B-F V.S Fisher/Lester V.S Wiser

[Roasted Almond]

Can anyone clarify the following points:

1) When Calculating IBNR in B-F paper, we use "Weighted Average LDF's" but Fisher/Lester's comparisions of Reserving methods are based on "Simple Average LDF's"?

2) When It comes to Wiser, he uses many different Averages for LDFs and sometimes uses "JUDGEMENTALLY selected LDF's" to clarify his point.

Can anyone tell me the reason for above? and
How would you tackle the exam questions on each of these readings?

Many thanks in advance.

[blackjack]

You are correct in your observations. On the exam, the best thing to do is to just state the assumptions you make. For example, if you use the simple avg on a Fisher/Lange problem, then state that in your answer. A simple "Assumed straight avg. LDFs" should suffice. Many (if not all) the problems leaving you open to this type of "judgment" state in the question itself to "State all your assumptions."

[L. Mo]

That's kind of the point. There's lots of judgement involved, and each author has his/her own favorite picks.

Such is the joy of Part 6!

[blackjack]

Regarding your question as to why each author is different in their selection of LDFs, the answer is that there is no one correct "method" to select LDFs. In most real world cases, you are faced with making a judgement of what looks reasonable. Sometimes the averages are skewed by points too high/low so you make a judgement on what is most representative of the development period. This is where applying actuarial methods becomes more of an art than a science.

[Wigmeister General]

Can anyone clarify the following points:

1) When Calculating IBNR in B-F paper, we use "Weighted Average LDF's" but Fisher/Lester's comparisions of Reserving methods are based on "Simple Average LDF's"?

2) When It comes to Wiser, he uses many different Averages for LDFs and sometimes uses "JUDGEMENTALLY selected LDF's" to clarify his point.

Can anyone tell me the reason for above? and
How would you tackle the exam questions on each of these readings?

Many thanks in advance.

If asked directly, all of the authors would emphatically state that no average is correct in any and all circumstances. The reasons for the differences are matters of convenience and (the author's) personal preference to illustrate the B-F and F-L methodsWiser is broadly discussing loss reserving in general.

For B-F, use the weighted average. (If you forget during the exam, then use some reasonable method, stating you forgot which selection B-F use.)

For F-L, use the simple average. (If you forget during the exam, then use some reasonable method, stating you forgot which selection F-L use.)

For Wiser, you're unlikely to come across a computational problem. If you do, you're free to use judgment, but I would explain your judgmentally selected factors clearly. For example, "Although the weighted average and simple average of all years indicates a 1.50 selection, I've chosen 1.65 because the LDF show a rising trend in the most recent years and the 1.65 is consistent with the 3-year-average."