Ah, this is a problem regarding conditional probability. We define a probability space $\Omega = \\{a, b, c\\}$, where
* $a$ means that the fund has beat the market,
* $b$ means that the fund has not beat the market, and the investors did not withdraw,
* and $c$ means that the fund has not beat the market, and the investors withdrew their money.
Now from the assumption, it follows that if $P$ is the probability measure, then $P(\\{a\\}) = 0.4$, and $P(X) = 0.6$, where $X = \\{b,c\\}$ is the event that the fund underperformed. Moreover, $$ P(\\{c\\}|X) = 0.5 = P(\\{b\\}|X). $$ But since $\\{c\\} \cap X = \\{c\\}$ and $\\{b\\} \cap X = \\{b\\}$, we may deduce from that that $$ P(\\{c\\}) = P(\\{b\\}) = 0.5*0.6 = 0.3. $$
Now note that the event that the fund still exists is $Y := \\{a, b\\}$. Then $P(Y) = 0.4 + 0.3 = 0.7$ and moreover
$$
P({a}|Y) = P({a}|Y)/P(Y)=0.4/0.7 = 4/7 > 1/2,
$$
which is what we were looking for.