Not sure what to use for predicting investment returns!

[CantThinkofAUserName]

I'm working on calculating my target RE year. I've read a ton of messages on this board, but of course not all! I've finally gotten my health insurance budget to a point I'm comfortable with, which was the one part of my budget I was having trouble with. So now I'm trying to figure what return on investment to use in my calculations. I figure I should expect a slightly larger return on investment accounts that wouldn't be touched for at least 25 years, since I would have none of that in cash, and probably a 70/30 or 75/25 allocation. And I would expect a smaller return on $ that I would be using sooner, since I'd have some in cash, and the rest would probably be 65/35 or 60/40. But I still can't settle on what figures to use. 6% for 25 yr+ and 4% or the shorter? What do you all use? I don't want to be too conservative nor too optimistic. I have used FireCalc also, but want to do my own calculations as well.

 

[justplainbll]

You might want to consider using inflation adjusted (real) rates of return and the effect of the tax bite on returns in non-TDAs.

 

[Coolius]

Since I feel that we will be in a low interest environment for some time, and I definitely do not want to put too much allocation into stocks, I am using a 4% return for my planning.

I tend to be more conservative, so have about 30% in cash earning almost zilch, 40% in bonds, and the rest in stocks ( Mostly preferred issues ).

The disadvantages are that I probably will not participate in any market recovery, but the advantage is security and less anxiety when the volatility is high, like now.

 

[youbet]

Whatever percentages you use, don't make the mistake of using that percentage as an average you get every year. Early losses are more harmful than later losses and early gains are more helpful than later gains........ even if in all cases the average annual return is the same.

 

[utrecht]

Whatever percentages you use, don't make the mistake of using that percentage as an average you get every year. Early losses are more harmful than later losses and early gains are more helpful than later gains........ even if in all cases the average annual return is the same.

Isnt that only true in the withdrawal phase? During accumulation, it works out the same no matter when your gains and losses come.

 

[rescueme]

Isnt that only true in the withdrawal phase?

In the withdrawl phase (where I've been in since early '07) it's the same as the accumulation phase, and actually a bit easier to manage overall.

I have no need to max out my returns since prior to retirement I hit my "number". Additionally, since I/DW have more than several years in cash (to provide current retirment income), the current action of the market means little.

We simply sell when the market is up (as it was, earlier this year) and hold tight when the market is down (which it is now, and may be many years in the future - who knows?)

Unless you are a person who keeps their retirement investments in the market (both equity/bonds) and draw directly from it to meet your immediate income needs, or are a bit short in your retirement investments overall, market flux need not be as great an immediate impact to your plan, IMHO.

 

[Animorph]

Isnt that only true in the withdrawal phase? During accumulation, it works out the same no matter when your gains and losses come.

That's only true if no money is added or subtracted from the portfolio for the entire period. During accumulation you are most likely regularly adding to the portfolio, in which case the value will be most sensitive to the market returns near retirement, which are applied to more of your contributions. Similarly, its final value at the end of retirement will be most sensitive to the market returns at the start of retirement, which are applied to more of your money before it is withdrawn.

 

[utrecht]

In the withdrawl phase (where I've been in since early '07) it's the same as the accumulation phase, and actually a bit easier to manage overall.

I have no need to max out my returns since prior to retirement I hit my "number". Additionally, since I/DW have more than several years in cash (to provide current retirment income), the current action of the market means little.

We simply sell when the market is up (as it was, earlier this year) and hold tight when the market is down (which it is now, and may be many years in the future - who knows?)

Unless you are a person who keeps their retirement investments in the market (both equity/bonds) and draw directly from it to meet your immediate income needs, or are a bit short in your retirement investments overall, market flux need not be as great an immediate impact to your plan, IMHO.

I think you misunderstood the question. Someone mentioned that if you are planning for an avg of 6% per year, getting 6% EVERY year is not the same as 9%, 3%, -10%, 16%..ect (or whatever averages out to 6% per year). Big losses in the beginning hurt worse and raise your chances of running out of money. But I'm pretty sure that during the accumulation phase it makes no difference in what order your returns come.

 

[CantThinkofAUserName]

Hmm, start with $100.

First year 3% loss, second year 9% gain (avg 6% gain): $105.73

First and second years 6% gain: $112.36

First year 9% gain, second year 3% loss: $105.73 Ok I'm surprised here, how did this end up the same?

Is something wrong with my math or logic here? I would have thought either all 3 would be the same, or the third would have been higher than the other 2?

 

[Animorph]

Hmm, start with $100.

First year 3% loss, second year 9% gain (avg 6% gain): $105.73

First and second years 6% gain: $112.36

First year 9% gain, second year 3% loss: $105.73 Ok I'm surprised here, how did this end up the same?

Is something wrong with my math or logic here? I would have thought either all 3 would be the same, or the third would have been higher than the other 2?

That's correct. The order of returns makes no difference if you never add or subtract from the portfolio. Which is a different problem from arithmetically averaging the returns and then applying the average.

 

[justplainbll]

Hmm, start with $100.

First year 3% loss, second year 9% gain (avg 6% gain): $105.73

First and second years 6% gain: $112.36

First year 9% gain, second year 3% loss: $105.73 Ok I'm surprised here, how did this end up the same?

Is something wrong with my math or logic here? I would have thought either all 3 would be the same, or the third would have been higher than the other 2?

It would help if you bone up on the concept of internal rates of return (IRR).

 

[CyclingInvestor]

Hmm, start with $100.

First year 3% loss, second year 9% gain (avg 6% gain): $105.73

First and second years 6% gain: $112.36

First year 9% gain, second year 3% loss: $105.73 Ok I'm surprised here, how did this end up the same?

Is something wrong with my math or logic here? I would have thought either all 3 would be the same, or the third would have been higher than the other 2?

First year 3% loss, second year 9% gain = avg 3% gain

 

[MichaelB]

Average annual return is much different that annualized return. The first is the simple arithmetic average of all the yearly returns. The second is real return after the effects of volatility. The greater the volatility, the bigger the difference between the two.

Consider to sequences of returns. At 6% per year for 5 years, $100 grows to $133.8. With yearly returns of 10%, 10%, -20%, 0, 30%, the annual average is still 6%, but the portfolio only grows to $125.8 - a much lower annualized rate of return.

 

[CantThinkofAUserName]

Ok thanks everyone, I need to think about this some more. Now I'm even further from knowing how in the world to predict the future LOL. Initially when starting this thread I thought my main problem was trying to zero in on whether to assume, say 4%, 5%, or 6% avg returns per year. Now it turns out, not only that, but I need to also consider the situation where things are bad for the first bunch of years, then good later on. Uggh.

Oh, and CyclingInvestor, thanks for pointing out the error in my math, how embarrassing!

 

[jebmke]

Just make up a number and run the calculations. Nobody knows what future returns will be. My base plan had something like ~ 4% expected return (real) but right now, with the situation in Europe, I am thinking zero for the next 20 years or so is more likely.

Whatever you choose for your "expected return" - see how you would fare with a bad situation (zero or some similarly lousy real return). If you are eating pine cones and road pizza in the extreme situation, re-think your plan.

One last thought. Do your numbers in real dollars -- even if you have to put escalators on things that inflate above normal inflation. Nominal dollars are illusory.

 

[clifp]

Ok thanks everyone, I need to think about this some more. Now I'm even further from knowing how in the world to predict the future LOL. Initially when starting this thread I thought my main problem was trying to zero in on whether to assume, say 4%, 5%, or 6% avg returns per year. Now it turns out, not only that, but I need to also consider the situation where things are bad for the first bunch of years, then good later on. Uggh.

Oh, and CyclingInvestor, thanks for pointing out the error in my math, how embarrassing!

It is a difficult thing to try and figure out sequences of returns. That is why there is this really handy tool FIRECalc which measure the impact performance of a historical sequence of returns will have over the last 100+ years. Now if you thing that future will look nothing like the past then it is not a useful tool, but it is easier than trying figure not only what reasonable future return should be but trying to predict the sequencing.

To be honest I think you'll driving yourself crazy trying to do the later.

 

[utrecht]

Average annual return is much different that annualized return. The first is the simple arithmetic average of all the yearly returns. The second is real return after the effects of volatility. The greater the volatility, the bigger the difference between the two.

Consider to sequences of returns. At 6% per year for 5 years, $100 grows to $133.8. With yearly returns of 10%, 10%, -20%, 0, 30%, the annual average is still 6%, but the portfolio only grows to $125.8 - a much lower annualized rate of return.

Whats even worse is this:

Start with $1,000,000 and with draw 4% ($40,000) at the beginning of each year. See what happens under these 2 scenarios:

Year 1: -15%.
Year 2: -10%
Year 3: 18%
Year 4: 12%

Scenario 2

Year 1: 12%
Year 2: 18%
Year 3 -10%
Year 4: -15%

Exact same returns but in reverse order. The avg return is the same obviously. The result would be exactly the same if you weren't withdrawing, but when you withdraw money and have bad returns at the very beginning its much worse than having those same bad returns at the end of your retirement life span.

Scenario #1 ends 4 years with $825,431
Scenario #2 ends 4 years with $869,874

Thats more than a 5% difference in what you are left with. Imagine what could happen over 20 years if two people have the same returns but in a different order.

 

[chemist]

What you are trying to do manually is best left to calculators like FIREcalc and Fidelity Retirement Income Planner.

Assuming a constant return is folly. Google "retirement planner from hell" by Bill Bernstein.

 

[palomalou]

I've been using 3% in ORP--a good conservative estimate? Or still too high?

 

[Chuckanut]

i have used several of the Monte Carlo calculators including FireCalc. All of them give me a 100% survival rate over 30 years if I choose a 3% withdrawal rate. However 4% moves it to about 90%, still not bad. The biggest risk is that I live more than 30 years past retirement. To help with that, I am waiting to take SS until I am 66 when I can get the full amount, or, if the market is doing well, I may even take SS later than that.

Of course, unexpected events could sabatoge even the best plans. An exteneded serious illness, a complete market melt-down that lasts for years, the state could renege on my pension, or the Feds could cut SS. One can't worry about everything.

 

[REWahoo]

i have used several of the Monte Carlo calculators including FireCalc.

My apology if you are already aware of this, but FIRECalc is not a Monte Carlo calculator. The real value of FIRECalc is that it uses actual market performance history (although it does have a Monte Carlo option if you select the "random performance" button on the "Portfolio Perfomance" tab). By using actual history...

If you get a "100% success rate" with what you have and what you plan to spend, this means that you would have been able to maintain your standard of living and not run out of money, despite the worst that we've ever seen, including the Great Depression.

For more information see How It Works.

 

[Rustward]

I've been using 3% in ORP--a good conservative estimate? Or still too high?

What are you using for inflation?