A blockchain replaces central counterparties with time-consuming consensus protocols to record the transfer of ownership. This settlement latency slows cross-exchange trading, exposing arbitrageurs to price risk. Off-chain settlement, instead, exposes arbitrageurs to costly default risk. We show with Bitcoin network and order book data that cross-exchange price differences coincide with periods of high settlement latency, asset flows chase arbitrage opportunities, and price differences across exchanges with low default risk are smaller. Blockchain-based trading thus faces a dilemma: Reliable consensus protocols require time-consuming settlement latency, leading to arbitrage limits. Circumventing such arbitrage costs is possible only by reinstalling trusted intermediation, which mitigates default risk.

Blockchain-based markets impose substantial costs on cross-market trading due to the decentralized and time-consuming settlement process. I quantify the impact of the time-consuming settlement process in the market for Bitcoin on arbitrageurs activity. The estimation rests on a novel threshold error correction model that exploits the notion that arbitrageurs suspend trading activity when arbitrage costs exceed price differences. I estimate substantial arbitrage costs that explain 63% of the observed price differences, where more than 75% of these costs can be attributed to the settlement process. I also find that a 10 bp decrease in latency-related arbitrage costs simultaneously results in a 3 bp increase of the quoted bid-ask spreads. I reconcile this finding in a theoretical model in which liquidity providers set larger spreads to cope with high adverse selection risks imposed by increased arbitrage activity. Consequently, efforts to reduce the latency of blockchain-based settlement might have unintended consequences for liquidity provision. In markets with substantial adverse selection risk, faster settlement may even harm price informativeness.

In statistics, samples are drawn from a population in a data-generating process (DGP). Standard errors measure the uncertainty in estimates of population parameters. In science, evidence is generated to test hypotheses in an evidence-generating process (EGP). We claim that EGP variation across researchers adds uncertainty: Non-standard errors (NSEs). We study NSEs by letting 164 teams test the same hypotheses on the same data. NSEs turn out to be sizable, but smaller for better reproducible or higher rated research. Adding peer-review stages reduces NSEs. We further find that this type of uncertainty is underestimated by participants.

This paper develops a comprehensive framework to address uncertainty about the correct factor model. Asset pricing inferences draw on a composite model that integrates over competing factor models weighted by posterior probabilities. Evidence shows that unconditional models record zero probabilities, and post-earnings announcement drift, quality-minus-junk, and intermediary capital are incremental factors in conditional asset pricing. The integrated model tilts away from the subsequently underperforming factors, and delivers stable and admissible strategies. Model uncertainty makes equities appear considerably riskier, while model disagreement about expected returns spikes during crash episodes. Disagreement spans all return components involving mispricing, factor loadings, and risk premia.

We theoretically and empirically study portfolio optimization under transaction costs and establish a link between turnover penalization and covariance shrinkage with the penalization governed by transaction costs. We show how the ex ante incorporation of transaction costs shifts optimal portfolios towards regularized versions of efficient allocations. The regulatory effect of transaction costs is studied in an econometric setting incorporating parameter uncertainty and optimally combining predictive distributions resulting from high-frequency and low-frequency data. In an extensive empirical study, we illustrate that turnover penalization is more effective than commonly employed shrinkage methods and is crucial in order to construct empirically well-performing portfolios.

During the spring semester 2021 I designed a new course Advanced Empirical Finance: Topics and Data Science for the master students at KU. The course aims at providing a unified coding framework in R to tackle many (probably too many) common issues in empirical finance:

We are pleased to announce that the “Vienna–Copenhagen Conference on Financial Econometrics (VieCo)” takes place June 2-4, 2022 in Copenhagen. It is jointly organized by the Department of Economics of the University of Copenhagen and the Department of Statistics and Operations Research of the University of Vienna.

I regularly supervise empirical bachelor projects or master thesis in Finance on the following topics: Machine Learning Applications in Asset Pricing, Optimal Portfolio Choice problems or Microstructure of Decentralized Exchanges.

Bellman Equation outside of Finance. A short report on my first Kaggle Competition

A short series of posts on handling high-frequency data from Lobster and R