Research

The Bright Side of Multinational Tax Planning: Cost of Capital and Real Investment [Link to Paper]

Multinational enterprises (MNEs) can lower their tax liabilities by strategically choosing to re-allocate capital internally between subsidiaries (debt shifting) and by pricing these internal capital flows advantageously (transfer pricing). In this paper, I present a new channel that allows MNEs to decrease their effective cost of capital through tax planning by exploiting the spread between the external and internal cost of capital. MNE subsidiaries are optimally larger and less profitable than comparable individual firms. While profits are shifted from high-tax to low-tax countries reducing overall tax receipts, production expands. Implications for tax avoidance policies that maintain the production- enhancing properties of internal debt are discussed.

Presentations: SMYE 2022, ITFA Conference 2022, International Workshop for Early Career Economists: Shaping Globalisation (Mainz), IIPF Congress 2022, ETSG 2022, European Commission - JRC: Fiscal Unit, University of Passau

Commitment or Exit: Dynamic Equity Compensation with Heterogeneous Shareholders [Link to Paper]

I consider equity-based compensation when the blockholder engaged in corporate governance and traderss setting prices in financial markets have heterogeneous preferences over the firm policy. Tying the manager's compensation to the stock price solves the managerial moral hazard but also ties managerial incentives to the preferences of the marginal trader. Instead, the blockholder may rely on her dynamic relationship with the manager. This is easier to achieve for blockholders, such as founding families or index funds, exogenously locked into their investment than for blockholders who can easily exit. Dynamic incentive provision with shareholder heterogeneity implies higher compensation for the manager.

Presentations: 16th RGS Doctoral Conference, NBIM-Oxford Conference on Common Ownership / Experimental Governance and IO, 6th Erasmus Corporate Governance Conference, ESSFM

Generically Separable Belief-Types [Link to Paper]

In most mechanism design applications, it is assumed that different types have different preferences but share the same belief. The agent earns information rents. Relaxing the unrealistic assumption of common beliefs, Crémer and McLean (1985,1988) consider an environment where no two types share the same belief or the same preferences. The agent does not earn information rents if the beliefs satisfy a generic convex independence condition. I re-assess this convex independence condition in a richer type space, where two different types may share preferences or beliefs. There, the optimal mechanism generically separates the types by the latter: the allocation and the agent's expected utility carry over from the optimal mechanism in the restricted sub-game including only types with the same belief. The assumption that all agents share the same belief is generically without loss of generality in determining the efficiency properties of the optimal mechanism. 

Inefficient Full Surplus Extraction [Link to Paper]

Full surplus extraction (FSE) occurs when a principal designs an incentive compatible mechanism under which the privately informed agent earns no information rents. Previous FSE results imply that FSE is (weakly) efficiency enhancing and thus only beset with distributional concerns. In this paper, I derive the necessary and sufficient condition  for FSE in the canonical adverse selection model with type-dependent reservation utilities and general quasi-linear utility functions. There, I show that FSE can be inefficient: the FSE mechanism can be optimal even though the best non-FSE mechanism implies a higher total surplus.

Polarisation Pass-Through with Marcin Roter

The recent increase in political polarisation has received significant attention in the media and academic literature. In this paper, we study how exogenous changes in the polarisation of party elites and/or voters affect the polarisation in equilibrium policies, namely the polarisation pass-through. To this end, we develop a two-party model of party design with a two-stage policy-making process. After the party leaders choose their party line, they compete in an electoral competition. In the second stage, the winning party leader proposes a policy that passes only if it receives enough parliamentary support. Even though, generally in equilibrium, an increase in elite polarisation leads to more polarised policies, the pass-through is rarely complete and can even be zero or negative. The pass-through is also structurally different depending on the level of mass polarisation and office rents.

Presentations: Swiss Society of Economics and Statistics, ECWE - Verona