Let me take that.
Again, I think what Philip was articulating is that we look at fraud very differently. Both of those, be it a fictitious person or a real person, will manifest itself in identity fraud, if you want, or fraud on our shores. However, we would view them as significant risks.
Synthetic ID is a big concern for us. One of the challenges we have is to ensure that when we authenticate a client and validate the transactional activity that they're doing, we have some mechanism we can actually reconcile to, for example, to say that this is Mr. Andrews and these are Mr. Andrews' transactions, and therefore we will allow, conduct, the transactions for him.
It's difficult for us. Synthetic ID is a big concern because we have no control over the establishment of the ID, the production of that ID, but we deal with the manifestation when it's presented to us to conduct those transactions.
They're both very big topics for us. To quantify them would require a deeper dive into our respective systems to pull out the number.