Kay Giesecke
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Kay Giesecke has a total of 28 co-authors. Here are the top co-authors by number of shared publications:
Justin Sirignano
Department of Industrial and Enterprise Systems Engineering, University of Illinois at Urbana–Champaign, Urbana, Illinois 61801
3 shared publications- Justin A. Sirignano2 shared publications
- Agostino Capponi
Columbia University
1 shared publication - Alex Shkolnik
Department of Statistics & Applied Probability University of California Santa Barbara Santa Barbara 93106 CA
1 shared publication - Apaar Sadhwani1 shared publication
- Bruno Biais
Area of Finance, HEC Paris, 78350 Jouy-en-Josas, France
1 shared publication
Latest Publications
- Open Access
- 23 September 2026
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 123 (39)
AICO: Feature significance tests for supervised learning
Machine learning is central to modern science, industry, and policy, yet its predictive power often comes at the cost of transparency: We rarely know which input features drive a model’s predictions. Without such understanding, researchers cannot draw reliable conclusions, practitioners cannot ensure fairness or accountability, and policymakers cannot trust or govern model-based decisions. Existing tools for assessing feature influence are limited; most lack statistical guarantees, and many require costly retraining or surrogate modeling, making them impractical for large modern models. We introduce AICO (Add-In COvariates), a broadly applicable framework that turns model interpretability into an efficient statistical exercise. AICO tests whether each feature contributes to predictive performance by masking its information and measuring the resulting change. The method provides exact, finite-sample feature P -values and CIs for feature importance through a simple, nonasymptotic hypothesis testing procedure. It requires no retraining, surrogate modeling, or distributional assumptions, making it feasible for large-scale algorithms. In both controlled experiments and real applications, from credit scoring to mortgage-behavior prediction, AICO identifies variables that contribute to model behavior, providing a scalable and statistically principled path toward transparent and trustworthy machine learning.
- 23 September 2026
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 123 (39)
Open AccessAICO: Feature significance tests for supervised learning
Machine learning is central to modern science, industry, and policy, yet its predictive power often comes at the cost of transparency: We rarely know which input features drive a model’s predictions. Without such understanding, researchers cannot draw reliable conclusions, practitioners cannot ensure fairness or accountability, and policymakers cannot trust or govern model-based decisions. Existing tools for assessing feature influence are limited; most lack statistical guarantees, and many require costly retraining or surrogate modeling, making them impractical for large modern models. We introduce AICO (Add-In COvariates), a broadly applicable framework that turns model interpretability into an efficient statistical exercise. AICO tests whether each feature contributes to predictive performance by masking its information and measuring the resulting change. The method provides exact, finite-sample feature P -values and CIs for feature importance through a simple, nonasymptotic hypothesis testing procedure. It requires no retraining, surrogate modeling, or distributional assumptions, making it feasible for large-scale algorithms. In both controlled experiments and real applications, from credit scoring to mortgage-behavior prediction, AICO identifies variables that contribute to model behavior, providing a scalable and statistically principled path toward transparent and trustworthy machine learning.
- 1 May 2022
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Mathematics of Operations Research
- Vol. 47 (2) , 969-988
Reducing Bias in Event Time Simulations via Measure Changes
Stochastic point process models of event timing are common in many areas, including finance, insurance, and reliability. Monte Carlo simulation is often used to perform computations for these models. The standard sampling algorithm, which is based on a time-change argument, is widely applicable but generates biased simulation estimators. This article develops and analyzes a change of probability measure that can reduce or even eliminate the bias without restricting the scope of the algorithm. A result of independent interest offers new conditions that guarantee the existence of a broad class of point process martingales inducing changes of measure. Numerical results illustrate our approach.
- 1 May 2022
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Mathematics of Operations Research
- Vol. 47 (2) , 969-988
Reducing Bias in Event Time Simulations via Measure Changes
Stochastic point process models of event timing are common in many areas, including finance, insurance, and reliability. Monte Carlo simulation is often used to perform computations for these models. The standard sampling algorithm, which is based on a time-change argument, is widely applicable but generates biased simulation estimators. This article develops and analyzes a change of probability measure that can reduce or even eliminate the bias without restricting the scope of the algorithm. A result of independent interest offers new conditions that guarantee the existence of a broad class of point process martingales inducing changes of measure. Numerical results illustrate our approach.
- 1 March 2022
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 68 (3) , 1591-1594
Introduction to the Special Section on Data-Driven Prescriptive Analytics
- 1 March 2022
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 68 (3) , 1591-1594
Introduction to the Special Section on Data-Driven Prescriptive Analytics
- 1 January 2021
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 67 (1) , 6-7
Call for Papers—Management Science Special Issue on Blockchains and Crypto Economics
- 1 January 2021
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 67 (1) , 6-7
Call for Papers—Management Science Special Issue on Blockchains and Crypto Economics
- Open Access
Significance Tests for Neural Networks
We develop a pivotal test to assess the statistical significance of the feature variables in a single-layer feedforward neural network regression model. We propose a gradient-based test statistic and study its asymptotics using nonparametric techniques. Under technical conditions, the limiting distribution is given by a mixture of chi-square distributions. The tests enable one to discern the impact of individual variables on the prediction of a neural network. The test statistic can be used to rank variables according to their influence. Simulation results illustrate the computational efficiency and the performance of the test. An empirical application to house price valuation highlights the behavior of the test using actual data.
Open AccessSignificance Tests for Neural Networks
We develop a pivotal test to assess the statistical significance of the feature variables in a single-layer feedforward neural network regression model. We propose a gradient-based test statistic and study its asymptotics using nonparametric techniques. Under technical conditions, the limiting distribution is given by a mixture of chi-square distributions. The tests enable one to discern the impact of individual variables on the prediction of a neural network. The test statistic can be used to rank variables according to their influence. Simulation results illustrate the computational efficiency and the performance of the test. An empirical application to house price valuation highlights the behavior of the test using actual data.
- 1 January 2019
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 65 (1) , 107-121
Risk Analysis for Large Pools of Loans
Financial institutions, government-sponsored enterprises, and asset-backed security investors are often exposed to delinquency and prepayment risk from large numbers of loans. Examples include mortgages, credit cards, and auto, student, and business loans. Because of the size of the pools, the measurement of these exposures is computationally expensive. This paper develops and tests efficient numerical methods for the analysis of large pools of loans as well as asset-backed securities backed by such pools. For a broad class of statistical and machine learning models of loan-level risk in discrete time, we prove a law of large numbers and a central limit theorem for the pool-level risk. The asymptotics are then used to construct computationally efficient approximations for a large pool. The approximations aggregate the full loan-level dynamics, making it possible to take advantage of the detailed loan-level data often available in practice. We furthermore prove a convergence rate for the approximation and a uniform integrability result. The latter allows the approximation to be applied to a wide class of pool-level risk statistics. To demonstrate the effectiveness of our approach, we implement it on a data set of over 25 million mortgages. The results show the accuracy and speed of the approximation in comparison to brute-force simulation of a pool, for a variety of pools with different risk profiles. The online appendix is available at https://doi.org/10.1287/mnsc.2017.2947 . This paper was accepted by Noah Gans, stochastic models.
- 1 January 2019
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 65 (1) , 107-121
Risk Analysis for Large Pools of Loans
Financial institutions, government-sponsored enterprises, and asset-backed security investors are often exposed to delinquency and prepayment risk from large numbers of loans. Examples include mortgages, credit cards, and auto, student, and business loans. Because of the size of the pools, the measurement of these exposures is computationally expensive. This paper develops and tests efficient numerical methods for the analysis of large pools of loans as well as asset-backed securities backed by such pools. For a broad class of statistical and machine learning models of loan-level risk in discrete time, we prove a law of large numbers and a central limit theorem for the pool-level risk. The asymptotics are then used to construct computationally efficient approximations for a large pool. The approximations aggregate the full loan-level dynamics, making it possible to take advantage of the detailed loan-level data often available in practice. We furthermore prove a convergence rate for the approximation and a uniform integrability result. The latter allows the approximation to be applied to a wide class of pool-level risk statistics. To demonstrate the effectiveness of our approach, we implement it on a data set of over 25 million mortgages. The results show the accuracy and speed of the approximation in comparison to brute-force simulation of a pool, for a variety of pools with different risk profiles. The online appendix is available at https://doi.org/10.1287/mnsc.2017.2947 . This paper was accepted by Noah Gans, stochastic models.
- Open Access
Sensitivity based Neural Networks Explanations
Although neural networks can achieve very high predictive performance on various different tasks such as image recognition or natural language processing, they are often considered as opaque "black boxes". The difficulty of interpreting the predictions of a neural network often prevents its use in fields where explainability is important, such as the financial industry where regulators and auditors often insist on this aspect. In this paper, we present a way to assess the relative input features importance of a neural network based on the sensitivity of the model output with respect to its input. This method has the advantage of being fast to compute, it can provide both global and local levels of explanations and is applicable for many types of neural network architectures. We illustrate the performance of this method on both synthetic and real data and compare it with other interpretation techniques. This method is implemented into an open-source Python package that allows its users to easily generate and visualize explanations for their neural networks.
Open AccessSensitivity based Neural Networks Explanations
Although neural networks can achieve very high predictive performance on various different tasks such as image recognition or natural language processing, they are often considered as opaque "black boxes". The difficulty of interpreting the predictions of a neural network often prevents its use in fields where explainability is important, such as the financial industry where regulators and auditors often insist on this aspect. In this paper, we present a way to assess the relative input features importance of a neural network based on the sensitivity of the model output with respect to its input. This method has the advantage of being fast to compute, it can provide both global and local levels of explanations and is applicable for many types of neural network architectures. We illustrate the performance of this method on both synthetic and real data and compare it with other interpretation techniques. This method is implemented into an open-source Python package that allows its users to easily generate and visualize explanations for their neural networks.
- Open Access
Deep Learning for Mortgage Risk
We develop a deep learning model of multi-period mortgage risk and use it to analyze an unprecedented dataset of origination and monthly performance records for over 120 million mortgages originated across the US between 1995 and 2014. Our estimators of term structures of conditional probabilities of prepayment, foreclosure and various states of delinquency incorporate the dynamics of a large number of loan-specific as well as macroeconomic variables down to the zip-code level. The estimators uncover the highly nonlinear nature of the relationship between the variables and borrower behavior, especially prepayment. They also highlight the effects of local economic conditions on borrower behavior. State unemployment has the greatest explanatory power among all variables, offering strong evidence of the tight connection between housing finance markets and the macroeconomy. The sensitivity of a borrower to changes in unemployment strongly depends upon current unemployment. It also significantly varies across the entire borrower population, which highlights the interaction of unemployment and many other variables. These findings have important implications for mortgage-backed security investors, rating agencies, and housing finance policymakers.
Open AccessDeep Learning for Mortgage Risk
We develop a deep learning model of multi-period mortgage risk and use it to analyze an unprecedented dataset of origination and monthly performance records for over 120 million mortgages originated across the US between 1995 and 2014. Our estimators of term structures of conditional probabilities of prepayment, foreclosure and various states of delinquency incorporate the dynamics of a large number of loan-specific as well as macroeconomic variables down to the zip-code level. The estimators uncover the highly nonlinear nature of the relationship between the variables and borrower behavior, especially prepayment. They also highlight the effects of local economic conditions on borrower behavior. State unemployment has the greatest explanatory power among all variables, offering strong evidence of the tight connection between housing finance markets and the macroeconomy. The sensitivity of a borrower to changes in unemployment strongly depends upon current unemployment. It also significantly varies across the entire borrower population, which highlights the interaction of unemployment and many other variables. These findings have important implications for mortgage-backed security investors, rating agencies, and housing finance policymakers.
- 27 October 2017
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 65 (5) , 2015-2040
Dynamic Portfolio Execution
We analyze the optimal execution problem of a portfolio manager trading multiple assets. In addition to the liquidity and risk of each individual asset, we consider cross asset interactions in these two dimensions, which substantially enriches the nature of the problem. Focusing on the market microstructure, we develop a tractable order book model to capture liquidity supply/demand dynamics in a multiasset setting, which allows us to formulate and solve the optimal portfolio execution problem. We find that cross asset risk and liquidity considerations are of critical importance in constructing the optimal execution policy. We show that even when the goal is to trade a single asset, its optimal execution may involve transitory trades in other assets. In general, optimally managing the risk of the portfolio during the execution process affects the time synchronization of trading in different assets. Moreover, links in the liquidity across assets lead to complex patterns in the optimal execution policy. In p...
- 27 October 2017
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 65 (5) , 2015-2040
Dynamic Portfolio Execution
We analyze the optimal execution problem of a portfolio manager trading multiple assets. In addition to the liquidity and risk of each individual asset, we consider cross asset interactions in these two dimensions, which substantially enriches the nature of the problem. Focusing on the market microstructure, we develop a tractable order book model to capture liquidity supply/demand dynamics in a multiasset setting, which allows us to formulate and solve the optimal portfolio execution problem. We find that cross asset risk and liquidity considerations are of critical importance in constructing the optimal execution policy. We show that even when the goal is to trade a single asset, its optimal execution may involve transitory trades in other assets. In general, optimally managing the risk of the portfolio during the execution process affects the time synchronization of trading in different assets. Moreover, links in the liquidity across assets lead to complex patterns in the optimal execution policy. In p...
- Open Access
Fluctuation Analysis for the Loss From Default
We analyze the fluctuation of the loss from default around its large portfolio limit in a class of reduced-form models of correlated firm-by-firm default timing. We prove a weak convergence result for the fluctuation process and use it for developing a conditionally Gaussian approximation to the loss distribution. Numerical results illustrate the accuracy and computational efficiency of the approximation.
Open AccessFluctuation Analysis for the Loss From Default
We analyze the fluctuation of the loss from default around its large portfolio limit in a class of reduced-form models of correlated firm-by-firm default timing. We prove a weak convergence result for the fluctuation process and use it for developing a conditionally Gaussian approximation to the loss distribution. Numerical results illustrate the accuracy and computational efficiency of the approximation.