When professional investors such as pension funds, fund-of-funds etc. are asked by VC firms to invest into their funds, they face a problem of adverse selection: VC firms’ skills to generate high returns are a hidden characteristic (Akerlof, 1970). Hence, VC firms provide them with information on the portfolio of current VC funds they manage. At that time current funds are often not older than 3-4 years and final performance of these funds is hard to predict. While previous literature has shown a conflict between the general partners of a VC fund and its investors (Gompers, 1996; Kandel, Leshchinskii & Yuklea, 2011), this study goes one step further. We aim at identifying certain investment and divestment patterns of the current VC fund until the time of raising a new fund, which might convey reliable information to predict final fund performance. Thereby, we help investors to distinguish the ‘good’ from the ‘bad’, which is particularly important given the headwind the VC industry is currently facing due to its underperformance in comparison to other asset classes (Harris, Jenkinson & Kaplan, 2012).