There have been some good and interesting answers given already. I agree with Billy Van Jura in that there are regulatory risks that will differ from jurisdiction to jurisdiction; however, these risks are pretty easily managed. I also have to agree with Kapil Mehta and Stan Belot that the broker has a duty to act in the best interest of the customer and while in my experience this is not normally a problem with brokerages owned by insurers it is certainly a risk.
I would argue that the largest risk to insurers in owning a brokerage is the stability of their distribution channels. Obviously, any insurer that is going to invest in partial or total ownership of a brokerage likely employs a broker distribution model. With this sort of distribution model the good will and faith between the insurer and the broker is of key importance. While many brokers will not care too much about an insurer taking an ownership stake in one of it’s competitors many others will. If the relationship between the insurer and it’s other brokers is not carefully managed during the period it is acquiring ownership stakes in brokers or in the immediate aftermath then it could find itself with more brokers walking away from their brokerage agreements than it can make up for with the expansion of business at the brokerage it acquired. This would certainly put the insurer in an unfavorable position with it’s remaining brokers who would now have leverage to negotiate more favorable terms when renewing agreements, as the insurer would be negotiating from a position of weakness. I am uncertain whether or not any of the insurers (including my own) buying up ownership positions in brokerages have encountered this problem; however, it is a real risk that is less easily managed than some of the others mentioned.