Finance Internal Seminar: Rasmus Højer Nielsen (PhD presentation) and Daniel Neukirchen

Presenter: Rasmus Højer Nielsen, AU

Title: Joint SPX-VIX calibration in Volterra-Heston models with jumps

Abstract: This paper examines the joint SPX-VIX calibration capability of state-independent jump extensions of the Volterra-Heston model class across many dates and a wide set of strikes and maturities. In particular, we augment the classical continuous affine Volterra pricing model by compound Poisson type jump discontinuities in both the log-price and the variance process. We show that this extension preserves the tractability of transforms required in the pricing of SPX and VIX options, since the relevant Ricatti-Volterra equation is unchanged compared to the model without jumps. We develop a simple and transparent data filtering and calibration methodology for the joint SPX-VIX calibration problem and apply this to tractable versions of the jump extended Volterra-Heston model for a broad set of recent days. The empirical results show that the addition of low frequency jumps in the log-price are useful and can meaningfully improve the fit of the left wing of the SPX implied volatility smile, while preserving a reasonable fit to short term VIX smile. Allowing for jumps in variance can improve the fit even further, but this improvement is mainly observed on dates where the non-jump models already produce reasonable fits.

Presenter: Daniel Neukirchen, AU

Title: Corporate Capital Allocation Strategies

Abstract: Although corporate capital allocation decisions are inherently interdependent, research has primarily examined them in isolation, leaving our understanding of capital allocation strategies limited. Using a large language model, we recover such strategies from forward-looking earnings call communication. Recovered strategies predict firms’ future capital allocation decisions, operating performance, and firm value. Firms with more explicit strategies exhibit lower analyst forecast dispersion and forecast errors, smaller earnings surprises, and an attenuated stock market reaction to earnings surprises. The evidence suggests that more explicit capital allocation strategies are associated with more profitable firm decisions and help financial markets interpret and predict corporate earnings. 

PhD presentation as part of mandatory 1st or 3rd year presentation. The time is extended to one hour; the presenter has 40 minutes for the presentation, 5-10 minutes for the discussant, and 5-10 minutes for questions.

Organizers: Stefan Hirth and Mads Markvart Kjær