Investment Yield from MSA Legend Paper
I am a bit confused how Investment Yield are supposed to be calculated.
In the new MSA Legend paper, the formula given used: Annualized Investment Return / Avg Investment Assets
However, the formula in old MSA paper used: Net Investment Income / (Avg Investment Assets - Net Investment Income/2)
Are we required to remove the "Net Investment Income" in the denominator? Does the definition for "Avg Investment Assets" changed under CCIR definition?
Comments
You are right they are inconsistent I noticed this too. The ratios in the MSA legend are definitely different.
Investment Yield as you mentioned is Investment Return / Avg(Invested Assets) which is different from the CCIR version in a few ways:
1) Investment Return here is the annualized investment return which means we take the investment return and multiply it by Q. Q varies depending what quarter we are looking at. If it is Q1 then Q =4, Q2 has Q=2, Q3 has Q= 4/3 and Q4 has Q=1. For the CCIR formula I guess it assumes we are using year end numbers.
2) Invested Assets is different then the CCIR definition. The CCIR definition was cash and cash equivalents +accrued investment income + investments. The MSA.Legend definition adds in a few more items including equity accounted investees, financial instrument derivative assets and investment properties.
3) There is no requirement to add the Share of Net Income (Loss) from Equity Accounted Investees in the numerator and subtract that and Investment Return from the denominator as you mentioned.
Interestingly enough the ROE formula is also slightly different. in the MSA legend it also has the annualized adjustment which is not in CCIR. Additionally, the Equity in the denominator for CCIR is line 699 from 20.11 which is Total Equity. The denominator for this calculation is line 699 and line 899 from 20.11which is Total Equity + Head Office Account, Reserves & AOCI. This difference is likely unimportant as line 899 only applies for foreign insurers.
Like CCIR both these ratios are multiplied by 100 and so they are not a %
If asked on the exam would either formula be valid?
When we write exam questions, we usually do so based on certain papers so there will be sufficient information to answer it based on one paper, but not the other paper (usually). You should also be able to tell what paper a question in testing on most of the time. However, in the case where the question can be calculated using two different papers which leads to differing answers, both methods will be accepted