- 6,480
Lets attempt to settle/explain this.
I believe this is what people are somewhat saying when thinking about this topic......but maybe not. Not saying this is the thought, but I think this is why there is a debate here as to whether 1 day a year ends up being equal to 12 months.
Here is comparison of the last 12 months had someone had 100% in an index product with 0% floor, 9% cap & 100% participation rate. $120,000 from 1 day last year to 1 day this year compared to having $10,000 each month having 1 day, so therefore 12 days. A person doing this the last 12 months would be substantially better, having been able to make almost 4% instead of a 13% loss (0% credited)
PS-- reminds me of how I would love to find a way to start dollar cost averaging back in the market right now on a daily incremental basis, but most places I find only allow it to be set up on a monthly basis & who wants to guess at mutual fund/ETF value on 21 different days a month?
View attachment 8175
Hey Allen, you are giving me credit for a lot of Annuity Technical Knowledge, Mathematical Knowledge, and Financial Theory Knowledge I do not possess.
scagent and I are talking past each other.
One reason is that back at the beginning of the thread when I made the comment about monthly crediting options, I did not even understand that there are at least 2 different monthly crediting computations and it looks like I have posted links to both of them at various points through the thread. So, a person reading the thread is not going to know what was in my mind when I said Monthly Crediting.
A second reason is that (I believe) scagent wants me to agree that Monthly Point to Point and Annual Point to Point are based on the same numbers and he uses whole number computations to demonstrate this. I agree with his math, but I do not agree with applying it to Annual Point to Point and Monthly Point to Point and to some company's definitions of Monthly Average because the companies use percentage values for those computations.
scagents whole number math does not work for percentage math.
Let's take a string of 5 numbers: 16, 20 14, 25, 28.
For convenience we'll say those numbers represent a value at the end of 5 periods called months. 5 months = the big period.
The overall change for 5 months is 28 - 16 = 12
The periodic changes for 5 months are 4, -6, 11, 5. These sum to 12.
This is the demonstration scagent is making.
However the indexes convert these whole number computations to percentage based computations. (We'll ignore the concept of caps to just focus on percent math)
The percentage change for the big period is (28-16)/16 = 75% (conceptually Annual P2P)
The percentage changes for the months are:
4/16= 25%; -6/20 = -30%; 11/14 = 78.6% 3/25 = 12%.
The percentage changes for the months sum to:
25.0%-30.0%+78.6%+12.0%=85.6%
When you use whole numbers, the sum of monthly changes = the big period change of 12.
When you use percents, the sum of monthly changes 85.6% does not = the big period percentage change of 75%.
Although I don't understand the math and financial theory involved, I extend from that to say that percentage based Annual Point to Point computations based on two points will yield different indexing results from percentage based Monthly Point to Point computations based on 13 points. (And probably Monthly Average computations based on the average of a sum of percents will yield different indexing results from Monthly Average whole number computations based on the average of 12 month end index values.