Side Project

[Marshac]

I have been working on a small calculator as a side project during my "free time". Basically, there is a chart of the market from 1930-> ~6/2004, and you can set up your monthly contributions, slide a "slider" up and down the chart, and see how your "portfolio" would have done during those years.

The first version is here:
http://thealterego.com/market/

The second version (which now includes Aaa bonds) can be found here:
http://thealterego.com/market/index2.html

The first one seems to be right, but I'm not sure about the math for the bond returns... I think they're right, but if anyone would like to eyeball it and tell me what you think, that would be great. Someone previously asked me if contributions are adjusted for inflation.... they are not.

 

[wabmester]

It's just plain scary to look at a chart of the Dow in linear scale.  Most charts use log scale.   A 1000 point increase from 500-1500 (200%) shouldn't look the same as a 1000 point increase from 9000-10000 (10%+).

Your calculation of annual percentage gain is incorrect.   I looked at the 25-years from 1975-2000, and you told me that the market went up 50%/year.   I wish  

Other than that, pretty cool -- nice use of client-side scripting.

 

[Cut-Throat]

It's just plain scary to look at a chart of the Dow in linear scale. Most charts use log scale. A 1000 point increase from 500-1500 (200%) shouldn't look the same as a 1000 point increase from 9000-10000 (10%+).

Wab,

You're right on that one. Maybe that's what ***** has been doing

 

[Marshac]

Your calculation of annual percentage gain is incorrect.   I looked at the 25-years from 1975-2000, and you told me that the market went up 50%/year.   I wish  

Are you talking about the number in between the two years on the old version? Yeah, that number isn't the annual gain, it's simply the gain between the beginning year, and the end year.... I can't remember why I put it there (It's gone in the second version http://www.thealterego.com/market/index2.html).... the number doesn't mean anything =) As for the effective RR, I calculate a RR of about 21% between 75 and 2000, and a bond RR of about 9.3%.... is that about right?

Thanks for the input

 

[wabmester]

As for the effective RR, I calculate a RR of about 21% between 75 and 2000 ... is that about right?

Only if your name is Warren Buffett  

Hopefully a math-wiz will chime in here to help you out, but the ERR calculation is more complex than what you're doing.

Assuming you're calculating the final balance correctly (it looked OK to me when I glanced at your source), then you want to find the rate that would have given you the same balance with simple compounding, right?

The formula to calculate the balance B after periodic contributions C for N periods at rate R is:

B = C[((1+R)^(N+1) - (1+R))/R]

You know B, C, and N -- find R.

 

[Marshac]

All I did was take the total gain / total contributions, and then divide that % by the number of years :-/

 

[wabmester]

In other words, you ignored both the effects of compounding and DCA'ing.   I believe you  

I'm tempted to offer more help, but my dogs keep barking at me every time a deer or raccoon wanders into the yard.   It makes working damn near impossible.   Good dogs  

 

[Hyperborea]

That's wrong for a number of reasons.  Even if the contribution was a lump sum at the beginning of the period of n years you would need to take the nth root of the returns %age.  Now if as in your case the contributions are on an ongoing basis you will need to account for each of the contributions separately using a series.  There should be something out there on the web explaining this in more detail if you google for it or a finance book will have it too.

 

[wabmester]

Now if as in your case the contributions are on an ongoing basis you will need to account for each of the contributions separately using a series.

The series simplifies to the expression I gave above.

 

[Hyperborea]

The series simplifies to the expression I gave above.

Missed it but you are correct. This the formula for the calculation of the sum of a geometric series.

 

[johnblake]

Nice application Marshac,

Are you taking dividends into account ?

--John