preferences, and computational problems can be solved more efficiently. In this survey, we give a thorough overview of many classic and modern restricted preference domains and explore their properties and applications. We do this from the viewpoint of computational social choice, letting computational problems drive our interest, but we include a comprehensive discussion of the economics and social choice literatures as well. Particular focus areas of our survey include algorithms for recognizing whether preferences belong to a particular preference domain, and algorithms for winner determination of voting rules that are hard to compute if preferences are unrestricted.
    Date: 2022–05
  7. By: Dim, Chukwuma; Koerner, Kevin; Wolski, Marcin; Zwart, Sanne
    Abstract: We develop a sovereign default risk index using natural language processing techniques and 10 million news articles covering over 100 countries. The index is a highfrequency measure of countries' default risk, particularly for those lacking marketbased measures: it correlates with sovereign CDS spreads, predicts rating downgrades, and reflects default risk information not fully captured by CDS spreads. We assess the influence of sovereign default concerns on equity markets and find that spikes in the index are negatively associated with same-week market returns, which reverses over the next week, indicating that investors might overreact to default concerns. Equity markets' reaction to default concerns is more pronounced and persistent for countries with tight fiscal constraints. The response to global, compared to country-specific, default concerns is much stronger, underlining the relevance of global "push" factors for local asset prices.
    Keywords: Sovereign default,Credit risk,Equity returns,Machine learning,Naturallanguage processing,Early warning indicators
    JEL: F30 G12 G15
    Date: 2022
  8. By: Joshua S. Gans
    Abstract: The adoption of artificial intelligence (AI) prediction of demand by a monopolist firm is examined. It is shown that, in the absence of AI prediction, firms face complex trade-offs in setting price and quantity ahead of demand that impact on the returns of AI adoption. Different industrial environments with differing flexibility of prices and/or quantity ex post, also impact on AI returns as does the time horizon of AI prediction. While AI has positive benefits for firms in terms of profitability, its impact on average price and quantity, as well as consumer welfare, is more nuanced and critically dependent on environmental characteristics.
    JEL: D21 D81 O31
    Date: 2022–04

This nep-cmp issue is ©2022 by Stan Miles. It is provided as is without any express or implied warranty. It may be freely redistributed in whole or in part for any purpose. If distributed in part, please include this notice.
General information on the NEP project can be found at For comments please write to the director of NEP, Marco Novarese at <>. Put “NEP” in the subject, otherwise your mail may be rejected.
NEP’s infrastructure is sponsored by the School of Economics and Finance of Massey University in New Zealand.