Yeah, if I can add to that Kai. I mean, obviously, we spent a lot of time analyzing results with us through the loss bridge [ph]. We identify factors like ceded in adjustment premiums, rates, trend, mix, experience, you name it. I think a couple of highlights; number one, we didn't have the adverse impact of Ogden, which as you know, hurt us last year, and that had a meaningful impact on the loss ratio. We continue to have growth in the A&H business, which tends to have a lower loss ratio -- a higher expense ratio, but lower loss ratio, through the business. We have had some ups and downs on the losses, so as Pete says, we didn't have any substantial large loss activity in the reinsurance division, good property activity. On the other hand, we had pretty poor aviation results this quarter. We had some additional losses on marine. Obviously, the property book in the U.S. was affected by the weather events. So every quarter, you are going to have something. But by and large, I think, not unreasonable. And of course, the addition of the Novae book makes a difference, because the Novae book, as Pete pointed out, has a higher acquisition expense, but generally a lower loss ratio. And fully two points of the improvement in the loss ratio is the inclusion of the Novae book in our consolidated results. So from our perspective, this really is consistent with the run-off of the discontinued lines that we have had. We got out of retail property and retail casualty in the U.S. in the end of 2016. But those lines hurt us in 2017. As those premiums ran off, you are now seeing a book, which is less exposed to the discontinued line, and more demonstrating the underlying profit potential of the ongoing continuing business.