1. Endogenous Life-Cycle Portfolio Allocation in the Presence of both Housing Investment and Tax-deferred Accounts with D. Pelletier (2026) (submitted)
This paper develops a life-cycle portfolio allocation model to address the joint effects of housing investment and tax-deferred accounts on the portfolio allocation of households. In addition to distinguishing between taxable and tax-deferred accounts, the model employs a comprehensive housing investment structure and Epstein-Zin recursive preferences. The baseline model results show moderate risk-taking behavior by households in the presence of both housing investment and a tax-deferred account. The results further show that while housing investment has a sizable crowding-out effect on risky asset investment, the presence of a tax-deferred account displays a smaller crowding-in effect..
2. Home Purchases with and without Mortgage Credits: Implications for Portfolio Investment, with A. Gunes (2024) (submitted)
This paper is an empirical assessment of the effects of home purchases on portfolio allocation at the time of purchase as buying a home is one of the largest purchasing decisions over the lifetime of a typical household in the US. Using the Panel Study of Income Dynamics data, the results indicate significant differentiation between portfolio behaviours of mortgage purchasers and non-mortgage purchasers at the time of purchasing decision. While mortgage purchasers are more inclined to save for the down payment in the form of riskless assets, non-mortgage purchasers who are planning to purchase a home without any mortgage loans invest mostly in the risky assets. The results imply that an optimal portfolio allocation for mortgage purchasers could help them make the purchase earlier or buy larger houses with the same duration of the saving.
3. Pass-through of Different Shocks into Domestic Prices with N. Solakoglu, S. Babuscu, and A. Hazar (2025) (submitted)
This paper estimates the pass-through effects of oil prices and exchange rates on domestic producer and consumer prices in Turkey for 2003Q1–2014Q3. It utilizes both a recursive VAR model with Cholesky decomposition and a Structural Vector Autoregressive (SVAR) model with block exogeneity feature. The impulse responses of both models suggest that the SVAR model represent the dynamics of a small open economy better than the recursive VAR model. The results indicate that the pass-through from oil prices is small and statistically insignificant, whereas the pass-through from exchange rates is both significant and substantial. It further finds that the pass-through estimates have dramatically increased as we have observed a sharp increase in the inflation rate over recent years. The results have important policy implications for Turkiye, as inflation is the country's most fundamental challenge
4. Portfolio Choice Under Health Risk: The Role of Insurance Across Age Groups with A. Gunes (2026) (submitted)
This paper examines how health insurance moderates the adverse effect of poor health on household risky asset holdings over the life cycle. Using panel data from the Panel Study of Income Dynamics for 1997-2023, we find that poor health significantly reduces households’ holdings of risky financial assets, while health insurance partially offsets this negative effect. However, this buffering role is concentrated among younger households, for whom insurance fully offsets the adverse effect of poor health on risky asset holdings. By contrast, for middle-aged and older households, the moderating effect of health insurance is substantially weaker and statistically insignificant, respectively.
5. Origin-Weighted Exchange Rates and Urban House Prices (with M. Kilinc, 2026)
Housing is geographically fixed, but demand for it can cross borders. We document how cities’ international population links shape the transmission of exchange-rate movements to local house prices. Combining predetermined population shares by origin with bilateral exchange rates, we construct city specific currency exposures for 32 cities in seven countries. Local projections compare cities within the same country and quarter, thereby separating differences in international exposure from common national housing and macroeconomic conditions. A one-percent larger origin-weighted depreciation is associated with 0.14-0.22 percent higher city house prices relative to the national index after three to six quarters. The response reveals a spatial dimension of exchange-rate transmission: cities’ links to different origin countries help explain why their housing markets respond differently to currency movements. A complementary specification using US-dollar depreciation and total foreign-origin population shares yields a delayed positive relative response in more internationally exposed cities, both in the same sample and in an expanded sample of 64 housing markets across twelve countries. Together, the findings show that international population links are an economically meaningful dimension of urban housing-market exposure to exchange rates. They connect the international purchasing power of potential buyers to local housing-price dynamics, revealing how an immobile asset participates in international demand.