You are peer reviewing the work of an actuary who has just built a GLM for an upcoming private passenger rate change. The actuary has included deductible in his GLM. The indicated deductible relativity for a deductible of $1,000 is higher than for a $500 deductible. You suggest that this result is counterintuitive and that the actuary should instead determine deductible relativities using the Loss Elimination Ratio approach. The modelling actuary argues that the relativities from the GLM should be implemented because the volume of data is credible enough and that the company’s historical experience does show that insureds with $1,000 deductibles have had higher losses on average than insureds with $500 and that we should expect this trend to continue in the future. Fully explain why it would be inappropriate to implement the modelling actuary’s deductible relativities, even if they do accurately depict past loss experience.