This paper shows that Leland's (1992) results on the positive effects of in
sider trading on investment are not robust to the introduction of noise in
the insider's information. The paper then considers two variations of his m
odel in which the insider is risk neutral (to ensure robustness), and the i
nvestment decision is prior to the placing of the stock in the market. It i
s shown that if insider trading takes place in the primary market, it has n
o effect on the level of investment, whereas if it takes place in the secon
dary market, it has a negative effect on investment.