Shortly after the crisis, the Fed decided it didn’t have enough knowledge of what was going on inside the banks. The Fed’s answer: More internal monitoring. So the Fed beefed up its staff dedicated to spending their days working out of the banks they were monitoring, to get as close a look as possible of what was going on.

That’s created closer ties between the Fed, particularly the New York Fed, and the banks, which should help with regulation. But it has led to its own problems. There was the case of Carmen Segarra, who was embedded at Goldman when she expressed concerns about a deal the bank was doing, only to be dismissed by her supervisors. Then last month, it was discovered that a former New York Fed employee who had joined Goldman Sachs passed along confidential supervisory information to the bank.

So, the Fed is rethinking its strategy again. In the past few months, according to The Wall Street Journal, the central bank has been recalling some of the teams of regulators it posted at the banks and relocating those employees to the Fed’s own offices. The Fed has also been looking to increase the restrictions on where bank regulators can work after they leave the central bank. The problem is they already have rules in place that restrict regulators from working for a bank they supervised, or to receive compensation from them for at least a year. And that hasn’t really worked.

Does the Fed’s pullback from internal monitoring make sense, though? With supervisors working at banks, some chumminess is going to develop. Yet closer supervision may ultimately be a good thing. Just the same, it has led to some embarrassing events, so it has to go.