Panos Toulis
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
Panos Toulis has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Tengyuan Liang
Booth School of Business University of Chicago Chicago Illinois USA
2 shared publications - Wenxuan Guo
Booth School of Business, The University of Chicago , Chicago, IL 60637
2 shared publications - Avi Feller
University of California, Berkeley, CA, USA
1 shared publication - David Puelz
Seminar participants Booth School of Business conference participants at Polmeth Advances with Field Experiments Design and Analysis of Experiments Atlantic Causal Inference University of Chicago The University of Chicago
1 shared publication - Dionysios Kehagias
Centre for Research and Technology Hellas, Thessaloniki, Greece
1 shared publication - Guillaume Basse
Department of Management Science and Engineering, Stanford University, Stanford, CA
1 shared publication
Latest Publications
- 22 April 2026
- Published by Oxford University Press (OUP) in Journal of the Royal Statistical Society Series B: Statistical Methodology
- Vol. 88 (5) , 1682-1703
Gaussianized design optimization for covariate balance in randomized experiments
Achieving covariate balance in randomized experiments enhances the precision of treatment effect estimation. However, existing methods often require heuristic adjustments based on domain knowledge and are primarily developed for binary treatments. This paper presents Gaussianized Design Optimization, a novel framework for optimally balancing covariates in experimental design. The core idea is to Gaussianize the treatment assignments: we model treatments as transformations of random variables drawn from a multivariate Gaussian distribution and convert the design problem into a nonlinear continuous optimization over Gaussian covariance matrices. Compared to existing methods, our approach offers significant flexibility in optimizing covariate balance across a diverse range of designs and covariate types. Adapting the Burer–Monteiro approach for solving semidefinite programmes, we introduce first-order local algorithms for optimizing covariate balance, improving upon several widely used designs. Furthermore, we develop inferential procedures for constructing design-based confidence intervals under Gaussianization and extend the framework to accommodate continuous treatments. Simulations demonstrate the effectiveness of Gaussianization in multiple practical scenarios.
- 22 April 2026
- Published by Oxford University Press (OUP) in Journal of the Royal Statistical Society Series B: Statistical Methodology
- Vol. 88 (5) , 1682-1703
Gaussianized design optimization for covariate balance in randomized experiments
Achieving covariate balance in randomized experiments enhances the precision of treatment effect estimation. However, existing methods often require heuristic adjustments based on domain knowledge and are primarily developed for binary treatments. This paper presents Gaussianized Design Optimization, a novel framework for optimally balancing covariates in experimental design. The core idea is to Gaussianize the treatment assignments: we model treatments as transformations of random variables drawn from a multivariate Gaussian distribution and convert the design problem into a nonlinear continuous optimization over Gaussian covariance matrices. Compared to existing methods, our approach offers significant flexibility in optimizing covariate balance across a diverse range of designs and covariate types. Adapting the Burer–Monteiro approach for solving semidefinite programmes, we introduce first-order local algorithms for optimizing covariate balance, improving upon several widely used designs. Furthermore, we develop inferential procedures for constructing design-based confidence intervals under Gaussianization and extend the framework to accommodate continuous treatments. Simulations demonstrate the effectiveness of Gaussianization in multiple practical scenarios.
- 26 November 2025
- Published by Oxford University Press (OUP) in Biometrika
- Vol. 113 (2)
Asymptotic validity and finite-sample properties of approximate randomization tests
Summary: Randomization tests rely on simple data transformations and possess an appealing robustness property. In addition to being finite-sample valid if the data distribution is invariant under the transformation, these tests can be asymptotically valid under a suitable studentization of the test statistic, even if the invariance does not hold. However, practical implementation often encounters noisy data, resulting in approximate randomization tests that may not be as robust. In this paper, one key theoretical contribution is a nonasymptotic bound on the discrepancy between the size of an approximate randomization test and the size of the idealized randomization test using noiseless data. This allows us to derive novel conditions for the validity of approximate randomization tests under data invariances, while being able to use existing results based on studentization if the invariance does not hold. We illustrate our theory through several examples, including significance tests in linear regression. These examples clarify key aspects of how randomization tests behave in small samples and address limitations of prior theoretical results.
- 26 November 2025
- Published by Oxford University Press (OUP) in Biometrika
- Vol. 113 (2)
Asymptotic validity and finite-sample properties of approximate randomization tests
Summary: Randomization tests rely on simple data transformations and possess an appealing robustness property. In addition to being finite-sample valid if the data distribution is invariant under the transformation, these tests can be asymptotically valid under a suitable studentization of the test statistic, even if the invariance does not hold. However, practical implementation often encounters noisy data, resulting in approximate randomization tests that may not be as robust. In this paper, one key theoretical contribution is a nonasymptotic bound on the discrepancy between the size of an approximate randomization test and the size of the idealized randomization test using noiseless data. This allows us to derive novel conditions for the validity of approximate randomization tests under data invariances, while being able to use existing results based on studentization if the invariance does not hold. We illustrate our theory through several examples, including significance tests in linear regression. These examples clarify key aspects of how randomization tests behave in small samples and address limitations of prior theoretical results.
- Open Access
- 2 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 918
Gaussianized Design Optimization for Covariate Balance in Randomized Experiments
- 2 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 918
Open AccessGaussianized Design Optimization for Covariate Balance in Randomized Experiments
- Open Access
Randomization Inference of Periodicity in Unequally Spaced Time Series with Application to Exoplanet Detection
The estimation of periodicity is a fundamental task in many scientific areas of study. Existing methods rely on theoretical assumptions that the observation times have equal or i.i.d. spacings, and that common estimators, such as the periodogram peak, are consistent and asymptotically normal. In practice, however, these assumptions are unrealistic as observation times usually exhibit deterministic patterns -- e.g., the nightly observation cycle in astronomy -- that imprint nuisance periodicities in the data. These nuisance signals also affect the finite-sample distribution of estimators, which can substantially deviate from normality. Here, we propose a set identification method, fusing ideas from randomization inference and partial identification. In particular, we develop a sharp test for any periodicity value, and then invert the test to build a confidence set. This approach is appropriate here because the construction of confidence sets does not rely on assumptions of regular or well-behaved asymptotics. Notably, our inference is valid in finite samples when our method is fully implemented, while it can be asymptotically valid under an approximate implementation designed to ease computation. Empirically, we validate our method in exoplanet detection using radial velocity data. In this context, our method correctly identifies the periodicity of the confirmed exoplanets in our sample. For some other, yet unconfirmed detections, we show that the statistical evidence is weak, which illustrates the failure of traditional statistical techniques. Last but not least, our method offers a constructive way to resolve these identification issues via improved observation designs. In exoplanet detection, these designs suggest meaningful improvements in identifying periodicity even when a moderate amount of randomization is introduced in scheduling radial velocity measurements.
Open AccessRandomization Inference of Periodicity in Unequally Spaced Time Series with Application to Exoplanet Detection
The estimation of periodicity is a fundamental task in many scientific areas of study. Existing methods rely on theoretical assumptions that the observation times have equal or i.i.d. spacings, and that common estimators, such as the periodogram peak, are consistent and asymptotically normal. In practice, however, these assumptions are unrealistic as observation times usually exhibit deterministic patterns -- e.g., the nightly observation cycle in astronomy -- that imprint nuisance periodicities in the data. These nuisance signals also affect the finite-sample distribution of estimators, which can substantially deviate from normality. Here, we propose a set identification method, fusing ideas from randomization inference and partial identification. In particular, we develop a sharp test for any periodicity value, and then invert the test to build a confidence set. This approach is appropriate here because the construction of confidence sets does not rely on assumptions of regular or well-behaved asymptotics. Notably, our inference is valid in finite samples when our method is fully implemented, while it can be asymptotically valid under an approximate implementation designed to ease computation. Empirically, we validate our method in exoplanet detection using radial velocity data. In this context, our method correctly identifies the periodicity of the confirmed exoplanets in our sample. For some other, yet unconfirmed detections, we show that the statistical evidence is weak, which illustrates the failure of traditional statistical techniques. Last but not least, our method offers a constructive way to resolve these identification issues via improved observation designs. In exoplanet detection, these designs suggest meaningful improvements in identifying periodicity even when a moderate amount of randomization is introduced in scheduling radial velocity measurements.
- Open Access
A Graph-Theoretic Approach to Randomization Tests of Causal Effects Under General Interference
Interference exists when a unit's outcome depends on another unit's treatment assignment. For example, intensive policing on one street could have a spillover effect on neighboring streets. Classical randomization tests typically break down in this setting because many null hypotheses of interest are no longer sharp under interference. A promising alternative is to instead construct a conditional randomization test on a subset of units and assignments for which a given null hypothesis is sharp. Finding these subsets is challenging, however, and existing methods are limited to special cases or have limited power. In this paper, we propose valid and easy-to-implement randomization tests for a general class of null hypotheses under arbitrary interference between units. Our key idea is to represent the hypothesis of interest as a bipartite graph between units and assignments, and to find an appropriate biclique of this graph. Importantly, the null hypothesis is sharp within this biclique, enabling conditional randomization-based tests. We also connect the size of the biclique to statistical power. Moreover, we can apply off-the-shelf graph clustering methods to find such bicliques efficiently and at scale. We illustrate our approach in settings with clustered interference and show advantages over methods designed specifically for that setting. We then apply our method to a large-scale policing experiment in Medellin, Colombia, where interference has a spatial structure.
Open AccessA Graph-Theoretic Approach to Randomization Tests of Causal Effects Under General Interference
Interference exists when a unit's outcome depends on another unit's treatment assignment. For example, intensive policing on one street could have a spillover effect on neighboring streets. Classical randomization tests typically break down in this setting because many null hypotheses of interest are no longer sharp under interference. A promising alternative is to instead construct a conditional randomization test on a subset of units and assignments for which a given null hypothesis is sharp. Finding these subsets is challenging, however, and existing methods are limited to special cases or have limited power. In this paper, we propose valid and easy-to-implement randomization tests for a general class of null hypotheses under arbitrary interference between units. Our key idea is to represent the hypothesis of interest as a bipartite graph between units and assignments, and to find an appropriate biclique of this graph. Importantly, the null hypothesis is sharp within this biclique, enabling conditional randomization-based tests. We also connect the size of the biclique to statistical power. Moreover, we can apply off-the-shelf graph clustering methods to find such bicliques efficiently and at scale. We illustrate our approach in settings with clustered interference and show advantages over methods designed specifically for that setting. We then apply our method to a large-scale policing experiment in Medellin, Colombia, where interference has a spatial structure.
- Open Access
- 8 May 2006
- Published by Association for Computing Machinery (ACM)
- Vol. 20, 1191-1198
Mertacor
In this paper we present the internal architecture and bidding mechanisms designed for Mertacor, a successful trading agent, which ended up first in the Classic Trading Agent Competition (TAC) of 2005. TAC provides a realistic benchmarking environment in which different trading agents compete with each other in order to best satisfy their clients' preferences and maximize their total profit. The "travel game" scenario of TAC involves three types of auctions; a) continuous one-sided that sell flight tickets, b) ascending multi-unit auctions for booking hotel rooms, and c) continuous double auctions for entertainment tickets. For each one of these types, we describe the techniques deployed by Mertacor. In flight auctions prices are updated according to a random walk process, thus the accurate prediction of the next update is not feasible. A key element of agent behavior in these auctions is its ability to accurately deduce the specific time in an auction, at which bidding will be profitable. In order to deal with the uncertainty due to price fluctuations in flight auctions and to provide our agent with an efficient decision mechanism, we have designed the z-heuristic framework. The goal of z-heuristic is to figure out when the price assumes its minimum value and recommend bidding at that moment. In the case of hotel auctions, Mertacor used fuzzy reasoning in conjunction with rule-based reasoning in order to predict the closing prices of hotel rooms, when historical data from past auctions are available. In order to bid efficiently in entertainment auctions, we have designed a bidding strategy whose goal is to preserve a pre-specified long-term profit. We finally present and discuss the results of agent benchmarking in the TAC Classic game.
- 8 May 2006
- Published by Association for Computing Machinery (ACM)
- Vol. 20, 1191-1198
Open AccessMertacor
In this paper we present the internal architecture and bidding mechanisms designed for Mertacor, a successful trading agent, which ended up first in the Classic Trading Agent Competition (TAC) of 2005. TAC provides a realistic benchmarking environment in which different trading agents compete with each other in order to best satisfy their clients' preferences and maximize their total profit. The "travel game" scenario of TAC involves three types of auctions; a) continuous one-sided that sell flight tickets, b) ascending multi-unit auctions for booking hotel rooms, and c) continuous double auctions for entertainment tickets. For each one of these types, we describe the techniques deployed by Mertacor. In flight auctions prices are updated according to a random walk process, thus the accurate prediction of the next update is not feasible. A key element of agent behavior in these auctions is its ability to accurately deduce the specific time in an auction, at which bidding will be profitable. In order to deal with the uncertainty due to price fluctuations in flight auctions and to provide our agent with an efficient decision mechanism, we have designed the z-heuristic framework. The goal of z-heuristic is to figure out when the price assumes its minimum value and recommend bidding at that moment. In the case of hotel auctions, Mertacor used fuzzy reasoning in conjunction with rule-based reasoning in order to predict the closing prices of hotel rooms, when historical data from past auctions are available. In order to bid efficiently in entertainment auctions, we have designed a bidding strategy whose goal is to preserve a pre-specified long-term profit. We finally present and discuss the results of agent benchmarking in the TAC Classic game.