I need some actuarial assistance, please. If there is a short tail property portfolio which has grown, let's say, from $5m Y1 to $10m Y2, $15m Y3 and $30m Y4 and the experience to date is 50% GLR Y1, 45% Y2, 40%, Y3 45% (estimated fully developed) with Y4 being too undeveloped. How much statistical confidence can you have in the loss ratios sustained in Y1 and Y2 being estimators of how the book is going to perform in Y5, given relative low premium volumes in Y1/Y2, if there has been further growth and re-underwriting of the portfolio? How much is past performance a true indicator for future performance and what statistical argument could you use against historic performance?