Compounding regular payments that increase at fixed intervals

Hi,

I wonder if someone could clarify the "correct" method to work out the result when compounding a series of monthly premiums over a fixed period of time when the premiums increase on an annual basis. I'm asking because a piece of software we have been using for a long time has been producing results that I just can't quantify by using Excel.

Therefore, I'm looking for a little clarification as to whether I am looking at this correctly.

For example, I was using a figure of £100, paid monthly from

01/01/1995 to 01/11/2007, increasing by 3% per annum and using a growth rate of 5% p.a.

While it may be possible to write a one-line formula to work this out, I used an Excel Formula / Macro to project each separate monthly payment.

First of all, I work out a monthly growth factor [1.05 ^ (1/12) - 1] 0.41%

As there are 154 months in this time period,

1 154 100.00 187.04 2 153 100.00 186.28 3 152 100.00 185.52 4 151 100.00 184.77 5 150 100.00 184.02 6 149 100.00 183.27 7 148 100.00 182.53 8 147 100.00 181.79 9 146 100.00 181.05 10 145 100.00 180.32 11 144 100.00 179.59 12 143 100.00 178.86 13 142 103.00 183.48 14 141 103.00 182.73 15 140 103.00 181.99 ............ 150 5 142.00 144.92 151 4 142.00 144.33 152 3 142.00 143.74 153 2 142.00 143.16 154 1 142.00 142.58

Each projected value is [Premium x Adjusted Growth Rate ^ remaining months]

The result is the sum total of the projected values. In this case my spreadsheet reports it as £25,102.28

Now, the unnamed piece of software I have using reports the same result as £25,691.21. If I change the end date to the middle of the month (15/11/2007), the answer does not change, so I assume that it is also using monthly periods within its calculation. However, I just can't arrive at this result, by tweaking my own calculations, adding an extra month, etc...

I'm aware that you can get different results by changing compounding periods, etc, and they are all technically "correct". I guess what I am interesting in knowing is:

· Is the way I perform my own calculations a valid one? · What other valid ways can you use to do this?

Many Thanks in advance for any comments or suggestions.

Regards Neil.