8

Why do investors offset their annual returns against inflation, when inflation isn't realised until the end of the holding period (when the investor is actually spending the money)?

An example:

Assume the following:

  • Stock with a 10% annualised return.
  • Stock receives 50% of its growth from dividends - these are 'immediately' re-invested.
  • Holding period is 50 years.
  • Inflation is 3% per year.

Most people appear to calculate the inflation-adjusted return as:

Inflation Adj. Return = (1 + (0.10 - 0.03))^50 = 29.457

Another method, which makes more sense to me (with my current understanding), is as follows:

Nominal Return = (1 + 0.10)^50 = 117.391
Inflation = (1 - 0.03)^50 = 0.218
Inflation Adj. Return = 117.391 * 0.218 = 25.599

The results are very different.

Method 2 only realises inflation at the end of the holding period, but still accounts for the compounding effect against the dollar over time.

Method 2 also assumes "spending" your dividends by re-investing in the same stock is not affected by inflation. The rational is since you are purchasing the same "product" as the one you've just sold, so there's no realization of any loss or gain. There is, of course, some cash drag, brokerage fees and taxes - but no/minimal inflation (to my understanding).

Why is method 1 then used instead of method 2?

Lawrence Wagerfield
  • 565
  • 4
  • 10
  • Holding period isn't necessarily a year or finishes at the end of the year. Secondly if somebody wants to modify his portfolio, he needs to understand how his portfolio is performing so he can rearrange his holdings. – DumbCoder Dec 09 '15 at 14:17
  • In this case, we're assuming a multi-year investment: 50 years, specifically. Re. your second comment: that's fine, but my question is: why would they use method 1 to do that, rather than method 2? Assuming you're only 4 years into your investment - you can still calculate inflation-adjusted earnings using method 2 instead of method 1. My question is: why wouldn't you? – Lawrence Wagerfield Dec 09 '15 at 14:35
  • 2
    if you're using an annual inflation rate over a time period of more than a year you need to take into account that it is compounded; a 1% inflation rate is the change of prices over the last year so to cover 2 years you must either use multiple inflation rates or compound the average rate. – MD-Tech Dec 09 '15 at 15:17
  • My question takes compounding into account in both cases - did I miss something (or did you)? Notice the ^50 in the calculations - that's for the 50 year holding period. – Lawrence Wagerfield Dec 09 '15 at 15:48

1 Answers1

7

I would use neither method. Taking a short example first, with just three compounding periods, with interest rate 10%. The start value y0 is 1.

y0 = 1;
y1 = (1 + 0.1) y0;
y2 = (1 + 0.1) y1;
y3 = (1 + 0.1) y2 = 1.331

So after three years the value is 1.331, the same as y0 (1 + 0.1)^3.

Depreciating (like inflation) by 10% (to demonstrate) gets us back to y0 = 1

y2 = y3/(1 + 0.1);
y1 = y2/(1 + 0.1);
y0 = y1/(1 + 0.1) = 1

Appreciating and depreciating by 10% cancels out:

y0 = 1;
y1 = (1 + 0.1) y0/(1 + 0.1);
y2 = (1 + 0.1) y1/(1 + 0.1);
y3 = (1 + 0.1) y2/(1 + 0.1) = 1

Appreciating by 10% interest and depreciating by 3% inflation:

y0 = 1;
y1 = (1 + 0.1) y0/(1 + 0.03);
y2 = (1 + 0.1) y1/(1 + 0.03);
y3 = (1 + 0.1) y2/(1 + 0.03) = 1.21805

This is the same as y0 (1 + 0.1)^3 (1 + 0.03)^-3 = 1.21805

So for 50 years the result is y0 (1 + 0.1)^50 (1 + 0.03)^-50 = 26.7777

Note

You can of course use subtraction but not using the inflation figure directly. E.g.

x = 0.03 (1 + 0.1)/(1 + 0.03) = 0.0320388

y0 (1 + (0.1 - x))^50 = 26.7777

(edit: This appears to be the Fisher equation.)

2nd Note

Further to comments, here is a chart to illustrate how much the relative performance improves when inflation is accounted for. The first fund's return is 6% and the second fund's return varies from 3% to 6%. Inflation is 3%.

enter image description here

Chris Degnen
  • 9,797
  • 1
  • 20
  • 36
  • Great answer, thanks Chris. Will re-read over then next few days and no doubt mark as the answer :) – Lawrence Wagerfield Dec 09 '15 at 16:16
  • Interestingly, John Bogle seems to disagree with both your method and my 2nd method. He states "deducting the same inflation rate from both figures (the CAGRs) further increases the comparative advantage of the investment with the higher return" (taken verbatim). According to Bogle, inflation has an exponential effect on the HPR that is relative to the annual investment return. This precludes you from deducing an 'inflation coefficient' for some period that can be applied to any investment of the same period (which is what both your method and my 2nd method imply). Could Bogle be wrong? – Lawrence Wagerfield Dec 09 '15 at 16:33
  • Simply put, he gives an example that makes 'equity Y' 57% better than 'equity Z' after inflation. However, before inflation, 'equity Y' is only 52% better than 'equity Z'. In both our methods (my method 2 and your method) the comparison would be unaffected by inflation. – Lawrence Wagerfield Dec 09 '15 at 16:43
  • Assuming a 50 year holding period, without inflation: 6% fund (1 + 0.06)^50 = 18.4202. 4% fund (1 + 0.04)^50 = 7.1067. That's a 38.58% relative difference. With 3% inflation: 6% fund (1 + 0.06)^50 * (1 + 0.03)^-50 = 4.2018. 4% fund (1 + 0.04)^50 * (1 + 0.03)^-50 = 1.6211. That's still a 38.58% relative difference. What am I missing? (The 6% fund did not perform any better, relatively speaking, with or without inflation, using the method you described.) I'm aware I must be missing something! – Lawrence Wagerfield Dec 09 '15 at 17:18
  • 1
    Without inflation, 6% fund gain divided 4% fund gain: (((1 + 0.06)^50) - 1)/(((1 + 0.04)^50) - 1) = 2.85264 times better. With inflation: (((1 + 0.06)^50 * (1 + 0.03)^-50) - 1)/(((1 + 0.04)^50 * (1 + 0.03)^-50) - 1) = 5.15512 times better. – Chris Degnen Dec 09 '15 at 21:00
  • I see now - the -1 constant that converts the change factor back to a relative change makes all the difference. Thank you for persisting! – Lawrence Wagerfield Dec 10 '15 at 07:16
  • I'm sorry, I still don't see why we'd want to compare the relative change between two funds, as opposed to the change factor (e.g. 6% vs 4% as opposed to 106% vs 104%). Specifically, if the lower fund's return was 3%, then we'd be dividing by 0 with this method of comparison. This does not compute: (((1 + 0.06)^50 * (1 + 0.03)^-50) - 1)/(((1 + 0.03)^50 * (1 + 0.03)^-50) - 1) == divide by zero. – Lawrence Wagerfield Dec 10 '15 at 09:35
  • (p.s. I am happy to ask this as a separate question - please up vote comment and I will do so) – Lawrence Wagerfield Dec 10 '15 at 09:43
  • I've added a chart to illustrate. – Chris Degnen Dec 10 '15 at 09:59
  • I've asked a separate question for this particular query here: http://money.stackexchange.com/questions/56889/why-compare-asset-performance-using-returns-instead-of-using-their-change-facto – Lawrence Wagerfield Dec 10 '15 at 14:20