Please use this identifier to cite or link to this item:
Tchamyou, Vanessa S.
Asongu, Simplice
Year of Publication: 
Series/Report no.: 
AGDI Working Paper No. WP/17/028
This study complements the scarce literature on conditional market timing in the mutual fund industry by assessing determinants of market timing throughout the distribution of market exposure. It builds on the intuition that the degree of responsiveness by fund managers to investigated factors (aggregate liquidity, information asymmetry, volatility and market excess return) is contingent on their levels of market exposure. To this end, we use a panel of 1467 active open-end mutual funds for the period 2004-2013. Fund-specific time-dynamic beta is employed and we avail room for more policy implications by disaggregating the dataset into market fundamentals of: equity, fixed income, allocation and tax preferred. The empirical evidence is based on Quantile regressions. The following findings are established. First, there is consistent positive threshold evidence of volatility and market return in market timing, with the slim exception of allocation funds for which the pattern of volatility is either U- or S-shaped. Second, the effect of volatility and market return are consistently positive and negative respectively in the bottom and top quintiles of market exposure, but for allocation funds. Third, the effects of information asymmetry and aggregate liquidity are positive and negative, contingent on specifications, level of market exposure and market fundamentals. The findings broadly suggest that blanket responses of market exposures to investigated factors are unlikely to represent feasible strategies for fund managers unless they are contingent on initial levels of market exposure and tailored differently across ‘highly exposed’-fund managers and ‘lowly exposed’-fund managers. Implications for investors and fund managers are discussed.
Mutual funds
Market timing
Quantile regression
Document Type: 
Working Paper

Files in This Item:

Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.