Author Topic: How do you quantify the risk differential between 85/15 and 70/30?  (Read 2347 times)

GOFU

  • Stubble
  • **
  • Posts: 170
As a quantitative matter, how much difference would it really make if I were 70/30 instead of 85/15?

This question refers to the accumulation stage, not the withdrawal stage.

Conceptually 85/15 carries more risk. But how do you calculate the extent of the risk and potential reward, quantitatively, so as to make a truly informed risk assessment?

If the bottom falls out of the market you are going to get clobbered whether you are 70% stocks or 85%.

Here is a simple example of how I put numbers to the concepts:

Let's say on a portfolio of $1 million, you have an asset mix of 85/15. If the market drops 50% overnight you are looking at a diminution in value of stock holdings from $850k down to $425k. With a 70/30 mix your stock holdings drop in value from $700k to $350k. A difference of $75k between the two down side results.

A 25% market drop results in a $37,500 disparity between these two down side results.

On the up side, let's say that over 10 years the 85/15 portfolio earns 6% (growth of $790,847) while the 70/30 earns 5% (growth of 628,894). That up side gain differential of $161,953 is 216% of the down side risk of a 50% market drop, and 432% of the down side of a 25% market drop.

Am I omitting any significant considerations from my way of thinking about this?
« Last Edit: April 29, 2018, 04:59:05 PM by GOFU »

AdrianC

  • Pencil Stache
  • ****
  • Posts: 915
  • Location: Cincinnati
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #1 on: April 29, 2018, 05:18:03 PM »
I think what youíre not considering is that during accumulation volatility is not risk. Volatility of the type you describe is desirable. You want stocks to go on sale.

Radagast

  • Handlebar Stache
  • *****
  • Posts: 1259
  • Location: West of the Mountains, East of the Sea
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #2 on: April 29, 2018, 05:19:56 PM »
https://portfoliocharts.com/calculators/

Portfolio Charts has some pretty sweet ways to play around with this.

Edit: There's a new calculator. Type your question into this:
https://portfoliocharts.com/portfolio/target-accuracy/

or this:
https://portfoliocharts.com/portfolio/portfolio-growth/
« Last Edit: April 29, 2018, 05:23:53 PM by Radagast »

GOFU

  • Stubble
  • **
  • Posts: 170
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #3 on: April 29, 2018, 05:33:25 PM »
https://portfoliocharts.com/calculators/

Portfolio Charts has some pretty sweet ways to play around with this.

Edit: There's a new calculator. Type your question into this:
https://portfoliocharts.com/portfolio/target-accuracy/

or this:
https://portfoliocharts.com/portfolio/portfolio-growth/

Great resource. Thanks.

GOFU

  • Stubble
  • **
  • Posts: 170
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #4 on: April 29, 2018, 06:46:17 PM »
I think what youíre not considering is that during accumulation volatility is not risk. Volatility of the type you describe is desirable. You want stocks to go on sale.

Which makes me question why the conventional wisdom of 80/20 or even more conservative is promoted. Someone that is, say, 25 or 30 years old with a 15, 20 or 30 year earning and investment horizon has little reason to hold anything but 100% stocks.

As I view it, the bond portion is there to help you through down markets during the withdrawal phase. During accumulation the bond portion is a drag on overall returns.

Radagast

  • Handlebar Stache
  • *****
  • Posts: 1259
  • Location: West of the Mountains, East of the Sea
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #5 on: April 29, 2018, 11:25:31 PM »
I think what youíre not considering is that during accumulation volatility is not risk. Volatility of the type you describe is desirable. You want stocks to go on sale.

Which makes me question why the conventional wisdom of 80/20 or even more conservative is promoted. Someone that is, say, 25 or 30 years old with a 15, 20 or 30 year earning and investment horizon has little reason to hold anything but 100% stocks.

As I view it, the bond portion is there to help you through down markets during the withdrawal phase. During accumulation the bond portion is a drag on overall returns.
I agree. The chief benefit is mental. If you are ok with that there is little reason for bonds if you will be making regular contributions for at least ten more years. Stocks went up? Great, you have more money! Stocks went down? Great, you get way more shares for your money! Its only really in the last 5-15 years or so where a potential crash starts to influence your plans.

dustinst22

  • Pencil Stache
  • ****
  • Posts: 577
  • Age: 40
  • Location: Huntington Beach, CA
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #6 on: April 30, 2018, 12:16:31 AM »
Bonds are important due to sequence of returns risk.

Obligatory Japan 1989-2013 (25 years) chart:



How would stocks having a negative return for 25 years affect your plans?  Suddenly you're 50 and have significantly less money than expected.  Bonds are at their best when taking distributions from the portfolio, planned or not.  While you may think this won't happen for many years, after all you're young, there are many scenarios where you might end up needing your portfolio before then.  Generally, 100% of anything is not recommended.  Here's an insightful post from the Bogleheads forum on this topic:

------------------------------------------------------------

Listen, there are two HUGE reasons other than psychology (itself not to be underestimated) to have some bonds:

1) you might need the money during a crash. As in, you get fired or get ill, both for the long term. The forced sale will be devastating and may well be the difference between the poor house and some degree of comfort. Do you really want to trade that in exchange for "more money if stocks do well", a situation in which you're already in good shape?

2) there are no guarantees that stocks rebound after any amount of time. Compare a stock certificate with a Treasury bond, no guarantees in sight for the former. If someone tells you that it's a certainty because of history blah blah, ask them for a written, notarized guarantee backed by THEIR personal fortune. Only then can you rightfully sleep well.

So don't do 100%. Add 20 or 30 percent bonds, which won't hurt your returns too badly and will be there for you when times generous. Unlike others above, I won't qualify this to depend on your worldview or vitamin levels. Just don't.

------------------------------------------------------------

http://www.bogleheads.org/forum/viewtopic.php?f=1&t=142825&newpost=2123372#p2120756

This is what the Cfiresim.com calculator tells me, regarding a 1,000,000 portfolio, a 60 year time-horizon, and a 4% withdraw rate:

Using Historical Data, 90% is the peak success rate:



Using Monte Carlo (randomized) data, between 70-80% is the peak success rate:



Every time you rerun the Monte Carlo, the results are different. Some Cfiresim.com Monte Carlo results are REALLY BAD for 100% equities:



Vanguard's Monte Carlo retirement nest egg calculator also maxes out the success rate between 70-80% stocks:

With 100 stocks / 0 bonds, a 4% withdraw rate, and a 50 year time horizon, here are the chances of success I get:

80%
79%
80%

When I change to 90 stocks / 10 bonds

81%
81%
82%

80 stocks / 20 bonds

82%
82%
81%

70 stocks / 30 bonds

82%
82%
84%

60 stocks / 40 bonds

81%
82%
80%



So, to answer your question, of "100% stocks for how long?"  I don't think there's ever a good time to be 100% stocks.




You might shave a few months off your FI date, but you're risking adding years if things don't go your way with 100% stocks.

To highlight this, let's model a family saving $2,333 a month, with a FIRE goal of $800,000.  Based on Vanguard's portfolio allocation model, they can expect to receive 10.2% with 100% stocks, and 9.6% with 80/20 stocks/bonds.



After 12 years of investing, what's the difference?
  • 100% Stocks - $666,000
  • 80/20 stocks/bonds - $638,000

Things are looking good!  Now let's put some real market data in there, and let's assume their income drops and they are unable to add anything else to their portfolio.  It doesn't matter why, job loss/starting a business/getting a less stressful job since they are so close to FIRE...whatever.  Let's see what this would look like if this family were unlucky enough to be at this stage in the years 1999 -2015, where the total bond market has outperformed the stock market:



Blue line is the total stock market, orange line is the total bond market, and the green line is intermediate bonds.  I included those for Mr. Bogle, as he prefers those to the total bond market (that's another topic).

Looking at the numbers, the 100% stock portfolio hits the $800,000 number in 2012:



While the 80/20 portfolio hits it in 2007:



That's a 5 year difference.  I can already hear the complaints, "But why didn't they add anything to the portfolio?  That will never happen to me!  If that happened to me I would've added 20% bonds to my portfolio immediately..etc."  The reason I didn't show what it looks like when they keep contributing $2,333 a month...is because there's no difference.  If you're still interested, this is what it looks like:



This is what it looks like for every starting point, usually it looks like a month or two difference from reaching the $800,000 goal.  You might shave a few months off your FI date, but you're risking adding years if things don't go your way with 100% stocks.  Note, this example only covers the accumulation phase, the consequences are much worse if you're trying to stay 100% stocks after retiring.  Click the "Monte Carlo" button on the right on CFiresim and you'll see 100% stocks can look pretty bad:



This can be devastating and may well be the difference between the poor house and some degree of comfort. Do you really want to trade that in exchange for "more money if stocks do well", a situation in which you're already in good shape?
« Last Edit: April 30, 2018, 12:24:32 AM by dustinst22 »

MustacheAndaHalf

  • Handlebar Stache
  • *****
  • Posts: 1361
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #7 on: April 30, 2018, 06:34:56 AM »
For both your original post (15-30% bonds) and your revised idea (0% bonds), I'd provide the same evidence: Vanguard Target Date funds allocate 10% bonds when you're very far from retirement.  Fidelity Target Retirement funds do the same.  Those companies have billions of retirement assets under management with those allocations.

So I'd recommend 10% bonds owing to the allocation you see at both Vanguard and Fidelity for target date retirement funds that are decades from retirement.

GOFU

  • Stubble
  • **
  • Posts: 170
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #8 on: April 30, 2018, 07:44:29 AM »
For both your original post (15-30% bonds) and your revised idea (0% bonds), I'd provide the same evidence: Vanguard Target Date funds allocate 10% bonds when you're very far from retirement.  Fidelity Target Retirement funds do the same.  Those companies have billions of retirement assets under management with those allocations.

So I'd recommend 10% bonds owing to the allocation you see at both Vanguard and Fidelity for target date retirement funds that are decades from retirement.

I won't try to match wits with Vanguard and Fidelity. Too many smart people with much experience for me to say they are wrong. But I am curious how and why they decide on that.

BobTheBuilder

  • Stubble
  • **
  • Posts: 102
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #9 on: April 30, 2018, 10:59:39 AM »
One part of the equation may be the rebalancing factor. Starting at a "neutral" market (e.g. stocks and bonds may go up and down equally likely) it can be shown that you can reap some extra return if you stick to mechanically rebalancing once or twice a year. So it is possible for a 70/30 allocation to outperform a 100% stocks allocation (if bonds are not starting at 0% yield) both on return and volatility. While you can argue about allocation percentages, if you have no bonds/gold/REITs/ or whatever in your portfolio except for stocks, you will have a hard time with rebalacing :-)

So bottom line as far as I see it: Minor allocation of something other than stocks can decrease volatility without sacrificing long term returns. How big that allocation should be however is back-testing guesswork, I guess.
If I had a porfolio big enough to matter/diversify further, I would allocate something like 70% stocks and something to rebalance against (bonds/gold). Right now I don't, so it's 100% stocks for me.

Edit: Found my "source" http://www.morganstanley.com/articles/rebalancing-effect
I am sure you get different results any time you change the time frame.
« Last Edit: April 30, 2018, 11:08:53 AM by BobTheBuilder »


dustinst22

  • Pencil Stache
  • ****
  • Posts: 577
  • Age: 40
  • Location: Huntington Beach, CA
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #11 on: April 30, 2018, 02:23:23 PM »
https://forum.mrmoneymustache.com/investor-alley/why-would-i-be-in-anything-other-than-100-stocks/

https://forum.mrmoneymustache.com/investor-alley/asset-allocation-100-stocks-for-how-long/

https://forum.mrmoneymustache.com/investor-alley/100-stocks-in-the-accumulation-phase/

https://forum.mrmoneymustache.com/investor-alley/revisiting-the-asset-allocation-question-the-case-for-100-stocks/

All very good threads, I encourage you to really read these -- particularly Dodge's prolific posts.  There is a lot of good info in regards to bonds allocation on the bogglehead forums as well -- where you'll find extremely seasoned/experienced investors.  They have a saying there -- "only when the tide goes out do you discover who has been swimming naked".
« Last Edit: April 30, 2018, 02:26:03 PM by dustinst22 »

GOFU

  • Stubble
  • **
  • Posts: 170
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #12 on: April 30, 2018, 08:43:42 PM »
Looks like I have some homework to do.

Mr Mark

  • Handlebar Stache
  • *****
  • Posts: 1150
  • Location: Planet Earth
  • Achieved Financial Independence summer 2014. RE'18
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #13 on: May 01, 2018, 12:18:16 AM »
When you're in solid accummulation mode, >>5 years from FIRE, I also favour something like a 90/10 for the reasons above. A nice way to 'autobalance' is to have a % in Wellington fund that has a 60/40 - 65/35 equity/bond tilt. So if you want 90/10, have 25% of your portfolio in Wellington. Or 16% in Wellesley fund that has a 40/60 tilt. Then just save save save. Look at the market maybe once a year on the same exact day (maybe on your birthday, or in mid Dec for tax optimisation) only to rebalance and check how things are going.

As you get closer to FIRE, gradually and regularly (rebalance annually to a written target - don't try to time market) increase bond allocation to around ~80/20 (I'm now at 75/25).

My personal target FIRE allocation for the "4% core stache" (~70% of liquid assets) is a bit more complex:
10% REIT
10% International ex-US
5% US Small cap value
15% Bond index
60% VTSAX

yet I still find myself attracted to JL Collins' super simple 70/30 VTSAX/VTBSX... sigh

For the other 30%, 15% is in rental properties. The other ~15% I have in a seperate and less volatile 60/30/10 US Equity/Bonds/CD system, as I'll need to spend it over the next 5-7 years and I need the DW to SWAN.


beltim

  • Magnum Stache
  • ******
  • Posts: 2796
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #14 on: May 01, 2018, 03:32:43 AM »
One part of the equation may be the rebalancing factor. Starting at a "neutral" market (e.g. stocks and bonds may go up and down equally likely) it can be shown that you can reap some extra return if you stick to mechanically rebalancing once or twice a year. So it is possible for a 70/30 allocation to outperform a 100% stocks allocation (if bonds are not starting at 0% yield) both on return and volatility. While you can argue about allocation percentages, if you have no bonds/gold/REITs/ or whatever in your portfolio except for stocks, you will have a hard time with rebalacing :-)

So bottom line as far as I see it: Minor allocation of something other than stocks can decrease volatility without sacrificing long term returns. How big that allocation should be however is back-testing guesswork, I guess.
If I had a porfolio big enough to matter/diversify further, I would allocate something like 70% stocks and something to rebalance against (bonds/gold). Right now I don't, so it's 100% stocks for me.

Edit: Found my "source" http://www.morganstanley.com/articles/rebalancing-effect
I am sure you get different results any time you change the time frame.

That is a super cherry-picked time frame corresponding to the longest bond bull market in history (CAGR of bonds over that time frame = 7.7%, CAGR of stocks over that time frame = 8.1%).  Over almost any other time period you look at, with more typical returns of stocks and bonds, including any percentage of bonds reduces annual returns (and also reduces volatility):

https://forum.mrmoneymustache.com/investor-alley/if-you-can-handle-volatility-and-a-15-year-timeline-invest-in-100-stocks/msg964703/#msg964703

MustacheAndaHalf

  • Handlebar Stache
  • *****
  • Posts: 1361
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #15 on: May 01, 2018, 06:37:28 AM »
Does this run contrary to conventional wisdom? Yes, I suppose it does.  So what.  Bonds are very risky right now since they do not sport high enough yields to ensure an investor maintains his purchasing power net of all taxes, fees, and inflation over a long term holding period.
In general I don't think people should ignore conventional wisdom without an explanation and evidence, so your explanation is important.  Come to think of it, people probably shouldn't follow convention wisdom without evidence, either.  :)

There's a reason to prefer 5% bonds over 0% bonds: rebalancing is an important behavior, and it won't happen without practice.  Selling stocks while everyone tells you how well stocks have done is not easy.  But at 5% bonds, it's much easier than at 30% bonds (the 70/30 portfolio OP mentioned).  Retirement date funds slowly increase the bond percentage, and I think people with 0% bonds will just stay at 0% bonds rather than adjusting.

BobTheBuilder

  • Stubble
  • **
  • Posts: 102
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #16 on: May 01, 2018, 08:11:14 AM »

That is a super cherry-picked time frame corresponding to the longest bond bull market in history (CAGR of bonds over that time frame = 7.7%, CAGR of stocks over that time frame = 8.1%).  Over almost any other time period you look at, with more typical returns of stocks and bonds, including any percentage of bonds reduces annual returns (and also reduces volatility):

https://forum.mrmoneymustache.com/investor-alley/if-you-can-handle-volatility-and-a-15-year-timeline-invest-in-100-stocks/msg964703/#msg964703

Aha, indeed. Thanks for the graph! So lets keep the stocks north of 70%.

AdrianC

  • Pencil Stache
  • ****
  • Posts: 915
  • Location: Cincinnati
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #17 on: May 07, 2018, 07:43:30 AM »
I think what youíre not considering is that during accumulation volatility is not risk. Volatility of the type you describe is desirable. You want stocks to go on sale.

Which makes me question why the conventional wisdom of 80/20 or even more conservative is promoted. Someone that is, say, 25 or 30 years old with a 15, 20 or 30 year earning and investment horizon has little reason to hold anything but 100% stocks.

As I view it, the bond portion is there to help you through down markets during the withdrawal phase. During accumulation the bond portion is a drag on overall returns.

Most people can't stand to "lose" 50% of their stash when stocks crash. They do stupid things like selling at the bottom and buying at the top. Bonds cushion the ride somewhat. If that's what it takes to stop someone from behaving irrationally then bonds (short or intermediate term) are good.

Assume stocks will drop 50% at some point. 100% stocks, you're down 50%. 50% stocks 50% bonds, you're down 25%.

Someone mentioned sequence of returns risk. Yes, bonds help mitigate sequence of returns risk when you're in withdrawal mode. There's no sequence of returns risk in accumulation, as long as you can trust yourself to not do something stupid when stocks go on sale.

frugal_c

  • Stubble
  • **
  • Posts: 175
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #18 on: May 07, 2018, 08:40:53 AM »
It's really hard to quantify as it depends on the time period you are looking at.  Bonds kicked butt during the great depression and the 2008 meltdown.  Bonds did poorly during the 70's and early 80's inflation.  Which scenario will the future look like?

Personally, if I had 20 years or more before retirement I would be 85-90% stocks.  The only thing holding me back from 100% stocks is I saw during 2008 & 09 that I might have actually needed the money right then.  People were getting laid off all over the place, there was a mortgage to be paid, jobs were brutal to find, and selling stocks that were down 50% or more didn't seem like good capital allocation.  I just wouldn't assume that you won't need the money until x years.

jacoavluha

  • Stubble
  • **
  • Posts: 231
Re: How do you quantify the risk differential between 85/15 and 70/30?
« Reply #19 on: May 13, 2018, 08:13:19 PM »
How much downside can you tolerate? Assume your equity allocation can drop in half at any moment. Are you ok staying the course if your portfolio fell by 40-45%? This could absolutely happen at 85/15 allocation.

Many of us havenít really experienced a bear market and itís hard to predict how weíll react.