This study investigates what drives investor return extrapolation through the lens of the X-CAPM framework (Barberis et al., 2015). By calibrating this model to stock market data, we quantify the extent of extrapolative weighting both market-wide and across individual stocks. Our findings point to a salience-based explanation: U.S. investors anchor their expectations on broad historical return distributions. A sentiment index derived from our estimates reflects market valuation shifts and inversely correlates with investor age. Internationally, sociodemographics—especially education—play a key role in shaping extrapolative beliefs. Extending the analysis across stocks, we show how industry classification and return distribution salience influence extrapolation behavior. 

We examine how mutual funds respond to corporate bond defaults within their portfolios, using a large panel of portfolio holdings of U.S. corporate bond mutual funds. Controlling for market-wide defaults and a wide set of fund characteristics, we find that funds significantly reduce exposure to rating and maturity categories in which they personally experienced defaults over the past year. We then investigate whether funds adjust overall risk-taking by de-risking or, conversely, by reaching for yield within unaffected rating and maturity categories. We assess the implications of these reallocations for fund performance and investor flows. Finally, we highlight the market-level implications of mutual fund rebalancing after defaults by studying whether it affects bond pricing, particularly through wider bond spreads and reduced liquidity for bonds held by default-affected funds.

We investigate whether adjusting for seasonality in intraday foreign exchange (FX) returns enhances volatility predictions. To achieve this, we first estimate and fit a Flexible Fourier Form (FFF) at a daily frequency to remove seasonal patterns from five-minute returns. Using GARCH and GAS specifications with different distributional assumptions, we analyze the EUR/USD, EUR/GBP, and GBP/USD exchange rates. Our findings reveal that deseasonalized returns do not improve volatility forecasts at the five-minute horizon. However, at intermediate horizons (1 to 4 hours), removing seasonality enhances forecast accuracy. This suggests that seasonal effects play a more significant role at these horizons than at very short-term intervals.

Presentations:

PhD Workshop @ VU Amsterdam (2025)

PhD Mentoring Day 31st Finance Forum @ ULL (2024)

PhD Workshop @ VU Amsterdam (2023)

Brownbag Finance Seminar @ VU Amsterdam (2022)

Research in Behavioral Finance Conference (RBFC) @ VU Amsterdam (2022)

Summer schools:
SoFiE Summer School in Financial Econometrics @ National Bank of Belgium (2024)

Stewardship (co-organization):

EFA 50th Annual Meeting @ VU Amsterdam (2023)

Empirical Asset Pricing Meeting @ VU Amsterdam (2023)

Conferences attended:

EFA 51st Annual Meeting @ National Bank of Slovakia (2024)

EFA 49th Annual Meeting @ IESE Business School (2022)