Konstantinos Gavriilidis

Institutional Herding and Mood

Drawing on a unique data set of daily portfolio holdings for Turkish mutual funds we investigate the relationship between mood and institutional herding on the premises of various established mood proxies (weekend effect; holiday effect; Ramadan; sunshine; new/full moon) for the January 2002 – August 2008 period. Results indicate that fund managers in Turkey herd significantly, with their herding growing in magnitude as the number of active funds per stock rises and appearing stronger on the buythan the sell-side. Although the relationship of mood with institutional herding occasionally assumes the correct sign as per theoretical expectations, institutional herding is found to be insignificantly different across various mood states, thus denoting that mood does not impact the propensity of fund

Exchange-Traded Funds: Do They Promote or Depress Noise Trading?

Abstract Investing in exchange-traded funds (ETFs) is related to specific behavioral biases and heuristics encountered particularly among retail traders. Despite the increasing retail participation in ETFs, the possibility of their introduction boosting noise trading has been largely unexplored. We address this issue drawing on a sample of eight European markets based on a model encompassing rational and noise traders. Our results indicate that the introduction of ETFs has improved the efficiency and dampened any existing noise trading significance at the underlying spot markets, with no evidence suggesting that noise traders migrate from the spot to the ETF market.

Parliamentary Activity and Voting Behaviour

This study examines whether the electorate votes for the most active incumbent parliament members in the context of the Greek elections held in January 2015. I introduce a new measure of parliamentary activity based on the parliamentary work of each incumbent during the previous parliamentary period. The findings suggest that incumbents with a higher parliamentary activity are more likely to get re-elected. The results hold when controlling for age, sex and familiarity. The findings bear important implications for the electorate and the parliament members.