What is the prediction of future long-term returns for your portfolio?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
flyingaway
Posts: 1885
Joined: Fri Jan 17, 2014 10:19 am

What is the prediction of future long-term returns for your portfolio?

Post by flyingaway » Tue Feb 07, 2017 12:12 pm

When many people talk about retirement planning, they use the 4% rule or similar things based on backtesting data. However, it is more informative if we can have an idea on what is the predictions of future long-term returns for a typical portfolio, e.g., 75/25 (stock/bond). I know this board had a few threads on this topics and some people did make (or report) some predictions.

There is a recent article on MSN: Despite risks, public pensions put faith in long-term returns http://www.msn.com/en-us/money/markets/ ... li=BBnbfcL

"The California Public Employees' Retirement System, the largest U.S. public pension fund, anticipates annual returns of 6.2 percent over the next decade.

However, CalPERS still expects its long-term return to align more closely with a discount rate that it plans to reduce to 7 percent by 2020, because it anticipates returns will jump to 7.83 percent in the decades to follow."

I thought pension funds are mostly invested conservatively (they do invest in hedge funds). So if they anticipate a long-term return between 6% and 8%, that is much better (rosier) than many predictions that I read.

Any thoughts?

User avatar
Cyclesafe
Posts: 658
Joined: Wed Dec 31, 2014 1:03 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by Cyclesafe » Tue Feb 07, 2017 12:33 pm

First of all nobody can predict future performance, especially to three significant digits. Pension funds project growth rates that are supposedly politically palatable, not those that have any chance whatsoever of being achieved.

So what works for us?

1) Postulate your retirement date and expected asset needs at that date.
2) Propose your annual contributions to savings and guess earnings for fixed and equity investment allocations.

If #1 and #2 don't jive, repeat with different dates, needs, savings, and earnings.

Repeat until #1 equals #2.

As long as earnings expectations are reasonable, you are done. If not, start over.

To start, I'd suggest a 1% real rate for fixed and 3% real rate for equity. Don't forget taxes.

If your investments fare better than this be happy.

zwzhang
Posts: 54
Joined: Thu Jul 25, 2013 10:43 am

Re: What is the prediction of future long-term returns for your portfolio?

Post by zwzhang » Tue Feb 07, 2017 12:37 pm

Here is my two cents.
It depends on how long we are looking. For example: For US total market, the very long-term return = 1.9 (current VTI dividend yield) + 1.5 (history dividend real growth rate) = 3.4% real return.

Now we will look at current valuation. CAPE 10 is 30% above its history mean. If we annualize it by 20 years, we got 1.3% per year. So for the next 20 year: (20 yr)long-term return = 1.9 (current VTI dividend yield) + 1.5 (history dividend real growth rate) - 1.3 (valuation) = 2.1% real return.

Thought?

flyingaway
Posts: 1885
Joined: Fri Jan 17, 2014 10:19 am

Re: What is the prediction of future long-term returns for your portfolio?

Post by flyingaway » Tue Feb 07, 2017 1:03 pm

Based on a retirement calculator, if we retire at 55, we will run out of money at 93 with a 2% real return. For me, that is not too bad.
But the curve with a 3% real return looks really nice, and might be used to convince my wife that we can retire.
Last edited by flyingaway on Sat Feb 11, 2017 5:14 pm, edited 1 time in total.

garlandwhizzer
Posts: 1973
Joined: Fri Aug 06, 2010 3:42 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by garlandwhizzer » Tue Feb 07, 2017 1:38 pm

Just my 2 cents worth. If those returns are real and inflation adjusted I believe CALPERS predictions over the next decade or so are markedly optimistic. If they are nominal returns (which I suspect) I believe they're mildly optimistic. I have a 2/3 equity 1/3 bond portfolio and my rather conservative expectation is for about 3% real over the next decade or so. I'd rather err on the pessimistic side in making that prediction and I suspect CALPERS prefers to err on the optimistic side for political reasons. Having said that, it's important to remember that predictions of future market returns are notoriously inaccurate no matter who makes them. Get a good investment plan and stick to it--that's about all we can do.

Garland Whizzer

User avatar
Portfolio7
Posts: 430
Joined: Tue Aug 02, 2016 3:53 am

Re: What is the prediction of future long-term returns for your portfolio?

Post by Portfolio7 » Tue Feb 07, 2017 3:07 pm

I've read several articles about unrealistic pension fund return assumptions. There is an incentive to overstate expected return; it reduces the amount of funding the organization is required to contribute. :greedy

I admit, I'm all over the map when it comes to returns. Long term nominal CAGR has been about 9% for us, though lower recently... so as our portfolio slowly shifts to bonds there is a small reduction in return, and if CAPE returns to the mean, it could mean a 2-4% drop in medium range nominal (10 years)... which puts us anywhere from 3%n to 7%n, but probably 4 or 5%n (i.e. ~ 2-3% real) Long term, I expect perhaps 7-8% nominal on that basis, but now we are talking 40 years.

Problem is, nobody knows. Sequence of Returns makes a difference too. To be honest, I am not all that sure that Grantham isn't right - maybe CAPE has shifted a bit. I don't see low interest rates going away, even with some MBS sales, but we'll see. I go back to the boglehead thing. I set my risk level, and stay the course. The real opportunities are easy to miss if I'm focused on the minutia.
An investment in knowledge pays the best interest.

itstoomuch
Posts: 5343
Joined: Mon Dec 15, 2014 12:17 pm
Location: midValley OR

Re: What is the prediction of future long-term returns for your portfolio?

Post by itstoomuch » Tue Feb 07, 2017 3:28 pm

Oregon PERs assumes a 7.5% growth.
Oregon State Budget is negative 1.5B because of PERs, not to mention local and school district participation in PERs. The budget assumes constant spending and increasing tax revenues. OR is #1 or #2 for new residents.
I'm betting on 4-6%.
Targeting 2017's Discretionary for 5% but I'm 50% cash and the remainder in conservative equity. No bonds.
Rev012718; 4 Incm stream buckets: SS+pension; dfr'd GLWB VA & FI anntys, by time & $$ laddered; Discretionary; Rentals. LTCi. Own, not asset. Tax TBT%. Early SS. FundRatio (FR) >1.1 67/70yo

heyyou
Posts: 3131
Joined: Tue Feb 20, 2007 4:58 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by heyyou » Wed Feb 08, 2017 12:28 am

You are too late to get a prediction, we are fresh out of them.

So far, the future is often slightly different from the past, in unexpected ways. Of the few who were right on a prediction, they noticeably failed on followup attempts. Many of those with predictions have an agenda that they want to share.

I manage a retirement plan with two recipients. Consider saving until your portfolio is 25 times your current spending to cover 30 years of retirement. Consider choosing a retirement withdrawal(WD) method that is based on your recent annual portfolio value, not one based solely on retirement day asset level. Store a decade of retirement spending in bonds, to use if stock prices are depressed. Expect to adapt to your portfolio WDs, you already have 30 years of experience at living within your income. Try to have most of your necessary expenses covered by steady income sources, so your discretionary spending is from those variable portfolio WDs.

No thanks, I don't need a prediction today.

User avatar
Pajamas
Posts: 6015
Joined: Sun Jun 03, 2012 6:32 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by Pajamas » Wed Feb 08, 2017 12:35 am

Very conservative assumptions about the future make me feel more secure: "Plan for the worst, hope for the best."

I expect my investment returns to track inflation when planning and hope that they do much better. I also assume that I will live to 90 when planning and hope that I live longer and in good health, although statistics and family history both point indicate a shorter rather than longer life.

User avatar
TinkerPDX
Posts: 276
Joined: Sun Jun 08, 2014 4:38 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by TinkerPDX » Wed Feb 08, 2017 12:53 am

For forecasting purposes for equities, I use expected earnings. So for instance, for VTI, the current price/prospective earnings is 18.86, and 1/18.86 = 5.3%.

For forecasting bond returns, I use the yield. For BND, that's about 2.41% right now.

Imperfect, but I need something for planning. Factor in 2% inflation (fed target), and that leaves about 3.3% real return for US equity and 0.4% for US bond. Seems reasonable enough.

Every once in a while I update those values for each slice of my equity portfolio and each element of my bond portfolio and it spits out a total expected return for the equity portion of my portfolio based on the allocations.

From the last time I updated it (maybe a month ago), my US equity was at 5.33% nominal, my ex-US equity was 7.4% nominal, and my fixed income was 1.69% nominal, resulting in a total "expected" nominal return for 2017 of 5.27% with a 77-23 AA. I don't really "expect" it, but I need a number, so that's the one I'm using.

Nate79
Posts: 3477
Joined: Thu Aug 11, 2016 6:24 pm
Location: Portland, OR

Re: What is the prediction of future long-term returns for your portfolio?

Post by Nate79 » Wed Feb 08, 2017 1:20 am

TinkerPDX wrote:For forecasting purposes for equities, I use expected earnings. So for instance, for VTI, the current price/prospective earnings is 18.86, and 1/18.86 = 5.3%.

For forecasting bond returns, I use the yield. For BND, that's about 2.41% right now.

Imperfect, but I need something for planning. Factor in 2% inflation (fed target), and that leaves about 3.3% real return for US equity and 0.4% for US bond. Seems reasonable enough.

Every once in a while I update those values for each slice of my equity portfolio and each element of my bond portfolio and it spits out a total expected return for the equity portion of my portfolio based on the allocations.

From the last time I updated it (maybe a month ago), my US equity was at 5.33% nominal, my ex-US equity was 7.4% nominal, and my fixed income was 1.69% nominal, resulting in a total "expected" nominal return for 2017 of 5.27% with a 77-23 AA. I don't really "expect" it, but I need a number, so that's the one I'm using.
For your method of the return for equity how has it compared to historical returns?

User avatar
Johnnie
Posts: 509
Joined: Sat May 28, 2016 3:18 pm
Location: Michigan

Re: What is the prediction of future long-term returns for your portfolio?

Post by Johnnie » Wed Feb 08, 2017 11:53 am

Pajamas wrote:Very conservative assumptions about the future make me feel more secure: "Plan for the worst, hope for the best."
That is also my instinct, but I also don't want to get goofy about it to the point of becoming unrealistic in the other direction.

For example, when I hear middle aged people opine that, "Social Security will be gone by the time I get there," I often suggest that maybe they shouldn't count on getting it all, but it just as unrealistic or more to assume zero. Current payroll taxes cover about 3/4 of current benefit payments (the rest comes from the 'lock box' of course :wink: ), so use that amount for a conservative assumption - not zero.
"I know nothing."

User avatar
Pajamas
Posts: 6015
Joined: Sun Jun 03, 2012 6:32 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by Pajamas » Wed Feb 08, 2017 12:12 pm

Johnnie wrote:
Pajamas wrote:Very conservative assumptions about the future make me feel more secure: "Plan for the worst, hope for the best."
That is also my instinct, but I also don't want to get goofy about it to the point of becoming unrealistic in the other direction.

For example, when I hear middle aged people opine that, "Social Security will be gone by the time I get there," I often suggest that maybe they shouldn't count on getting it all, but it just as unrealistic or more to assume zero. Current payroll taxes cover about 3/4 of current benefit payments (the rest comes from the 'lock box' of course :wink: ), so use that amount for a conservative assumption - not zero.
I do get goofy about, it per your standards. There have been several changes in social security during my life, none of them in my favor, and have no doubt that there will be several more. Full social security benefits are very far off for me and will be relatively low compared to what many people receive even if nothing changes. In addition, the COLA increases have been insufficient so the benefit is effectively decreasing. So I ignore social security in calculations and think of them as a potential but not guaranteed bonus that will be used for discretionary spending (hope for the best) or assistance with daily living (cleaning service to skilled nursing facility (plan for the worst).

flyingaway
Posts: 1885
Joined: Fri Jan 17, 2014 10:19 am

Re: What is the prediction of future long-term returns for your portfolio?

Post by flyingaway » Wed Feb 08, 2017 12:26 pm

There is another (older) thread talking about similar topics. But I think I am more interested in knowing if we may use the numbers used by pension funds to roughly predict the future long-term returns. For almost all retirement calculations, we do need at least one number, at least for planning purpose.

For my own planning, I do not include social security. I think if a pension fund fails, they can ask the government to increase tax. My backup strategy is my social security.

CurlyDave
Posts: 699
Joined: Thu Jul 28, 2016 11:37 am

Re: What is the prediction of future long-term returns for your portfolio?

Post by CurlyDave » Wed Feb 08, 2017 12:35 pm

I think public pension funds are not a very good place to look for predicted future returns.

The reason is that by predicting high investment returns they can reduce current required contributions without appearing underfunded with respect to projected future liabilities. This reduces their demands on the public budget and makes politicians look better.

The magnitude of this effect is hard to determine, but it is certainly not zero.

OutInThirteen
Posts: 381
Joined: Thu Mar 14, 2013 2:11 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by OutInThirteen » Thu Feb 09, 2017 9:05 am

In my Excel spreadsheet I assume 6% for stocks and 4% for inflation, resulting in 2% real. However, the information isn't particularly useful to me. COLA'd pensions and SS cover all expenses, with investments set aside for medical and/or long term care expenses. Whatever's left will go to the children.

petulant
Posts: 325
Joined: Thu Sep 22, 2016 1:09 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by petulant » Thu Feb 09, 2017 9:46 am

Pajamas wrote:
Johnnie wrote:
Pajamas wrote:Very conservative assumptions about the future make me feel more secure: "Plan for the worst, hope for the best."
That is also my instinct, but I also don't want to get goofy about it to the point of becoming unrealistic in the other direction.

For example, when I hear middle aged people opine that, "Social Security will be gone by the time I get there," I often suggest that maybe they shouldn't count on getting it all, but it just as unrealistic or more to assume zero. Current payroll taxes cover about 3/4 of current benefit payments (the rest comes from the 'lock box' of course :wink: ), so use that amount for a conservative assumption - not zero.
I do get goofy about, it per your standards. There have been several changes in social security during my life, none of them in my favor, and have no doubt that there will be several more. Full social security benefits are very far off for me and will be relatively low compared to what many people receive even if nothing changes. In addition, the COLA increases have been insufficient so the benefit is effectively decreasing. So I ignore social security in calculations and think of them as a potential but not guaranteed bonus that will be used for discretionary spending (hope for the best) or assistance with daily living (cleaning service to skilled nursing facility (plan for the worst).
Johnnie's right that with current numbers SS should pay out about 75% of entitled benefits starting in a couple decades. You're right that social security might get worse than that, and it would be completely unreasonable to plan for 100% if retirement includes any period after 2035.

That said, I still think it's excessively cautious to plan for 0% of social security benefits. Maybe dropping the 75% to 50% would be more realistic. Personally, I've done a range from 50% to 75% in my Monte Carlo simulations.

User avatar
Sheepdog
Posts: 5156
Joined: Tue Feb 27, 2007 3:05 pm
Location: Indiana, retired 1998 at age 65

Re: What is the prediction of future long-term returns for your portfolio?

Post by Sheepdog » Thu Feb 09, 2017 10:15 am

Let's see. My average nominal return for the last 18 years (1999-2016) was 5.31% with a very conservative in-retirement stock/bond ratio and with up and down markets. I'll predict that for my portfolio going forward which will take me to age 101 :shock: .

Amended prediction. For the last 12 years (2005-2016) (which eliminates the 2000-2002 down period) with my 22% stock level in that time period (I had more stock before that) my return was a higher nominal 5.91% so I'll take that as my prediction for me for the next 18 years to 2034!!!
It's not what you gather, but what you scatter which tells what kind of life you have lived---Helen Walton

User avatar
willthrill81
Posts: 5741
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: What is the prediction of future long-term returns for your portfolio?

Post by willthrill81 » Thu Feb 09, 2017 3:50 pm

This calculator is a very useful one for examining time periods in stocks.
https://dqydj.com/sp-500-historical-return-calculator/

I find several things notable from examining its results. Contrary to what a few suggest, volatility over the entire period being investigated (not year to year) decreases substantially as the time frame increases. Just to be clear, you could, historically speaking, lose 50% of your investment in equities at any given time. But over a 10, 20, 30, or 40 year period, the likelihood of losing money from beginning to end quickly plummets. I personally think that this dispels the notion that stocks' best days are behind them because there have been many times historically when their performance was lackluster, not just in some recent periods.

Minimum Real Annual Returns from S&P 500
10 yr. -5.9%
20 yr. -.2%
30 yr. 1.9%
40 yr. 3.2%

But that doesn't tell the whole story, just the worst of the worst. Another useful measure is the percentile real return for each period.

Percentile (Pct. Beating) 40 Years 30 Years 20 Years 10 Years
90 % 4.875 % 4.444 % 2.597 % -0.675 %
80 % 5.291 % 5.000 % 3.672 % 2.454 %
70 % 5.530 % 5.537 % 5.339 % 4.327 %
60 % 5.914 % 6.089 % 6.048 % 5.561 %
50 % 6.352 % 6.606 % 6.886 % 6.728 %
40 % 6.736 % 7.135 % 7.635 % 8.231 %
30 % 7.114 % 7.680 % 8.356 % 9.740 %
20 % 7.456 % 8.204 % 9.129 % 11.039 %
10 % 8.211 % 8.717 % 10.925 % 13.932 %

So for a 20 year period, the S&P 500 achieved at least a 2.6% real annual return 90% of the time. For a 30 year period, it returned at least 4.4% 90% of the time.

Is it necessary to assume the worst case scenario for your projections? I don't personally think so. I would actually be very comfortable using the 80% percentile, but this is obviously a subjective choice. At that level, the S&P 500 returned at least a 3.7% real return over all 20 year periods.

But there's more at work here as well. It's well documented that certain asset classes outperform the S&P 500, such as value and small caps. Unfortunately, I haven't find a calculator like the above one that examines these asset classes, but I do know that during the so-called "lost decade" of stocks, 1999-2009, small cap returned roughly 79%. I believe that there has only been one 20 year period in which small caps did not outperform the S&P 500. Consequently, my portfolio is tilted heavily toward value and small cap. It's volatile, but that's the price you pay for historically higher performance.

So for someone like me with a portfolio tilted toward historically higher performing asset classes who could easily wait 30 years before retiring, I feel comfortable using a projected real return of 5%. I also acknowledge that, using history as a guide, I probably have a 50% chance of achieving close to an 8% real return given the tilt of my portfolio.

I also want to include a quote from the authors of the above calculator.

"If you’ve seen our previous calculator on this subject, you’ve probably already drawn some conclusions about why this data is so interesting when presented this way. Let’s stress one main point for you again: the numbers look better and better the longer the time period you select.

That is, as your holding period increases:
-Volatility smooths out
-Returns become more predictable
-For long enough time periods negative returns actually disappear

Of course, at this point we need to caution you: past performance is no guarantee of future results. However, past performance is an excellent way to determine if the plans you’ve made based on future investment returns are reasonable. If you expect to return 12% real over a 40 year career it helps to know… that’s never happened before. A 10.282% annualized real return is the winner, so far."
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

flyingaway
Posts: 1885
Joined: Fri Jan 17, 2014 10:19 am

Re: What is the prediction of future long-term returns for your portfolio?

Post by flyingaway » Thu Feb 09, 2017 4:54 pm

Sheepdog wrote:Let's see. My average nominal return for the last 18 years (1999-2016) was 5.31% with a very conservative in-retirement stock/bond ratio and with up and down markets. I'll predict that for my portfolio going forward which will take me to age 101 :shock: .

Amended prediction. For the last 12 years (2005-2016) (which eliminates the 2000-2002 down period) with my 22% stock level in that time period (I had more stock before that) my return was a higher nominal 5.91% so I'll take that as my prediction for me for the next 18 years to 2034!!!
This is a piece of real-life recent valuable information. Thank you, Sheepdog.

User avatar
TinkerPDX
Posts: 276
Joined: Sun Jun 08, 2014 4:38 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by TinkerPDX » Fri Feb 10, 2017 12:16 am

Nate79 wrote:
TinkerPDX wrote:For forecasting purposes for equities, I use expected earnings. So for instance, for VTI, the current price/prospective earnings is 18.86, and 1/18.86 = 5.3%.

For forecasting bond returns, I use the yield. For BND, that's about 2.41% right now.

Imperfect, but I need something for planning. Factor in 2% inflation (fed target), and that leaves about 3.3% real return for US equity and 0.4% for US bond. Seems reasonable enough.

Every once in a while I update those values for each slice of my equity portfolio and each element of my bond portfolio and it spits out a total expected return for the equity portion of my portfolio based on the allocations.

From the last time I updated it (maybe a month ago), my US equity was at 5.33% nominal, my ex-US equity was 7.4% nominal, and my fixed income was 1.69% nominal, resulting in a total "expected" nominal return for 2017 of 5.27% with a 77-23 AA. I don't really "expect" it, but I need a number, so that's the one I'm using.
For your method of the return for equity how has it compared to historical returns?
Quite a bit lower.

User avatar
tadamsmar
Posts: 7767
Joined: Mon May 07, 2007 12:33 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by tadamsmar » Fri Feb 10, 2017 3:09 am

flyingaway wrote:Based on a retirement calculator, if we retire at 55, we will run out of money at 93 with a 2% real return. For me, that is not too bad.
But the cure with a 3% real return looks really nice, and might be used to convince my wife that we can retire.
2% real might be OK for planning purposes. 1% would be better.

I think it's good to use multiple methods. Some suggestions that I used:

1. Look at buying an inflaton-adjusted annuity just to see how that math pans out.

2. Do a stochastic projection using financialengines.com if you are 10+ years from retirement. The projection is not very useful when you are close to retirement since financialengines just assumes annuitization of your nest egg at retirement. At least, it worked that way the last time I used it, years ago.

User avatar
willthrill81
Posts: 5741
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: What is the prediction of future long-term returns for your portfolio?

Post by willthrill81 » Fri Feb 10, 2017 11:08 am

tadamsmar wrote:
flyingaway wrote:Based on a retirement calculator, if we retire at 55, we will run out of money at 93 with a 2% real return. For me, that is not too bad.
But the cure with a 3% real return looks really nice, and might be used to convince my wife that we can retire.
2% real might be OK for planning purposes. 1% would be better.
A 2% real return is very conservative, given historical precedent. 90% of 20 year periods in the S&P 500 yielded 2.6% or better, though it's obviously lower for 10 year periods. 70% of 20 periods periods yielded 5.3% or higher. If you want devise your planning around the worst ~5%, that's obviously your choice, but it is close to assuming the worst.

There is a risk in projecting returns that are too low. This might compel someone to save far more than they needed to, possibly to the detriment of their enjoyment of their lives during the accumulation phase. Granted, that could be argued to be a less important risk than having inadequate funds in retirement, but for someone trying to retire at 55, potentially working to age 60 due to lower than expected returns is hardly catastrophic either.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
tadamsmar
Posts: 7767
Joined: Mon May 07, 2007 12:33 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by tadamsmar » Fri Feb 10, 2017 11:36 am

willthrill81 wrote:
tadamsmar wrote:
flyingaway wrote:Based on a retirement calculator, if we retire at 55, we will run out of money at 93 with a 2% real return. For me, that is not too bad.
But the cure with a 3% real return looks really nice, and might be used to convince my wife that we can retire.
2% real might be OK for planning purposes. 1% would be better.
A 2% real return is very conservative, given historical precedent. 90% of 20 year periods in the S&P 500 yielded 2.6% or better, though it's obviously lower for 10 year periods. 70% of 20 periods periods yielded 5.3% or higher. If you want devise your planning around the worst ~5%, that's obviously your choice, but it is close to assuming the worst.

There is a risk in projecting returns that are too low. This might compel someone to save far more than they needed to, possibly to the detriment of their enjoyment of their lives during the accumulation phase. Granted, that could be argued to be a less important risk than having inadequate funds in retirement, but for someone trying to retire at 55, potentially working to age 60 due to lower than expected returns is hardly catastrophic either.
The typical AA is no going to be 100% in the SP500.

If you look at the Trinity Study:

https://www.bogleheads.org/wiki/File:TrinityTable3.jpg

If you are in 100% stocks, and you withdraw at a 4% rate, then you get 100% reliability for only 25 years. That amounts to a 0% real return being the safe assumption. By that measure, my 1% assumption is not conservative enough. None of the stock and bond AAs in the Trinity give you 100% reliability at any rate above 0% real. To exceed 0% real, you have to take on at least a 2% risk of failure to cover a 30 year retirement period.

Per the Study definitions, that's 4% of your nest egg at retirement, inflation adjusted thereafter. So, the next egg would be exhausted in precisely 25 years if the rate was a steady 0% real.

User avatar
willthrill81
Posts: 5741
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: What is the prediction of future long-term returns for your portfolio?

Post by willthrill81 » Fri Feb 10, 2017 11:48 am

tadamsmar wrote:
willthrill81 wrote:
tadamsmar wrote:
flyingaway wrote:Based on a retirement calculator, if we retire at 55, we will run out of money at 93 with a 2% real return. For me, that is not too bad.
But the cure with a 3% real return looks really nice, and might be used to convince my wife that we can retire.
2% real might be OK for planning purposes. 1% would be better.
A 2% real return is very conservative, given historical precedent. 90% of 20 year periods in the S&P 500 yielded 2.6% or better, though it's obviously lower for 10 year periods. 70% of 20 periods periods yielded 5.3% or higher. If you want devise your planning around the worst ~5%, that's obviously your choice, but it is close to assuming the worst.

There is a risk in projecting returns that are too low. This might compel someone to save far more than they needed to, possibly to the detriment of their enjoyment of their lives during the accumulation phase. Granted, that could be argued to be a less important risk than having inadequate funds in retirement, but for someone trying to retire at 55, potentially working to age 60 due to lower than expected returns is hardly catastrophic either.
The typical AA is no going to be 100% in the SP500.

If you look at the Trinity Study:

https://www.bogleheads.org/wiki/File:TrinityTable3.jpg

If you are in 100% stocks, and you withdraw at a 4% rate, then you get 100% reliability for only 25 years. That amounts to a 0% real return being the safe assumption. By that measure, my 1% assumption is not conservative enough. None of the stock and bond AAs in the Trinity give you 100% reliability at any rate above 0% real. To exceed 0% real, you have to take on at least a 2% risk of failure to cover a 30 year retirement period.
What you mean by 'real return' is different from what he is referring to. If he is including a 4%+inflation WR (Trinity study) in his calculations, a 2% real return above that would virtually never run out of money.

And a 2% risk of failure should not be a problem for someone with a pension plus SS.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
Aptenodytes
Posts: 3751
Joined: Tue Feb 08, 2011 8:39 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by Aptenodytes » Fri Feb 10, 2017 2:38 pm

If someone asks you "what number do I need?" your first question to them should be "what do you want to do with it?" In your case you neglected to say what you will do with your prediction, so there's no way to answer.

What decision are you planning to make that depends on having a prediction of portfolio returns?

User avatar
tadamsmar
Posts: 7767
Joined: Mon May 07, 2007 12:33 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by tadamsmar » Fri Feb 10, 2017 2:45 pm

willthrill81 wrote:
What you mean by 'real return' is different from what he is referring to. If he is including a 4%+inflation WR (Trinity study) in his calculations, a 2% real return above that would virtually never run out of money.

And a 2% risk of failure should not be a problem for someone with a pension plus SS.
It's a backtest for the period 1926 to 1995. It used the SP500 for the stock allocation.

The study shows that a 100% stock allocation with the 4% WR lasted 25 years but not 30 years with 100% reliability. Therefore, if a constant 2% real return would virtually never run out if money, then it follows the the Trinity results must be equivalent to a lower constant real return for the period 1926 to 1995.

The math is simple. Suppose there was no inflation. If you start with $100 and consume $4 per year, it last 25 years. If you inflate the starting sum and the withdrawal rate using the same inflation rate, then you get the same result, it lasts 25 years. That's approximately the result they got, it lasted 25 but not 30 years.
Last edited by tadamsmar on Fri Feb 10, 2017 4:02 pm, edited 1 time in total.

User avatar
jazman12
Posts: 212
Joined: Tue Jan 26, 2016 4:38 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by jazman12 » Fri Feb 10, 2017 2:55 pm

I expect my equity allocation to track the overall market by a fractional percentage point (given expenses of low cost index funds) I expect total bonds to hold somewhat steady for 2017
Act soon... time is running out

bigred77
Posts: 1998
Joined: Sat Jun 11, 2011 4:53 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by bigred77 » Fri Feb 10, 2017 3:14 pm

For planning purposes, I use 5% real for a 75/25 portfolio (with a little tilting towards small, value, and EM) over the next 3 decades.

If I get that, I can retire early and wealthy.
If I get more, I can retire earlier and wealthier.
If I get less I can retire either early or wealthy, probably not both.
If I get 0% real, I will still be OK and will retire on time.

I don't pay it too much attention other than to day dream. Bogleheads frustrate me sometimes because IMO they are a little too pessimistic. Yes, past performance does not equal future results, but it can provide some information on an appropriate range to consider. Using planning parameters that are worse than anything we've ever observed in the past strikes me as unnecessary. Should we consider what we would do if that occurs (and yes it could occur)? Yes. Should we plan on it? Probably not.

Using CAPE 10, PE ratio's, forward PE ratios, etc. has proven to explain, at best, about 40% of actual observable returns. Using these tools to estimate expected returns going forward gives us some valuable information, but not anything I would use to justify drastic changes to previously well considered plan.

BTW.. The previous 17 years have been pretty grim for US equities (S&P 500 returned about 2.5% real). Can't we all advocate for some positive reversion to the mean? Isn't it about time for a 17 year period of 9%+ real returns? Heck I may be able to retire before age 50 if that happens.

User avatar
willthrill81
Posts: 5741
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: What is the prediction of future long-term returns for your portfolio?

Post by willthrill81 » Fri Feb 10, 2017 3:34 pm

bigred77 wrote:For planning purposes, I use 5% real for a 75/25 portfolio (with a little tilting towards small, value, and EM) over the next 3 decades.

If I get that, I can retire early and wealthy.
If I get more, I can retire earlier and wealthier.
If I get less I can retire either early or wealthy, probably not both.
If I get 0% real, I will still be OK and will retire on time.

I don't pay it too much attention other than to day dream. Bogleheads frustrate me sometimes because IMO they are a little too pessimistic. Yes, past performance does not equal future results, but it can provide some information on an appropriate range to consider. Using planning parameters that are worse than anything we've ever observed in the past strikes me as unnecessary. Should we consider what we would do if that occurs (and yes it could occur)? Yes. Should we plan on it? Probably not.

Using CAPE 10, PE ratio's, forward PE ratios, etc. has proven to explain, at best, about 40% of actual observable returns. Using these tools to estimate expected returns going forward gives us some valuable information, but not anything I would use to justify drastic changes to previously well considered plan.

BTW.. The previous 17 years have been pretty grim for US equities (S&P 500 returned about 2.5% real). Can't we all advocate for some positive reversion to the mean? Isn't it about time for a 17 year period of 9%+ real returns? Heck I may be able to retire before age 50 if that happens.
+1

Spot on. :sharebeer
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
zaboomafoozarg
Posts: 1954
Joined: Sun Jun 12, 2011 12:34 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by zaboomafoozarg » Fri Feb 10, 2017 4:16 pm

bigred77 wrote:BTW.. The previous 17 years have been pretty grim for US equities (S&P 500 returned about 2.5% real). Can't we all advocate for some positive reversion to the mean? Isn't it about time for a 17 year period of 9%+ real returns? Heck I may be able to retire before age 50 if that happens.
I'd think a positive RTM would be likely if we were currently at early 1980's CAPE levels, but instead we're almost at Black Tuesday CAPE levels (less than 5% below).

Personally, I'm expecting nothing more than 1% real for US stocks for the next 20 years.

User avatar
Toons
Posts: 12958
Joined: Fri Nov 21, 2008 10:20 am
Location: Hills of Tennessee

Re: What is the prediction of future long-term returns for your portfolio?

Post by Toons » Fri Feb 10, 2017 4:16 pm

5%
:happy
"One does not accumulate but eliminate. It is not daily increase but daily decrease. The height of cultivation always runs to simplicity" –Bruce Lee

User avatar
AllieTB1323
Posts: 194
Joined: Tue Mar 24, 2015 10:17 am
Location: Desert Washington State

Re: What is the prediction of future long-term returns for your portfolio?

Post by AllieTB1323 » Fri Feb 10, 2017 5:03 pm

Expect 5.0%

Plan using 3.0%.

I'm a "Glass Half Empty" type of a person.

kolea
Posts: 1234
Joined: Fri Jul 11, 2014 5:30 pm
Location: Maui and Columbia River Gorge

Re: What is the prediction of future long-term returns for your portfolio?

Post by kolea » Fri Feb 10, 2017 7:00 pm

I don't really have a plan or an expectation. I am not going to change my AA or my investments so the die is pretty much cast. I keep an eye on spending though; that I can control. BTW, I am retired.
Kolea (pron. ko-lay-uh). Golden plover.

flyingaway
Posts: 1885
Joined: Fri Jan 17, 2014 10:19 am

Re: What is the prediction of future long-term returns for your portfolio?

Post by flyingaway » Fri Feb 10, 2017 7:38 pm

kolea wrote:I don't really have a plan or an expectation. I am not going to change my AA or my investments so the die is pretty much cast. I keep an eye on spending though; that I can control. BTW, I am retired.
Since I am not fully retired, I do need a number to help me determine when I will have enough for retirement and how the future might look like. Once I am retired, I cannot do anything but just take what the market will give me.

User avatar
Johnnie
Posts: 509
Joined: Sat May 28, 2016 3:18 pm
Location: Michigan

Re: What is the prediction of future long-term returns for your portfolio?

Post by Johnnie » Fri Feb 10, 2017 9:32 pm

bigred77 wrote:For planning purposes, I use 5% real for a 75/25 portfolio (with a little tilting towards small, value, and EM) over the next 3 decades.

If I get that, I can retire early and wealthy.
If I get more, I can retire earlier and wealthier.
If I get less I can retire either early or wealthy, probably not both.
If I get 0% real, I will still be OK and will retire on time.

I don't pay it too much attention other than to day dream. Bogleheads frustrate me sometimes because IMO they are a little too pessimistic. Yes, past performance does not equal future results, but it can provide some information on an appropriate range to consider. Using planning parameters that are worse than anything we've ever observed in the past strikes me as unnecessary. Should we consider what we would do if that occurs (and yes it could occur)? Yes. Should we plan on it? Probably not.

Using CAPE 10, PE ratio's, forward PE ratios, etc. has proven to explain, at best, about 40% of actual observable returns. Using these tools to estimate expected returns going forward gives us some valuable information, but not anything I would use to justify drastic changes to previously well considered plan.

BTW.. The previous 17 years have been pretty grim for US equities (S&P 500 returned about 2.5% real). Can't we all advocate for some positive reversion to the mean? Isn't it about time for a 17 year period of 9%+ real returns? Heck I may be able to retire before age 50 if that happens.
+1 :happy <rowr>
"I know nothing."

marcopolo
Posts: 1120
Joined: Sat Dec 03, 2016 10:22 am

Re: What is the prediction of future long-term returns for your portfolio?

Post by marcopolo » Fri Feb 10, 2017 11:41 pm

petulant wrote:
Pajamas wrote:
Johnnie wrote:
Pajamas wrote:Very conservative assumptions about the future make me feel more secure: "Plan for the worst, hope for the best."

Johnnie's right that with current numbers SS should pay out about 75% of entitled benefits starting in a couple decades. You're right that social security might get worse than that, and it would be completely unreasonable to plan for 100% if retirement includes any period after 2035.

That said, I still think it's excessively cautious to plan for 0% of social security benefits. Maybe dropping the 75% to 50% would be more realistic. Personally, I've done a range from 50% to 75% in my Monte Carlo simulations.

In aggregate it is correct that current taxes will cover 75% of entitled benefits. But, I am not at all convinced that any cuts in the future will be spread evenly across the board. I think there is a not-insignificant possibility that the politicians will decide that people that have "sufficient" other source of income/assets don't "need" Social Security benefits. This could effectively wipe out (or more sseverly reduce) any SS payment for many Bogleheads that prudently planned for their retirements. I know this really changes the entire nature of Social Security and makes it more of a welfare program than an earned entitlement. But, I don't think it is that far-fetched. In my planning i treat SS as a safety margin in my planning. Don't count on it, but think of it as a possible positive surprise that could help offset a negative surprise relative to by planning assumptions.
Once in a while you get shown the light, in the strangest of places if you look at it right.

User avatar
FIREchief
Posts: 2684
Joined: Fri Aug 19, 2016 6:40 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by FIREchief » Sat Feb 11, 2017 2:41 am

bigred77 wrote: BTW.. The previous 17 years have been pretty grim for US equities (S&P 500 returned about 2.5% real). Can't we all advocate for some positive reversion to the mean? Isn't it about time for a 17 year period of 9%+ real returns? Heck I may be able to retire before age 50 if that happens.
I see that your number is correct. That said, for the previous 20 years, the real return for the S&% 500 has been 5.3% For the previous 15 years, it has been 4.6%. For the previous ten, 5.0%. So, why reference 17 years? That's an odd number to choose. I suppose if somebody has exactly a 17 year planning horizon, they may chose to look at the past 17 years. :confused

Why don't we use numbers that will parallel typical retirement planning timelines. Let's choose 25, 30 and 35 years.

6.7%, 7.0%, 8.6%.

Sure the academics will scream "foul," and toss large studies in our faces. But I think they were doing the same thing 25, 30 and 35 years ago as well.
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.

petulant
Posts: 325
Joined: Thu Sep 22, 2016 1:09 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by petulant » Sat Feb 11, 2017 12:52 pm

marcopolo wrote:
petulant wrote:
Pajamas wrote:
Johnnie wrote:
Pajamas wrote:Very conservative assumptions about the future make me feel more secure: "Plan for the worst, hope for the best."

Johnnie's right that with current numbers SS should pay out about 75% of entitled benefits starting in a couple decades. You're right that social security might get worse than that, and it would be completely unreasonable to plan for 100% if retirement includes any period after 2035.

That said, I still think it's excessively cautious to plan for 0% of social security benefits. Maybe dropping the 75% to 50% would be more realistic. Personally, I've done a range from 50% to 75% in my Monte Carlo simulations.

In aggregate it is correct that current taxes will cover 75% of entitled benefits. But, I am not at all convinced that any cuts in the future will be spread evenly across the board. I think there is a not-insignificant possibility that the politicians will decide that people that have "sufficient" other source of income/assets don't "need" Social Security benefits. This could effectively wipe out (or more sseverly reduce) any SS payment for many Bogleheads that prudently planned for their retirements. I know this really changes the entire nature of Social Security and makes it more of a welfare program than an earned entitlement. But, I don't think it is that far-fetched. In my planning i treat SS as a safety margin in my planning. Don't count on it, but think of it as a possible positive surprise that could help offset a negative surprise relative to by planning assumptions.
I agree that a result like that is possible, and I see where you're coming from. But I don't think that particular solution is probable, or at least I don't think it's sufficiently probable to base a plan on it. If my number one financial goal was financial security in retirement, I might be more likely to adopt your approach. But my goal is to absolutely minimize retirement savings based on prudent assumptions about the future. That frees up my finances for things that are more important for my values. From that standpoint, I think it's irresponsible to assume 100% of expected social security would be available, but I think it's also excessively conservative to set a 0% social security number.

User avatar
Pajamas
Posts: 6015
Joined: Sun Jun 03, 2012 6:32 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by Pajamas » Sat Feb 11, 2017 3:29 pm

petulant wrote: If my number one financial goal was financial security in retirement, I might be more likely to adopt your approach. But my goal is to absolutely minimize retirement savings based on prudent assumptions about the future. That frees up my finances for things that are more important for my values. From that standpoint, I think it's irresponsible to assume 100% of expected social security would be available, but I think it's also excessively conservative to set a 0% social security number.
Different people do have different financial situations and different levels of comfort in their planning. Financial security in retirement is very important to me as I stopped working early and it only gets harder to start working again as time goes by. The only things that I have to worry about are extreme situations such as the markets plunging by 50% and then never recovering or becoming completely unable to care for myself at a relatively young age.

The Social Security benefits that I will receive will be relatively small even in a best-case scenario and would certainly not be hard to spend or give away if they are not needed, plus they are far off, so it makes sense for me to just ignore them in my planning. If I were older, had more resources, and were going to receive larger benefits in just a few years, I would include them. I may even do so as I approach 70, if the age for maximum benefits is still 70 when I get there. It may be pushed out even further.

User avatar
tadamsmar
Posts: 7767
Joined: Mon May 07, 2007 12:33 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by tadamsmar » Sat Feb 11, 2017 3:52 pm

bigred77 wrote: BTW.. The previous 17 years have been pretty grim for US equities (S&P 500 returned about 2.5% real). Can't we all advocate for some positive reversion to the mean? Isn't it about time for a 17 year period of 9%+ real returns? Heck I may be able to retire before age 50 if that happens.
The 2.5% is the compound annual growth rate (CAGR) for the past 17 years.

Whereas 9% percent must be your conception of the average annual growth rate boosted by your belief in reversion to the mean.

But CAGR is always lower than average annual growth. It's called volatility drag.

Unfortunately, CAGR is the one you are going to be living off of when you retire. :(

This shows the CAGR was 2.27 and the average return was a 3.89 since 2000.

http://www.moneychimp.com/features/market_cagr.htm

You are comparing apples with anti-log log apples.

User avatar
warowits
Posts: 355
Joined: Tue Sep 29, 2015 2:38 am

Re: What is the prediction of future long-term returns for your portfolio?

Post by warowits » Sat Feb 11, 2017 4:34 pm

While it is great that many of you will be fine if we have 2% real returns for equities, I always worry about counseling people to expect that. A pessimistic prediction like this might convince someone there is no point in saving at all as they will not have enough to retire prior to 70 anyway. It may not be reasonable, but I have heard far too many people say they just aren't going to retire.
There are an army of people whose pay checks depend on convincing people to invest in ways that are against their self interest. This forum is the volunteer army that fights back!

User avatar
willthrill81
Posts: 5741
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: What is the prediction of future long-term returns for your portfolio?

Post by willthrill81 » Sat Feb 11, 2017 4:41 pm

warowits wrote:While it is great that many of you will be fine if we have 2% real returns for equities, I always worry about counseling people to expect that. A pessimistic prediction like this might convince someone there is no point in saving at all as they will not have enough to retire prior to 70 anyway. It may not be reasonable, but I have heard far too many people say they just aren't going to retire.
That's why I like to view data like I recently posted in a different thread (link below). It shows the percentile ranks for various time frames of the S&P 500's history. 2% real returns is very pessimistic for a long-term investor. Even for only a 5 or 10 year period, the likelihood of reasonable returns far outweighs the likelihood of very low ones.

For instance, over all 10 year periods, the S&P 500 has produced real returns of at least 2.45% 80% of the time.

viewtopic.php?f=10&t=210755
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
willthrill81
Posts: 5741
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: What is the prediction of future long-term returns for your portfolio?

Post by willthrill81 » Sat Feb 11, 2017 4:44 pm

tadamsmar wrote:
bigred77 wrote: BTW.. The previous 17 years have been pretty grim for US equities (S&P 500 returned about 2.5% real). Can't we all advocate for some positive reversion to the mean? Isn't it about time for a 17 year period of 9%+ real returns? Heck I may be able to retire before age 50 if that happens.
The 2.5% is the compound annual growth rate (CAGR) for the past 17 years.

Whereas 9% percent must be your conception of the average annual growth rate boosted by your belief in reversion to the mean.

But CAGR is always lower than average annual growth. It's called volatility drag.

Unfortunately, CAGR is the one you are going to be living off of when you retire. :(

This shows the CAGR was 2.27 and the average return was a 3.89 since 2000.

http://www.moneychimp.com/features/market_cagr.htm

You are comparing apples with anti-log log apples.
That's why I tilt toward small-cap. Small cap had a CAGR of 8.12% over that same period; mid-cap was 9.00%.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

flyingaway
Posts: 1885
Joined: Fri Jan 17, 2014 10:19 am

Re: What is the prediction of future long-term returns for your portfolio?

Post by flyingaway » Sat Feb 11, 2017 4:57 pm

I am actually financially independent in the sense of the 4% rule. So the number that I will be using to plug into the retirement calculators is for the future retirement portfolio with about 75%/25% stock/bond. I think I am OK with the backtesting data results. But I need a number to give me a simple way for visualization. As I said before, 2% real return will make me barely to my 90s and I am OK with that. 3% real return will make me happy. The numbers used by the pension funds imply 4% real returns, I was wondering if I need to find new ways to spend more money.

Dimitri
Posts: 427
Joined: Sun May 31, 2015 10:51 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by Dimitri » Sat Feb 11, 2017 5:15 pm

Good question. According to Vanguard over the past ten years I've achieved 3.1% on my retirement accounts and 8.3% on my non-retirement accounts. My wife's figures were 4.8% and 4.0% respectively. Long term prediction? Not much. I'm less than optimistic.
Let's never come here again because it would never be as much fun.

User avatar
willthrill81
Posts: 5741
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: What is the prediction of future long-term returns for your portfolio?

Post by willthrill81 » Sat Feb 11, 2017 5:26 pm

flyingaway wrote:I am actually financially independent in the sense of the 4% rule. So the number that I will be using to plug into the retirement calculators is for the future retirement portfolio with about 75%/25% stock/bond. I think I am OK with the backtesting data results. But I need a number to give me a simple way for visualization. As I said before, 2% real return will make me barely to my 90s and I am OK with that. 3% real return will make me happy. The numbers used by the pension funds imply 4% real returns, I was wondering if I need to find new ways to spend more money.
It depends on how far into retirement you are. Analysis has consistently shown that the first decade or so of a traditional 30 year retirement largely dictates the latter's success.

So will you actually be withdrawing 4%? 4% plus inflation (constant sum + CPI), or 4% of your assets (constant percentage)? Or a hybrid?
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

User avatar
galeno
Posts: 1340
Joined: Fri Dec 21, 2007 12:06 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by galeno » Sun Feb 12, 2017 3:10 pm

3% expected real CAGR for equities. 0% for FI. At 40% equities our port's expected real CAGR = 1.2%. 60% equities should produce 1.8%.
AA = 40/55/5. Expected CAGR = 3.8%. GSD (5y) = 6.2%. USD inflation (10 y) = 1.8%. AWR = 4.0%. TER = 0.4%. Port Yield = 2.13%. Term = 34 yr. FI Duration = 6.2 yr. Portfolio survival probability = 95%.

Clive
Posts: 1951
Joined: Sat Jun 13, 2009 5:49 am

Re: What is the prediction of future long-term returns for your portfolio?

Post by Clive » Sun Feb 12, 2017 4:33 pm

Main chart is log scaled (linear scaled and the tall stock bars are clearly 'way up there', as is more evident in the bottom right chart
Image
(Avg. CAGR is the 20 year annualised of the average of the median and average values, and is after the 3% SWR amount (so add 3% onto those figures for the overall real average)).

Since 1925 for a UK investor (GB£, UK inflation), 3% SWR over 20 years had 100% success rate of ending 20 years with at least the same inflation adjusted amount as at the start date for a 3-way blend of US stocks, UK stocks and gold. 50/50 UK/US stocks did see a worst case 46% of the inflation adjusted start date amount remaining after 20 years, but had much taller better cases, and broadly on average was more rewarding.

Question is, do you go with the better probability of leaving 100%+ of the original start date amount, that in effect covers longevity (assuming of course historical averages continue into the future) - more often a reasonable amount of expansion of wealth/value remaining for longevity/heirs; Or look to potentially shoot the lights out, accept maybe less than half of original wealth remaining, potential for seriously more, average case of a decent amount more?

Personally, of those two, I'm more attracted to the 3-way. I see 50/50 stock and gold barbell as a form of volatile bond with lower taxation risk, but higher volatility than bonds over interim years. When looked at that way US stock, UK stock, gold is a form of 33 stock/67 bond ... somewhat like a Larry Portfolio. US$ (holding US stocks with those $$$'s) along with gold is a form of long/short US$. Foreign currency can be as good as gold if the domestic currency falters/fails.

Foregoing right tail potential (larger inheritance for heirs) for better safety is a acceptable 'cost' if you already have enough, especially as right tails tend to occur across periods of transition from low to high valuations (high to low yields), such as starting near/at a historic low (after a major crash), measured to some other later date relative high. The prospects that we're currently at such a low look pretty unlikely given near 0% interest rates (low yields = high prices). Even the 2008/9 lows weren't deep enough down to likely lead to a extreme right tail forward time outcome.

Will historical outcomes continue into the future? Reasonable chance IMO. Historical values contain a lot of high levels of inflation (and hence yields). To achieve a 5% real when yields are 6% you need 11% gains. Whilst in nominal terms forward times might see 6%, relative to perhaps 1% inflation ... and that's no different to in the past when inflation and yields were much higher.

If you look at the individual US stock, UK stock, gold assets, then since 1925 the best asset each year gained 20% on average. If that was one of the stock 33% allocations the remainder was a stock/gold 50/50 barbell (bond like). That's reflected in the average of the other two assets since 1925 having been 3%. When gold is the best asset, the remainder is invested in stocks, 50/50 domestic/foreign. Whatever drove the price of gold up to be the years best may very well have foreign stock values also up (weak domestic currency).

Stocks and gold are like undated bonds. Long maturities, price sensitive. Not unreasonable to assume that both stocks and gold might at least broadly pace inflation over the mid to longer term, so if the long maturity/price sensitivity has one of those assets down perhaps 20% real one year, then it may very well gain 25% real later to revert back (compound to 0%) ... perhaps making it that years best performing asset.

Some have been saying to expect lower forward time outcomes, yet looking back since 1925 there have been some pretty horrible conditions, wars, crises ...etc. UK/US/Gold real rewards since the new millennium are holding up OK :

since 2000 8.5% annualised real (11.4% nominal)
since 2005 10.8% annualised real (13.8% nominal)
since 2010 10% annualised real (12.8% nominal)

... compared to 9% annualised real since 1986 (12.5% nominal).

The high yields of 1980's, low yields of present, do not seem to have slowed so far. At some point a yield direction reversal will occur and no doubt have a slowing/downward pressure on prices - which is the primary factor for some estimating lower forward time rewards. However the current/recent gains made will compensate for that, same as it has done in the past. Extended out to 20 year periods and it tends to all get washed out.

Its reasonable however to plan for the worst, hope for the best. 2% SWR is a fine target (50x) providing you don't have to work too much longer/harder to get there. Missing out on years when you and your partner were well and fit enough to enjoy the proceeds of your retirement savings/planning to instead strive to get to perhaps 50x instead of accepting 30x often turns out disastrous - such as one or the other partners physical or mental health deteriorating. I saw a cartoon image once, but can't recall exactly where or when, of a picture of a worker rushing past a graveyard - muttering 'time is money', in the background one of the gravestones had the words 'would have given all my money for some extra time'. Haven't managed to find a copy of that cartoon since, shame - as its a very apt one to have hanging up as a reminder for retirement planning.

bigred77
Posts: 1998
Joined: Sat Jun 11, 2011 4:53 pm

Re: What is the prediction of future long-term returns for your portfolio?

Post by bigred77 » Sun Feb 12, 2017 5:49 pm

FIREchief wrote:
bigred77 wrote: BTW.. The previous 17 years have been pretty grim for US equities (S&P 500 returned about 2.5% real). Can't we all advocate for some positive reversion to the mean? Isn't it about time for a 17 year period of 9%+ real returns? Heck I may be able to retire before age 50 if that happens.
I see that your number is correct. That said, for the previous 20 years, the real return for the S&% 500 has been 5.3% For the previous 15 years, it has been 4.6%. For the previous ten, 5.0%. So, why reference 17 years? That's an odd number to choose. I suppose if somebody has exactly a 17 year planning horizon, they may chose to look at the past 17 years. :confused

Why don't we use numbers that will parallel typical retirement planning timelines. Let's choose 25, 30 and 35 years.

6.7%, 7.0%, 8.6%.

Sure the academics will scream "foul," and toss large studies in our faces. But I think they were doing the same thing 25, 30 and 35 years ago as well.
I picked 17 years because it's the turn of the century till today. No other rhyme or reason (although admittedly I knew it would include the tech crash and great recession). Just pointing out the previous 17 years have been really bad, maybe the next 17 years will be better?

We can take your numbers for the past 25, 30, and 35 years as well. Those are roughly in the middle third for real returns over their stated time frames. I'd agree with you that seeing those numbers repeat going forward would not surprise me.
tadamsmar wrote:
bigred77 wrote: BTW.. The previous 17 years have been pretty grim for US equities (S&P 500 returned about 2.5% real). Can't we all advocate for some positive reversion to the mean? Isn't it about time for a 17 year period of 9%+ real returns? Heck I may be able to retire before age 50 if that happens.
The 2.5% is the compound annual growth rate (CAGR) for the past 17 years.

Whereas 9% percent must be your conception of the average annual growth rate boosted by your belief in reversion to the mean.

But CAGR is always lower than average annual growth. It's called volatility drag.

Unfortunately, CAGR is the one you are going to be living off of when you retire. :(

This shows the CAGR was 2.27 and the average return was a 3.89 since 2000.

http://www.moneychimp.com/features/market_cagr.htm

You are comparing apples with anti-log log apples.
9% wasn't my conception of average annual returns. I really did propose it was possible we could see 9%+ real CAGR returns over the next 17 years.

If we just observed 17 years with a CAGR that would fall within the lowest decile or so of observed returns over that time frame, then maybe we will observe a CAGR over the following 17 years that lies in the top decile of observed returns? Possible no?

Post Reply