That's a great question, and it's commonly misunderstood what is recoverable versus irrevocable. Our rules basically have been written to allow for certain categories of high-risk payments or known fraud, etc., to actually be required to be returned. We set up rules that say here are the mechanisms by which the sending and receiving organization under certain mechanisms will agree to actually return funds.
The irrevocability has to do with when that doesn't kick in. If that rule wasn't an indicator of needing to return, then it's irrevocable, and even when money has to be returned, it's the second transaction that comes back. That's the way to think about that.
As far as putting holds on it are concerned, we go back to the discussion we just had around the fraud indicators and the score. Before an RTR transaction, an actual payment transaction is initiated in subsecond time, that fraud check, a check against the risk list and an ingestion of the score would go into the participants' algorithm, and they could decide to place a hold and investigate or decide what to do with it, case by case.