Page 1 of 1

### Question about percent of expenses funded

Posted: Fri Feb 29, 2008 8:51 pm
Hi Jim!

I’m really glad I came across your Flexible Retirement Planner…I can’t wait to implement it with clients. I do however need one clarification.

I am having a hard time getting my arms around the “% of expenses funded” figure and want to see if I’ve got it:

The Median Total Withdrawal is the amount, “chosen” during the iterations, where half of the income figures used come in above the amount specified as annual spending and half come in below the amount specified….correct?

THEN, the “% of expenses funded” says that out of x-thousand iterations, the amount “chosen” during the iterations was on average x% of the spending amount specified?

Example:

Annual Retirement Spending Specified: \$50,000

Median Total Withdrawal: \$50,438 – the mark where 50% of the iterations used an amount above this and 50% of the iterations used an amount below this.

% expenses funded: 97.3% This says that, on average, over all the iterations, the spending (and thus, withdrawal) amount used was 97.3% of \$50,000 (the annual retirement spending specified)

I appreciate you getting back with me! Once again…this is a SUPER tool!

### Re: Question about percent of expenses funded

Posted: Fri Feb 29, 2008 8:55 pm
frp user wrote:I am having a hard time getting my arms around the “% of expenses funded” figure and want to see if I’ve got it:

The Median Total Withdrawal is the amount, “chosen” during the iterations, where half of the income figures used come in above the amount specified as annual spending and half come in below the amount specified….correct?
I think you've got it. As you are probably aware, the median total withdrawal is reported for each year of the simulation, rather than overall for the plan. You can see the values for each year by clicking on the graph or by looking at the additional output tab.
THEN, the “% of expenses funded” says that out of x-thousand iterations, the amount “chosen” during the iterations was on average x% of the spending amount specified?
This is also a median value, so the value reported is that value where half the iterations had higher values and half had lower values. Using a median instead of an average is more reliable, especially since the % expenses funded will be zero for each of the failed simulation iterations (with 95%, that means 500 iterations failed) which would skew an average if I calculated one.
Example:
Annual Retirement Spending Specified: \$50,000

Median Total Withdrawal: \$50,438 – the mark where 50% of the iterations used an amount above this and 50% of the iterations used an amount below this.
Yes, this is exactly right.
% expenses funded: 97.3% This says that, on average, over all the iterations, the spending (and thus, withdrawal) amount used was 97.3% of \$50,000 (the annual retirement spending specified)
Not quite - This is also a median value so the number reported is the halfway mark rather than the average. Again, If I used an average, the failures in the simulation would skew that value because they have 0% of expenses funded.

As I review the documentation for these outputs, I see what I have written is pretty light. I appreciate your question and the opportunity to correct a weakness in the documentation.