Long-memory and volatility spillovers across petroleum futures
Lovcha, Y. and Pérez-Laborda, A. (2022): "Long-memory and volatility spillovers across petroleum futures" Energy
We analyze volatility connectedness across petroleum markets, making two significant contributions to the literature. First, we relax the questionable assumptions on the persistence of volatilities in previous studies by accounting for the long-memory possibility, which is more in accordance with the empirical evidence. To do so, we rely on the connectedness framework of Diebold and Yilmaz (2012;2014), but, unlike previous literature, we employ a fractionally integrated VAR. The second contribution extends the frequency-domain approach proposed in Barunik and Krehlik (2016) to long-memory specifications. This extension, never considered before, allows us to compare connectedness measures based on short- and long-memory specifications across frequency ranges. We find overwhelming evidence of long memory in petroleum volatilities, with the use of standard short memory models magnifying spillover risk, especially at low frequencies.