I am interested in macroeconomics, in particular in business cycles, in the role of information, expectations and learning, and in the role of heterogeneity in determining aggregate outcomes. Such models often pose specific computational challenges and require the development of new tools.

• Papers
  • NEW VERSION (April 2025): Firm Cyclicality and Financial Frictions (slides, 2023 version , R&R at REStud, with Alex Clymo)
    How firms respond to shocks is informative about the frictions they face. Using administrative micro data we document heterogeneity in cyclicality with respect to size and age: for the youngest firms, small firms are more cyclical than large firms, but the reverse is true among older firms. Two mechanisms acting at different parts of the joint size-age firm can account for this heterogeneity: finance and heterogeneous returns to scale. Micro-level evidence supports the presence of these two mechanisms. We show that larger firms have higher returns to scale, compare the effect of leverage and returns to scale on cyclicality, and show that financial situation at entry affects firm survival. In a quantitative heterogeneous firm model, the existence of unconstrained high returns to scale firms fundamentally alters the propagation of financial frictions: Financial recessions are milder, but feature more pronounced missing generation effects.
    
    
  • Firm Dynamics and Earnings Risk (with Philipp Grübener)
    We study the role of firm and worker level shocks for individual labor earnings dynamics. A key feature of the distribution of earnings changes is excess kurtosis, with substantial earnings changes for a significant proportion of workers. Using Danish matched employer-employee data, we show that large worker earnings changes occur along the entire firm revenue growth distribution but more frequently in the tails. In particular, large earnings losses are more likely in shrinking firms due to more employment separations and wage losses of stayers. We interpret the evidence through the lens of an equilibrium search model with two-sided heterogeneity. The model reveals that while worker shocks account for the majority of earnings fluctuations, firm shocks are important for generating endogenous separations and large wage losses for stayers. Finally, the model implies significant endogenous responses of earnings dynamics to policy changes aiming to insure workers directly or indirectly through firms.
    
    
  • What can old firms tell us about the effect of age on firm size
    (revised version of Firm-level Entry and Exit over the Danish Business Cycle) (older Danmarks Nationalbank WP version)
    Revision requested at Scandinavian Journal of Economics (with Svend Greniman Andersen)
    If old firms are on average much larger than young firms, does it mean that firms get better with age? Using Danish administrative data, we construct age profiles of firm size (measured by the number of workers, turnover and value added) for ages 0-60. We find that the positive effect of aging stops around the age of 10 after which firms start to deteriorate. The positive correlation of size and age is caused by a strong selection effect where firms from the bottom of the distribution are more likely to exit.
    
    
  • Schumpeterian business cycles
    This paper presents an economy where business cycles and long term growth are both endogenously generated by the same type of iid shocks. I embed a multi-sector real business cycle model into an endogenous growth framework where innovating firms replace incumbent production firms. The only source of uncertainty is the imperfectly observed quality of innovation projects. As long as the goods are complements, a successful innovation in one sector increases demand for the output of other sectors. Higher profits motivate higher innovation efforts in the other sectors. The increase in productivity in one sector is thus followed by increases in productivity in the other sectors and the initial innovation generates persistent movement in aggregate productivity.
    
    Solving high-dimensional models with truncated shocks (complimentary paper to Schumpeterian Business Cycles)
    This note describes how a projection algorithm can be extended to solve dynamic stochastic equilibrium models of high dimensionality which feature several types of agents and non-smooth dynamics. I show how this framework can be applied to solve a multi-sector business cycle model with endogenous growth.
    
    
• Work in progress
  • Firm exit (with Petr Sedlacek)
    
    
  • Durable cycle and MPC (with Kathrin Schlafmann)
    
    
• Published papers
  • Overpersistence bias in individual income expectations and its aggregate implications
    (slides, Nationalbank WP) (with Kathrin Schlafmann)
    American Economic Journal: Macroeconomics, vol. 15, no. 4, October 2023
    Using micro level data, we document a systematic, income-related component in household income forecast errors. We show that these errors can be formalized by a modest deviation from rational expectations, where agents overestimate the persistence of their income process but otherwise form expectations in a perfectly rational and forward-looking manner. We then investigate the implications of these distortions in expectations on consumption and saving behavior and find two effects. First, low income households with this bias are too pessimistic and hence choose to borrow less than their fully rational counterparts even though their borrowing constraint is not binding. This allows a quantitative model to match the joint distribution of liquid assets and income. Second, the bias alters the distribution of marginal propensities to consume which makes government stimulus policies less effective.
    
    
  • Does Money Help Predict Inflation? An Empirical Assessment for Central Europe, (with Roman Horvath and Lubos Komarek), Economic Systems, Volume 35, Issue 4, December 2011
  Old papers and sleeping projects
  • The Optimal Monetary Policy Rule: the Role of Asymmetries and Reputation
    I study the optimal policy rule of a central bank under an asymmetric information and reputation building setting. I find that the optimal reaction function is nonlinear, despite a standard quadratic loss function and linear Phillips curve. The asymmetry arises from the signal extraction problem of the agents. Given this setting, I analyze the propositions of Rogoff and Blinder how to solve inflationary bias. I find that in the both cases the optimal policy function is nonlinear. Furthermore, the Blinder setting delivers higher value of social welfare.
    
    
  • Computer Adoption and the Changing Labor Market (with Miguel Morin)