Liquidity pricing is very critical in explaining fund performances, especially during periods where liquidity experiences sudden surges or dry-ups (i.e. liquidity imbalances, LIs) connected with a significant price movements. Hence, in the first part of the thesis we decided to investigate this relationship, focusing on an illiquid markets where the effect should be more evident. Even if transaction volumes do not seem to provide any pricing information, we find that return chasing behaviour and smart effects coexist, with the former being associated to investment activities and the latter to disinvestments. Since our empirical results confirm that liquidity and prices influences each other, we decide to investigate the formation of investors' choices through a theoretical model that embeds liquidity, risk and return expectations in the pricing of mutual funds. We model net flows at fund-level and we observe a LI upon the occurrence of four market conditions: high transaction costs, wrong endogenous pricing policy, growing uncertainty of fund returns and a negative "tail" state of the economy. Finally, since manager's actions are not considered in our models but seem to be important in the pricing of such vehicles, we study their effect on investors' decisions via a principal-agent dilemma setting. We discover that among "direct" (i.e. remuneration policies) measures, fund price volatility sensitive-remuneration solves the dilemma and different patterns among specialised investors and the aggregate market exist. Although similar variables drive preferences in direct real estate and the equity market, with performance related variables leading the trend, they often present opposite effects. We conclude that a limited and time-varying rationality and predictability exist in mutual fund markets.