Bornhuetter-Ferguson method: treatment of the first year

Hi,

When applying the Bornhuetter-Ferguson method, I noticed for the first year (the year with full run-off), sometimes the initial expected ultimate loss amount for this year is different from the actual amount paid out.

And I noticed in the course notes that the emerging liability for the first year is always coerced to 0 even if it is not zero if we apply the formula applicable for the other years.

I am wondering what is the reasoning behind this.

 

Thanks Steve. Related question about the B-F method and the past exam mentioned above.

Part (ii) of CM2B April 2021 Question 4 asks to comment on differences in reserves calculated by the B-F versus basic chain ladder. And the Examiners Report includes that "The chain ladder method gives a higher reserve.. because premiums have been increasing year-on-year....And the chain ladder method projects the existing claims forwards....
Whereas the Bornhuetter-Ferguson method relies on the estimated loss ratio from 2016"

I don't follow this explanation -if premiums have been increasing, wouldn't this imply a lower Loss ratio (=Claims/Premiums) for similar claims and so the B-F method (which relied on a higher estimate loss ratio from 2016) should produce higher reserves?

 

Firstly, a huge thanks for all those tutors and individuals who respond helpfully on Acted forums to students' posts! It is an invaluable tool for exam preparation and often provides me with knowledge (of theory and practical application) as as well as moral support.

But please can Acted reconsider this antiquated policy of moderating forum posts during an exam period. https://www.acted.co.uk/forums/index.php?threads/using-forums-during-exam-period.3426/ Since exams moved online, the IFoA require all students to attempt exams at exactly the same time worldwide (within 2 hours of each other) since from the moment that exam begins, all students have full access to download and share the exam paper. So there is no longer a risk of a student in one timezone sharing confidential information before others across the world have attempted the exam. (And has any student ever successfully hacked the Chief Examiner's laptop/IFoA exam system prior to the day of the exam and attempted to release the paper questions via an Acted forum?) Requiring moderation/vetting also places an additional burden on Acted tutors/staff, for no apparent gain.

In contrast there is a HUGE potential cost to students failing to receive urgent responses prior to exams. For example, I posted the above request for clarification on the B-F method a few days prior to the 13th Sept 2023 CM2 exam. Yet due to the vetting requirement my post only appeared on this forum after the exam and I didn’t receive a response in time. Unknowingly, this precise topic came up in the exam (Paper B Q2(iii) and so the absence of a response was costly to me (although thankfully still passed). It is common for students to have questions while studying/sitting practice papers in the week of an exam and would recommend removing this vetting requirement or ensuring tutors promptly review and approve posts within a few hours. Thanks very much again for all who enable these helpful forums to function