Working Papers & Work in Progress
Policy Uncertainty and Foreign Investors' Entry Mode Choice: Greenfield vs. Cross-Border M&AR&R (2nd round) · J. of International Economics
with Ivan Kandilov and Asli Leblebicioglu
Abstract
We document the distinct responses of two main components of foreign direct investment (FDI), greenfield (GF) projects and cross-border mergers and acquisitions (M&As), to economic policy uncertainty across U.S. states. To this end, we construct a novel dataset that covers the period from 2005 to 2019 using granular U.S. inbound GF and M&A transactions matched to individual foreign investors. We empirically demonstrate that higher uncertainty leads to a decline in the incidence of GF projects, but an increase in the incidence of M&As within investor in a given state. Our estimates also suggest that greater local policy uncertainty has a larger positive effect on the likelihood of cross-border investment via M&A (as opposed to GF) in industries where investment is more irreversible, in states with a more challenging business environment, and for foreign investors with no prior GF experience. Our findings are in line with a stylized model we set up highlighting the importance of entry costs and local synergies in the choice of foreign investment mode.
Euro Area Inflation and Unemployment: Phillips Curve and Forecastingsubmitted
with Emrehan Aktug and Saroj Bhattarai
Abstract
We study Euro Area inflation and unemployment dynamics by addressing the identification challenges inherent to estimating the Phillips curve and providing new frameworks and results for inflation forecasting. Phillips curve slope estimation is challenging due to the simultaneity between monetary policy and economic conditions that biases OLS estimates downward. To overcome this challenge, we adopt a panel local projection IV framework. We estimate the cumulative impacts of monetary policy shocks on inflation and unemployment over a given horizon, where we use U.S. monetary policy shocks as IVs to ensure a causal interpretation of our results. We show that the policy trade-off identified through this panel IV estimation of the Phillips curve relationship is not just qualitatively, but also quantitatively, different from OLS estimation. We find no evidence that this trade-off has changed recently in the post-pandemic period. Finally, we provide novel tools for inflation forecasting based on the panel Phillips curve model. We document that panel data models yield accurate one-year ahead forecasts for 17 member states in the 2009-2023 period and clearly outperform the commonly used, naïve single-country models.
Measuring and Forecasting Inflation in China with Disaggregated Price Datasubmitted
with Barry Goodwin and Zhongyuan You
Abstract
Motivated by the limited success of existing macroeconomic models in forecasting inflation for advanced and emerging countries, we examine whether two unobserved components (UC) models by Stock and Watson (2016) are a good fit for China’s inflation. We estimate these models with monthly CPI data during the December 2006-February 2023 period to dissect the persistent and non-persistent components of inflation. Then we run a forecasting competition among UC models, and other leading competitors from the literature, including the Bayesian vector autoregression models (BVAR) and time series models for the January 2015-February 2023 period. The multi-sector UC model delivers solid forecasting performance across horizons from 1 to 30 months, with the most notable gains concentrated at short horizons, particularly at 1- to 3-month-ahead forecasts.
The 1934 Turkish Surname Law as an Early Application of Matching Theorysubmitted
with Umut Dur
Abstract
We analyze the allocation problem inherent in the 1934 Turkish Surname Law through the lens of matching theory. The law imposed three allocation rules: (i) exclusivity within each locality, (ii) first-come priority, and (iii) a genealogical override that allowed later applicants to displace earlier applicants. This structure is equivalent to a priority matching market with endowed priorities, decades before Gale and Shapley’s (1962) formal model. We formalize the historical procedure as the Turkish Surname Mechanism (TSM) and define a stability notion that meets the requirements of the surname law. We show that it does not satisfy stability, strategy-proofness, and Pareto efficiency. We then propose a simple modification that is outcome-equivalent to the Deferred Acceptance mechanism (DA) and implemented similarly to the NH4 mechanism used by MIT’s housing office. Our analysis reframes the Surname Law not only as a nation-building reform but also as a natural experiment in early matching design, yielding lessons for contemporary allocation problems involving priority classes.
Policy Uncertainty and Foreign Investment: State-Level Analysisin progress
with Ivan Kandilov and Asli Leblebicioglu
Abstract
We study the short- and long-term impacts of policy uncertainty across U.S. states on inbound foreign investment. Leveraging granular data on cross-border M&As and greenfield investment over a 15-year period, we demonstrate that higher policy uncertainty leads to a decline in greenfield investment, but a rise in the incidence of M&As. Investors also pursue smaller deals to preserve flexibility. We compare our results against those from our firm-level analysis in Dur, Kandilov, and Leblebicioglu (2025).
Credit and Exchange Ratesin progress
with Saroj Bhattarai
Abstract
We propose nominal credit as a new fundamental to explain nominal exchange rate movements. We first show that for various countries, higher credit today is robustly correlated with future exchange rate depreciation against the US dollar. In a formal econometric exercise, we next find that including credit in the empirical model improves both in-sample and out-of-sample fit substantially. We present a dynamic equilibrium open economy model that rationalizes these empirical findings.
Forecasting with Inflation Componentsin progress
with Andrew Glover
Sudden Stops and Misallocationin progress
with Sevcan Yesiltas
Publications
Uninsured Income Risk and the Welfare Effects of Reducing Global Imbalances
with Andy Glover and Jacek Rothert · European Economic Review (2025)
Abstract
We study the welfare effect of policies that balance net foreign assets when households face uninsurable income risk and borrowing constraints. Subsidizing savings in debtor economies balances net foreign assets and raises the welfare of almost all citizens by increasing world capital and raising wages: in the presence of uninsurable income risk, higher wages create a positive pecuniary externality that raises welfare for low-wealth households. The same balancing of net foreign assets is achieved by discouraging savings in lender economies. However, this policy hurts most households by reducing global capital. These results suggest that balancing global imbalances may be a positive byproduct of raising investment rates, especially in debtor countries.
Seat Allocation Problem in Public Transportation
with Oguz Afacan and Umut Dur · The Scandinavian Journal of Economics (2025)
Abstract
We study a seat allocation problem in public transportation with potential seating restrictions. We first demonstrate that the commonly used procedure suffers from significant drawbacks, including welfare losses due to unfilled seats. We then introduce a new mechanism that addresses these deficiencies while also satisfying several other desirable properties under various seating restrictions. We also show that our proposed mechanism ensures that the maximum number of agents are seated while respecting agents’ preferences and priority rankings.
A Generalized Time Iteration Method for Solving Dynamic Optimization Problems with Occasionally Binding Constraints
with Enrique Martinez-Garcia · Computational Economics (2021)
Abstract
We study a generalized version of Coleman (1990)’s time iteration method (GTI) for solving dynamic optimization problems. Our benchmark framework is an irreversible investment model with labor-leisure choice. The GTI algorithm is simple to implement and provides advantages in terms of speed relative to Howard (1960) improvement algorithm. A second application on a heterogeneous agent-incomplete markets model further explores the performance of GTI.
Mind the Gap! — A Monetarist View of the Open-Economy Phillips Curve
with Enrique Martinez-Garcia · Journal of Economic Dynamics and Control (2020)
Abstract
In many countries, inflation has become less responsive to domestic factors and more responsive to global factors over the past decades. We introduce money and credit into the workhorse open-economy New Keynesian model. With this framework, we show that: (i) an efficient forecast of domestic inflation is based solely on domestic and foreign slack, and (ii) global liquidity (global money as well as global credit) is tied to global slack in equilibrium. Then, motivated by the theory, we evaluate empirically the performance of open-economy Phillips-curve-based forecasts constructed using global liquidity measures (such as G7 credit growth and G7 money supply growth) instead of global slack as predictive regressors. Using 50 years of quarterly U.S. data, we document that these global liquidity variables perform significantly better than their domestic counterparts and outperform in practice the poorly-measured indicators of global slack that global liquidity proxies for.
Appendix
The Turkish Current Account Deficit
with Osman Furkan Abbasoğlu and Ayşe Imrohoroğlu · Economic Inquiry (2019)
Abstract
During the 2011-2015 period, Turkey’s current account deficit as a percentage of GDP was one of the largest among the OECD countries. In this paper, we examine if this deficit can be considered optimal using the Engel and Rogers (2006) approach. In this framework, the current account of a country is determined by the expected discounted present value of its future share of world GDP relative to its current share. A country, whose income is anticipated to rise relative to the rest of the world is expected to borrow now and run a current account deficit. Our findings suggest that Turkey’s current account deficit in 2015 may be considered optimal if the Turkish economy’s share in the world economy could continue to grow at rates similar to the past or to the predictions from professional forecasts. The same approach, however, indicates that the current account deficit in 2011, at its peak, was unlikely to be optimal.
Inflation as a Global Phenomenon — Some Implications for Inflation Modeling and Forecasting
with Enrique Martinez-Garcia · Journal of Economic Dynamics and Control (2018)
Abstract
We model local inflation dynamics using global inflation and domestic slack motivated by a novel interpretation of the implications of the workhorse open-economy New Keynesian model. We evaluate the performance of inflation forecasts based on the single-equation forecasting specification implied by the model, exploiting the spatial pattern of international linkages underpinning global inflation. We find that incorporating cross-country interactions yields significantly more accurate forecasts of local inflation for a diverse group of 14 advanced countries (including the U.S.) than either a simple autoregressive model or a standard closed-economy Phillips curve-based forecasting model. We argue that modelling the temporal dimension—but not the cross-country spillovers—of inflation does limit a model’s explanatory power in-sample and its (pseudo) out-of-sample forecasting performance. Moreover, we also show that global inflation (without domestic slack) often contributes the most to achieve the gains on forecasting accuracy observed during our sample period (1984:Q1-2015:Q1)—this observation, according to theory, is crucially related to the flattening of the Phillips curve during this time period of increased globalization.
Appendix
The Winners and Losers of Tax Reform: An Assessment under Financial Integration
Journal of Economic Dynamics and Control (2017)
Abstract
I quantify the macroeconomic and redistributive effects of the unilateral elimination of the capital income tax in a two-country, heterogeneous-agent incomplete markets model with progressive labor income taxes. Home, by implementing the reform, induces government responses where labor income is taxed in Home and mostly subsidized in Foreign. In addition, post-reform price dynamics reduce Home’s wealth and suppress households’ ability to do consumption smoothing, with negative effects on the majority—particularly on the poor. In turn, Foreign accumulates wealth, and price movements work particularly in favor of the poor. As a result, a large majority in Home prefers the status quo whereas Foreign supports the reform unanimously. These findings are robust to alternative scenarios where (i) the borrowing constraints are relaxed, (ii) both countries jointly eliminate capital income taxes, (iii) foreign interest income is taxed, and (iv) Home capital income tax is reduced from 40% to 35%.
