While considerable evidence exists that institutions herd, the issue of why herding takes place remains unresolved. Using monthly holdings data for Portugal, we find clear evidence of herding and investigate whether such behaviour is intentional or spurious. By analysing herding under different market conditions, we conclude it is intentional. Month-of-the-quarter analysis suggests reputational reasons drive behaviour. Results are consistent with herding interacting with window dressing to determine funds, buy and sell decisions. The findings are important in understanding market dynamics and fund managers' behaviour and are of great significance to investors in managed funds.

We investigate herding in eight African frontier stock markets between January 2002 and July 2015, given the limited evidence on herding in frontier markets. Herding appears significant throughout the 2002-2015 period for all markets, with smaller stocks found to enhance its magnitude. Herding entails no clear asymmetries conditional on market performance; conversely, it appears notably asymmetric when conditioned on market volatility, as it is significant (or stronger) mainly during low volatility days, without this pattern, however, surviving when accounting for the 2007-2009 crisis. The US and South African markets motivate herding on a small number of occasions only, while the return dynamics of a regional economic initiative's member-markets are found to induce herding in each other very rarely, thus demonstrating that investors' behaviour in markets with low integration in the international financial system is not significantly affected by non-domestic factors.

The present paper aims at investigating the extent to which institutional herding at the industry level is motivated by intent or not. We assess the presence of intent using both market and sector states based on three variables (returns; volatility; volume), in order to gauge whether herding intent at the sector level is more relevant to conditions prevailing in a sector or the market as a whole. Using a unique database of quarterly portfolio holdings of Spanish funds, we produce evidence denoting that institutional herding in the Spanish market is intentional for most sectors, manifesting itself mainly during periods when the market as a whole or the specific sector under examination has underperformed, generated rising/high volatility and exhibited rising/high volume.

In view of evidence linking herding and social mood, we examine whether the positive mood documented during Ramadan translates into higher herding compared to non-Ramadan days.Drawing on a sample of seven majority Muslim countries, we report significant herding during Ramadan in most of our sample markets. Additionally, we show that herding appears significantly stronger within rather than outside Ramadan for most tests whereby its significance is manifested on both Ramadan-and non-Ramadan-days. Overall, herding significance within/outside Ramadan exhibits some variation in its levels across markets in relation to variables reflective of market states, both domestically (market returns; market volume) and internationally (US market returns; US investors' sentiment; global financial crisis) market states.

 We examine whether joining an exchange group confers an effect over herding in the group's member-markets and if this effect persists when accounting for various domestic and international market states, the dynamics of the group's member-markets and the outbreak of financial crises  We test for the above in the context of the Euronext, which contains four equity markets (Belgium, France, the Netherlands and Portugal)  Herding is significant post-merger in all four markets; herding in Portugal is significant (yet less strong) pre-merger as well. These results are robust when controlling for various domestic/international market states and the dynamics of the group's markets  Herding following the outbreak of the euro-zone crisis is significant in Belgium, the Netherlands and Portugal, and is motivated by the dynamics of the group's two largest markets (France and the Netherlands). *Highlights (for review)Page 2 of 33 A c c e p t e d M a n u s c r i p t Herding Dynamics in Exchange Groups: Evidence from Euronext AbstractThis study investigates in the context of the Euronext, whether joining an exchange group affects herding in the group"s member-markets and if this effect persists when accounting for various domestic and international market states, the dynamics of the group"s member-markets and the outbreak of financial crises. We find that herding is significant post-merger in all four constituent equity markets (Belgium, France, the Netherlands and Portugal) of the Euronext, with herding in Portugal being significant (yet less strong) pre-merger as well. These results are robust when controlling for various domestic and international market states, as well as the dynamics of the group"s markets. The period following the outbreak of the euro-zone sovereign debt crisis produces significant herding in Belgium, the Netherlands and Portugal, with this herding being motivated by the dynamics of the group"s two largest markets (France and the Netherlands).JEL classification: G12, G15, G31