The argument that investors are not harmed by insider trading rests on research which shows that the effects of insider trading on prices is fairly limited. The problem with this is that it also undermines the argument that insider trading makes markets more efficient — it cannot make markets more efficient unless it affects prices.
I think it is obvious that there would be far more insider trading if it were legal. There would be no need to keep transactions unobtrusive (except for long enough get them done before the market notices). Insiders would buy as much as they could, and sell information on to well financed traders. This would mean a far greater impact on prices.
It is also very easy to think of situations where it is very clear who loses. For example, insider trading ahead of a takeover bid benefits insiders at a cost to the (would be) acquirer's shareholders.
It is less clear who the losers are when it comes to trading on financial reports. A seller who would not have sold, or a buyer who would have stayed clear if the news had been known to them, are obvious examples.
The losers are not necessarily shareholders of the company, they are a diverse group. This is one reason why the pro-deregulation lobby's proposed solution of privatising the ban on insider trading by making it a contractual matter fails. Another is that the shareholders do not have a contractual relationship with employees of a company, so it is not clear how they could be compensated for a breach of contract. In addition variations in contracts would make insider trading rules variable, unpredictable and opaque to investors: markets need a certain degree of uniformity.
Another problem is that the substantial gains of insider trading require strong deterrents. Contracts could certainly provide this as damages could bankrupt offenders. However, combining this with the lower level of proof (probable, rather than beyond reasonable doubt) of civil law, is likely to lead to many innocent people being punished.
The practical problem with such contracts are numerous. Who needs to sign them? What happens if a group of people who have de facto, if not strictly official, access to information (secretaries, IT people) are not made to sign such contracts? What about temps? What about outsourcers' staff? Suppliers staff? How are investors supposed to keep track of which companies make which staff agree to what terms? Who is going to enforce the contracts?
Of course, changes to the law could address many of these issues, but the end result of that would come very close to fixing the problems of removing the ban on insider trading by re-inventing it in a different form.
Another wide effect of legalising insider trading would be to make fundamental analysis less worthwhile. Why bother, when there is more money to be made chasing the latest insider information?
I have seen, in a minor market, what happens fundamental analysis is neglected in favour of trading on insider information. The result is that information in the accounts is taken at face value, and the scope for misleading investors through creative accounting increases greatly. With no ones analysing the accounts in debt, who is to question discrepancies?
The argument that it is a useful incentive is even worse. Certainly, directors could make money by buying ahead of good results. They could make just as much by selling ahead of bad results. What allowing insider trading is likely to encourage them to do is to take risks, so that the outcomes of those risks would create a stream of price sensitive information to trade on.
the market efficiency argument assume that laws on disclosure cannot be changed. There is a problem with exceptions to the disclosure of price sensitive information. Rather than allow them to be disclosed by insider trading, simply tighten disclosure requirements. I am, in any case, very sceptical about claimed business reasons for keeping information private.
Legalising insider trading would also encourage manipulation of the market and make manipulation by rumour more destabilising because it would be more credible. If there was a greater stream of real inside information being leaked, how easy it would be to insert a few well chosen falsehoods into that stream. it would also encourage the hoarding of information, rather than its public release, to maximise opportunities to trade.
This is, in many way, just part of a wider debate on how regulated capital markets should be. I believe that they need heavy regulation, because there are so many opportunities to cheat, and these ultimately undermine the markets themselves. I was planning to write a post on the wider issues, when the FT drew my attention to this particular one.