Reit taxation | Page 2 | General financial discussion | Discussion forum

Norman1 said

rhvic said

…

Also, I do regard the effective purchase price as being the original price minus whatever Slate told me is ROC over the years - hence the $6.31 figure, since ROC is supposed to reduce cost base - I see this as a 'refund' of part of the purchase price. As to distributions received, I have subtracted from that column whatever equates to the reported ROC, since that is not really a dividend in my view, simply paying back some of what I paid in the first instance. At least, that is my way of doing the accounting.

That's the accounting for tax purposes and not for investment return purposes. There are lots of instances where the two don't match.

Just because one can deduct 4% of a building each year for tax purposes doesn't mean the building has actually deteriorated by 4% each year and no more than 4% each year.

I don't calculate returns using the grossed-up value of dividends received, in spite of the fact that I need to report the grossed-up value and not the actual value of dividends received for tax purposes.

Similarly, I use $7 for cost of some bank shares when I calculate investment returns. I don't use the tax cost of $14 for those shares, which is from artificially crystallizing the unrealized capital gains when the government changed the capital gains exemption rules.  

Norman, perhaps I am being simplistic, but I prefer simple. Note that I hold this investment in a registered account, so capital gains do not come into play here.

I use this simple idea for my investments:
total % gain/loss = {(current value of investment+total cumulative dividends-ROC)/(cost base)}-1.
Cost base of course is the total paid for the investment minus ROC.

This can then be calculated over a time period for an annualized compound value.

rhvic said

Norman, perhaps I am being simplistic, but I prefer simple. Note that I hold this investment in a registered account, so capital gains do not come into play here.

I use this simple idea for my investments:
total % gain/loss = {(current value of investment+total cumulative dividends-ROC)/(cost base)}-1.
Cost base of course is the total paid for the investment minus ROC.

This can then be calculated over a time period for an annualized compound value.

Unfortunately, that formula gives a number that has nothing to do with rate of return.

When the cumulative ROC distributions approach the original investment, the "cost base" approaches zero and the formula blows up trying to divide by zero.

The formula also gives different numbers when the ROC portion of the distributions changes. That makes no sense. Rate of return is the same when one receives $4 of distributions regardless if they were $4 of income + $0 of ROC or $1 of income + $3 of ROC.

topgun said

I know the original cost of my RRSP bank shares in 1993. $27.875 They split twice 2 for 1. I bought more over the years. Of course I sold some of them. They have a high rate of return compared to my expectations today. Somewhere about 12% annually. Once I get my RRIF I will be selling some annually. Merry Christmas.  

The rate of return will be quite meaningless

Norman1 said

rhvic said

Norman, perhaps I am being simplistic, but I prefer simple. Note that I hold this investment in a registered account, so capital gains do not come into play here.

I use this simple idea for my investments:
total % gain/loss = {(current value of investment+total cumulative dividends-ROC)/(cost base)}-1.
Cost base of course is the total paid for the investment minus ROC.

This can then be calculated over a time period for an annualized compound value.

Unfortunately, that formula gives a number that has nothing to do with rate of return.

When the cumulative ROC distributions approach the original investment, the "cost base" approaches zero and the formula blows up trying to divide by zero.

The formula also gives different numbers when the ROC portion of the distributions changes. That makes no sense. Rate of return is the same when one receives $4 of distributions regardless if they were $4 of income + $0 of ROC or $1 of income + $3 of ROC.  

The rate of return in % will be quite meaningless as Norman pointed out. But I think what RHVIC is trying to measure is how well the "true" return of the investment is. In other words, if the ROC nets out the original cost to zero, and still return $X dividend each year, that the asset is performing. But it cant be quantity as a true % return, esp when the ROC embedded in the dividend reduce the cost annually, so the %return changes every year. And on top, what about the reinvestment risk / return on the ROC ?

I agree it's a bit obscure, topgun, because I don't really pay attention to the various forums, I just come to the home page to see what's been posted lately on the various threads (I'm assuming like most people) and from that view you can't tell how to start a new thread. The key, I think, is to first select "Forum" at the top of the home page. (But I never end up starting any new threads because every few months when the urge hits I've forgotten how to do it again!)

Calculate the internal rate of return (IRR). That's what the rate on a savings account would have to be for the same investments, same payouts, and same ending balance. IRR also uses the timings so that it accounts for when the investments and payouts occur.

Spreadsheets function XIRR() that can find the IRR for an irregular sequence of cash flows.

One needs to look at the financial statements to determine whether or not the ROC part of distributions is a return of original investment. It actually doesn't matter for IRR calculations. Any return of the original investment will be reflected in the payouts and the ending value.

If one invests $5, gets a $1 distribution, and gets only $2 when the investment is liquidated, the investment return is the same, regardless of whether that $1 distribution was income or return of capital.

Distribution rate and actual rate of return differences are quite common. Lots of bonds, for example, have that because they trade above face value from the drop in interest rates.

If one pays 102% of face value for a bond, then one will have a lower return than what is calculated by simply dividing the interest payments by the purchase price. There will be a loss at maturity from paying 102% of face value now and receiving only 100% of face value at maturity.

XIRR() may give the same results as the way I did it. I will investigate further.

I tried XIRR(). With XIRR() it uses the actual dates. Jan 2020 has 31 days. Each month has a different number of days. In my model I assumed that each month is 1/12 year. The function XIRR() gets slightly different value. Not going to affect me. Cool.

In order to do the IRR calculation, one needs the date and amount of each investment and each distribution.

The total distributions and the time period are not enough.

Thanks Norman. I checked my spreadsheet, and if I look at one more significant figure in my compound annual return, I too had 0.46% as my rate of return - in agreement with your method exactly (0.455% if you want three significant figures). So my formula appears also to have some validity, as long as cost base is greater than zero.