Part of the danger is that the excess of funds is used to create permanent costs (e.g. hiring more people than needed, running on more servers than needed, creating more management layers than needed) that then become a danger to the organisation if there are ever lean times. It's very easy to increase costs, and very difficult to reduce them safely.

This gets exacerbated when the management get all excited about their pet projects and reroute funding & resources to those instead of the main thing that the funds were supposed to be for. Mozilla is the classic example of this - the Firefox team is not a priority when the funds get low, because of the various hare-brained projects the management have come up with.