Vanguard Target Retirement 2065

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
inverter
Posts: 1021
Joined: Mon Jul 27, 2015 1:40 pm
Location: New York, NY

Vanguard Target Retirement 2065

Post by inverter »

If any of you all are young (like me) or have children, Vanguard launched the Target Retirement 2065. Just moved some of my money over there for simplicity.

Cheers! :sharebeer
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Vanguard Target Retirement 2065

Post by willthrill81 »

Along with Paul Merriman, I have to say that I really have no idea why Vanguard has 10% of these funds in bonds when the target date is multiple decades into the future. The perceived volatility will be nearly the same as that of a 100/0 fund, and that 10% in bonds reduces the expected return by .4-.5% every year, which is a big crutch over a nearly 40 year period. Twenty years and in makes sense to me, but nearly 40?
The Sensible Steward
DSInvestor
Posts: 11647
Joined: Sat Oct 04, 2008 11:42 am

Re: Vanguard Target Retirement 2065

Post by DSInvestor »

willthrill81 wrote:Along with Paul Merriman, I have to say that I really have no idea why Vanguard has 10% of these funds in bonds when the target date is multiple decades into the future. The perceived volatility will be nearly the same as that of a 100/0 fund, and that 10% in bonds reduces the expected return by .4-.5% every year, which is a big crutch over a nearly 40 year period. Twenty years and in makes sense to me, but nearly 40?
For those who do not want any bonds, the solution is simple - hold stock funds.

I believe Benjamin Graham suggested asset allocations of 25/75 in either direction.
Last edited by DSInvestor on Sun Jul 16, 2017 2:06 pm, edited 1 time in total.
Wiki
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Vanguard Target Retirement 2065

Post by willthrill81 »

DSInvestor wrote:
willthrill81 wrote:Along with Paul Merriman, I have to say that I really have no idea why Vanguard has 10% of these funds in bonds when the target date is multiple decades into the future. The perceived volatility will be nearly the same as that of a 100/0 fund, and that 10% in bonds reduces the expected return by .4-.5% every year, which is a big crutch over a nearly 40 year period. Twenty years and in makes sense to me, but nearly 40?
For those who do not want any bonds, the solution is simple - hold stock funds.
I agree that the solution is simple, but it still does not address why Vanguard has 10% of a 2065 fund in bonds.
The Sensible Steward
DSInvestor
Posts: 11647
Joined: Sat Oct 04, 2008 11:42 am

Re: Vanguard Target Retirement 2065

Post by DSInvestor »

Vanguard's approach to target date funds:
https://advisors.vanguard.com/VGApp/iip ... EWIR062915

It's not just Vanguard's TR funds that are like this.

Fidelity's Freedom Index 2060 has 10% bonds.
https://fundresearch.fidelity.com/mutua ... /315793695

TRP's TR60 fund has 14% bonds/cash:
http://www3.troweprice.com/fb2/fbkweb/c ... cker=TRTFX

TSP Lifecycle 2050 has 17% bonds:
https://www.tsp.gov/InvestmentFunds/Fun ... L2050.html

Schwab TR 2060 has around 5% bonds:
http://www.schwab.com/public/schwab/inv ... ol%3DSWYNX
Wiki
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Vanguard Target Retirement 2065

Post by willthrill81 »

DSInvestor wrote:Vanguard's approach to target date funds:
https://advisors.vanguard.com/VGApp/iip ... EWIR062915
From that paper,
As described earlier, the human capital theory supports a larger commitment to equities for young individuals, declining to a more modest
allocation as the investor approaches retirement and eventually leaves the workforce. Vanguard TDFs maintain a significant level of equity exposure (90%) to age 40 because one’s human capital remains so dominant over the small balances in financial capital during the early stages of asset accumulation.
Based on that logic, one could easily argue, though not definitively, that a 100% equity allocation for young investors makes sense.
DSInvestor wrote:It's not just Vanguard's TR funds that are like this.

Fidelity's Freedom Index 2060 has 10% bonds.
https://fundresearch.fidelity.com/mutua ... /315793695

TRP's TR60 fund has 14% bonds/cash:
http://www3.troweprice.com/fb2/fbkweb/c ... cker=TRTFX

TSP Lifecycle 2050 has 17% bonds:
https://www.tsp.gov/InvestmentFunds/Fun ... L2050.html

Schwab TR 2060 has around 5% bonds:
http://www.schwab.com/public/schwab/inv ... ol%3DSWYNX
Again, why?
The Sensible Steward
User avatar
ClevrChico
Posts: 3258
Joined: Tue Apr 03, 2012 8:24 pm

Re: Vanguard Target Retirement 2065

Post by ClevrChico »

This makes me feel old!
User avatar
#Cruncher
Posts: 3977
Joined: Fri May 14, 2010 2:33 am
Location: New York City
Contact:

Re: Vanguard Target Retirement 2065

Post by #Cruncher »

willthrill81 in [url=https://www.bogleheads.org/forum/viewtopic.php?p=3451758#p3451758]this post[/url] wrote:... I really have no idea why Vanguard has 10% of these funds in bonds when the target date is multiple decades into the future. ... that 10% in bonds reduces the expected return by .4-.5% every year, ...
The benefit of re-balancing will reduce this differential. For example, assuming a constant 2% growth from bonds, if stocks grew 6% every year, the "drag" from the bonds would indeed be 0.4% (90% * 6% + 10% * 2% = 5.6%). Over eight years a $10,000 investment would grow to $15,464 (10000 * 1.056 ^ 8). However, instead of stocks growing 6% every year lets assume the following:
  • Stocks repeatedly grow 20% one year and then fall 6.3667% the next year. This is equivalent to a 6% compound growth rate.

    Code: Select all

    (1 + 20%) * (1 - 6.3667%) = 1.1236 = 1.06 * 1.06
  • At the end of each year stocks and bonds are re-balanced to 90% : 10%.
Instead of $15,464, with re-balancing the investment would grow to $15,547, a growth rate of 5.671%.

Code: Select all

Stock up    20.0000%                  42.2105%
Stock down  -6.3667%                 -20.9904%
 Year         Stocks  Bonds   Total     Stocks  Bonds   Total 
    0          9,000  1,000  10,000      9,000  1,000  10,000 
    1         10,638  1,182  11,820     12,437  1,382  13,819 
    2         10,050  1,117  11,166     10,112  1,124  11,236 
    3         11,879  1,320  13,199     13,974  1,553  15,527 
    4         11,222  1,247  12,469     11,362  1,262  12,625 
    5         13,264  1,474  14,738     15,701  1,745  17,446 
    6         12,531  1,392  13,923     12,767  1,419  14,185 
    7         14,811  1,646  16,457     17,642  1,960  19,602 
    8         13,992  1,555  15,547     14,345  1,594  15,938 
 Annual growth               5.671%                    6.000%
And if stocks alternated between growing 42.2105% and falling 20.9904% (still averaging 6%), with re balancing the investment would grow to $15,938, a growth rate of 6%. This is the same as a 100% stock portfolio growing 6% per year.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Vanguard Target Retirement 2065

Post by willthrill81 »

#Cruncher wrote:
willthrill81 in [url=https://www.bogleheads.org/forum/viewtopic.php?p=3451758#p3451758]this post[/url] wrote:... I really have no idea why Vanguard has 10% of these funds in bonds when the target date is multiple decades into the future. ... that 10% in bonds reduces the expected return by .4-.5% every year, ...
The benefit of re-balancing will reduce this differential. For example, assuming a constant 2% growth from bonds, if stocks grew 6% every year, the "drag" from the bonds would indeed be 0.4% (90% * 6% + 10% * 2% = 5.6%). Over eight years a $10,000 investment would grow to $15,464 (10000 * 1.056 ^ 8). However, instead of stocks growing 6% every year lets assume the following:
  • Stocks repeatedly grow 20% one year and then fall 6.3667% the next year. This is equivalent to a 6% compound growth rate.

    Code: Select all

    (1 + 20%) * (1 - 6.3667%) = 1.1236 = 1.06 * 1.06
  • At the end of each year stocks and bonds are re-balanced to 90% : 10%.
Instead of $15,464, with re-balancing the investment would grow to $15,547, a growth rate of 5.671%.

Code: Select all

Stock up    20.0000%                  42.2105%
Stock down  -6.3667%                 -20.9904%
 Year         Stocks  Bonds   Total     Stocks  Bonds   Total 
    0          9,000  1,000  10,000      9,000  1,000  10,000 
    1         10,638  1,182  11,820     12,437  1,382  13,819 
    2         10,050  1,117  11,166     10,112  1,124  11,236 
    3         11,879  1,320  13,199     13,974  1,553  15,527 
    4         11,222  1,247  12,469     11,362  1,262  12,625 
    5         13,264  1,474  14,738     15,701  1,745  17,446 
    6         12,531  1,392  13,923     12,767  1,419  14,185 
    7         14,811  1,646  16,457     17,642  1,960  19,602 
    8         13,992  1,555  15,547     14,345  1,594  15,938 
 Annual growth               5.671%                    6.000%
And if stocks alternated between growing 42.2105% and falling 20.9904% (still averaging 6%), with re balancing the investment would grow to $15,938, a growth rate of 6%. This is the same as a 100% stock portfolio growing 6% per year.
You are suggesting that there is a rebalancing bonus or at least that a portfolio with some amount of bonds will, over the long-term, perform just as well as a 100% stock portfolio. History has shown that that's not the case. Rebalancing maintains a desired level of risk but has not historically improved returns.

According to Vanguard, every 10% increase in a portfolio's allocation to bonds has reduced the CAGR by .3% to .5%, and that's with annual rebalancing.
The Sensible Steward
indexonlyplease
Posts: 1571
Joined: Thu Apr 30, 2015 12:30 pm
Location: Florida

Re: Vanguard Target Retirement 2065

Post by indexonlyplease »

The main point of the Target Dated Fund is someone that knows nothing about investing or will never want to know has a plan that is done for them. So, with all the mistakes they can make in 40-50 yrs of investing .5% (maybe) is a small price to pay. They could do a lot more damage playing with the investments.

This is why the market has returned a lot more than investors make.

Anyway it is the best way to start until someone wants to learn (maybe).
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Vanguard Target Retirement 2065

Post by willthrill81 »

indexonlyplease wrote:The main point of the Target Dated Fund is someone that knows nothing about investing or will never want to know has a plan that is done for them. So, with all the mistakes they can make in 40-50 yrs of investing .5% (maybe) is a small price to pay. They could do a lot more damage playing with the investments.
Certainly that's true, but it's beside the point. Including 10% of the fund in bonds isn't doing the investors any favors. It isn't enough to noticeably reduce the volatility, and over time it will likely cost investors a lot of money.
indexonlyplease wrote:This is why the market has returned a lot more than investors make.
That's an ongoing myth that's simply false. DALBAR's studies on the issue are completely bogus. Here's an article describing why. Most investors keep pace with the market just fine.
indexonlyplease wrote:Anyway it is the best way to start until someone wants to learn (maybe).
For now, that's probably true.
The Sensible Steward
indexonlyplease
Posts: 1571
Joined: Thu Apr 30, 2015 12:30 pm
Location: Florida

Re: Vanguard Target Retirement 2065

Post by indexonlyplease »

willthrill81 wrote:
indexonlyplease wrote:The main point of the Target Dated Fund is someone that knows nothing about investing or will never want to know has a plan that is done for them. So, with all the mistakes they can make in 40-50 yrs of investing .5% (maybe) is a small price to pay. They could do a lot more damage playing with the investments.
Certainly that's true, but it's beside the point. Including 10% of the fund in bonds isn't doing the investors any favors. It isn't enough to noticeably reduce the volatility, and over time it will likely cost investors a lot of money.
indexonlyplease wrote:This is why the market has returned a lot more than investors make.
That's an ongoing myth that's simply false. DALBAR's studies on the issue are completely bogus. Here's an article describing why. Most investors keep pace with the market just fine.
indexonlyplease wrote:Anyway it is the best way to start until someone wants to learn (maybe).
For now, that's probably true.
Thanks for that article. I will read it.
User avatar
#Cruncher
Posts: 3977
Joined: Fri May 14, 2010 2:33 am
Location: New York City
Contact:

Re: Vanguard Target Retirement 2065

Post by #Cruncher »

willthrill81 in [url=https://www.bogleheads.org/forum/viewtopic.php?p=3451872#p3451872]this post[/url] wrote:You are suggesting [1] that there is a rebalancing bonus or at least [2] that a portfolio with some amount of bonds will, over the long-term, perform just as well as a 100% stock portfolio.
I did suggest [1] that there is a re-balancing bonus. This was a mistake. I should have said "may be" because there can certainly be scenarios where re-balancing does not improve returns. However, I did not suggest [2] that a portfolio with bonds will, over the long term, perform as well as a 100% stock portfolio.
willthrill81 in same post wrote:Rebalancing ... has not historically improved returns. According to Vanguard, every 10% increase in a portfolio's allocation to bonds has reduced the CAGR by .3% to .5%, and that's with annual rebalancing.
I'd like to see some evidence for this. It's quite possible I'm wrong in thinking that re-balancing usually improves returns. Unfortunately, willthrill81, the Vanguard report you cite doesn't address this question. It simply shows that over the period 1926-2015 portfolios performed better the higher the stock allocation. But it doesn't provide any information about the effect of re-balancing. [*]

What I'd like to see is a report like the following one that I cobbled together -- except more comprehensive. It's based on the 2002-2016 returns for Vanguard Total Stock Market Index Fund Investor Shares (VTSMX) and Vanguard Total Bond Market Index Fund Investor Shares (VBMFX).

Code: Select all

             ------ Rebalancing ------ 
VTSMX VBMFX  Without    With     Diff
 100%    0%   7.212%     n/a
  90%   10%   6.968%   7.125%   0.157% 
  80%   20%   6.716%   6.986%   0.270% 
  70%   30%   6.455%   6.798%   0.343%
  60%   40%   6.185%   6.565%   0.380% 
  50%   50%   5.905%   6.289%   0.385% 
  40%   60%   5.614%   5.972%   0.359% 
  30%   70%   5.311%   5.616%   0.305% 
  20%   80%   4.996%   5.222%   0.226% 
  10%   90%   4.667%   4.790%   0.123% 
   0%  100%   4.323%     n/a
It shows a slight benefit from annual re-balancing for every portfolio with from 10% to 90% bonds. Admittedly 2002-2016 is a short period (that's all that is shown on Vanguard's Historical Returns for VTSMX and VBMFX) and there are other re-balancing options than annually. Here is the detail for 90% stocks 10% bonds:

Code: Select all

           % Change         Without Rebalancing      With Rebalancing
        ---------------    ---------------------   ---------------------
Year     VTSMX    VBMFX     VTSMX  VBMFX   Total    VTSMX  VBMFX   Total

Code: Select all

                            9,000  1,000  10,000    9,000  1,000  10,000
2002   (20.96%)   8.26%     7,114  1,083   8,196    7,377    820   8,196
2003    31.35%    3.97%     9,344  1,126  10,469    9,487  1,054  10,541
2004    12.52%    4.24%    10,514  1,173  11,687   10,596  1,177  11,774
2005     5.98%    2.40%    11,142  1,201  12,344   11,192  1,244  12,436
2006    15.51%    4.27%    12,870  1,253  14,123   12,802  1,422  14,225
2007     5.49%    6.92%    13,577  1,339  14,916   13,523  1,503  15,026
2008   (37.04%)   5.05%     8,548  1,407   9,955    9,084  1,009  10,093
2009    28.70%    5.93%    11,001  1,491  12,492   11,484  1,276  12,760
2010    17.09%    6.42%    12,882  1,586  14,468   13,324  1,480  14,804
2011     0.96%    7.56%    13,005  1,706  14,711   13,540  1,504  15,044
2012    16.25%    4.05%    15,119  1,775  16,894   15,575  1,731  17,305
2013    33.35%   (2.26%)   20,161  1,735  21,896   20,214  2,246  22,460
2014    12.43%    5.76%    22,667  1,835  24,502   22,592  2,510  25,102
2015     0.29%    0.30%    22,732  1,841  24,573   22,658  2,518  25,175
2016    12.53%    2.50%    25,581  1,887  27,467   25,269  2,808  28,077
                           ------ ------  ------                  ------
Return                     7.212% 4.323%  6.968%                  7.125%
Without re-balancing the 90:10 portfolio grew 0.244% points less annually than the 100% stock portfolio. But annual re-balancing trimmed this shortfall to only 0.087% points.

* Even though the Vanguard report doesn't address the issue of whether or not re-balancing helps returns, its results are quite interesting in their own right:

Code: Select all

                       Change
Stocks Bonds   Growth  Per 10%
------ -----   ------  -------
    0   100      5.4%
   20    80      6.7%    0.65%
   30    70      7.2%    0.50%
   40    60      7.8%    0.60%
   50    50      8.3%    0.50%
   60    40      8.7%    0.40%
   70    30      9.1%    0.40%
   80    20      9.5%    0.40%
  100     0     10.1%    0.30%
Adding a little bit of stocks to a bond heavy allocation has a large impact on return. But adding the same amount of stocks to an already-stock-heavy portfolio has less benefit. E.g., going from stocks:bonds 0:100 to 20:80 increases return 1.3% points. But going from 80:20 to 100:0 bumps return only 0.6% points.

It's too bad Vanguard didn't include the 10:90 and 90:10 portfolios to better show this effect near the extremes of allocation. It's especially unfortunate Vanguard didn't include the stock:bond 90:10 portfolio since that's the one for Vanguard Target Retirement 2065 Fund (VLXVX) which willthrill81 is criticizing.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Vanguard Target Retirement 2065

Post by willthrill81 »

#Cruncher wrote:
willthrill81 in [url=https://www.bogleheads.org/forum/viewtopic.php?p=3451872#p3451872]this post[/url] wrote:You are suggesting [1] that there is a rebalancing bonus or at least [2] that a portfolio with some amount of bonds will, over the long-term, perform just as well as a 100% stock portfolio.
I did suggest [1] that there is a re-balancing bonus. This was a mistake. I should have said "may be" because there can certainly be scenarios where re-balancing does not improve returns. However, I did not suggest [2] that a portfolio with bonds will, over the long term, perform as well as a 100% stock portfolio.
When you said this,
#Cruncher wrote:And if stocks alternated between growing 42.2105% and falling 20.9904% (still averaging 6%), with re balancing the investment would grow to $15,938, a growth rate of 6%. This is the same as a 100% stock portfolio growing 6% per year.
I thought that's what you were directly implying.
#Cruncher wrote:
willthrill81 in same post wrote:Rebalancing ... has not historically improved returns. According to Vanguard, every 10% increase in a portfolio's allocation to bonds has reduced the CAGR by .3% to .5%, and that's with annual rebalancing.
I'd like to see some evidence for this. It's quite possible I'm wrong in thinking that re-balancing usually improves returns. Unfortunately, willthrill81, the Vanguard report you cite doesn't address this question. It simply shows that over the period 1926-2015 portfolios performed better the higher the stock allocation. But it doesn't provide any information about the effect of re-balancing.
Fair enough. Just go to Portfolio Visualizer and compare different allocations between something like TSM and ITT (longer data than TBM but similar performance) with and without rebalancing. At least from 1972 to now, the differences I've seen range from zero to under 10 bps. But with other asset classes, there has actually been a big cost to rebalancing. For instance, with a single $10,000 investment, a portfolio comprised of 50% SCV and 50% ITT rebalanced annually from 1972 to now had a CAGR of 11.26% but 12.70% if no rebalancing was used. If $10,000 was added to the portfolio annually, then the final portfolio balance would be $21.9 million with rebalancing $37.1 million with no rebalancing. Since SCV and equities in general tend to outperform bonds significantly over the long-term, this is perfectly logical since the stocks will continue to grow and outpace bonds if the portfolio is never rebalanced.

Altogether, this analysis of historical data strongly suggests to me at least that there is virtually no rebalancing bonus in practice, but that there could very well be a big rebalancing cost.
The Sensible Steward
freebeer
Posts: 2014
Joined: Wed May 02, 2007 8:30 am
Location: Seattle area USA

Re: Vanguard Target Retirement 2065

Post by freebeer »

willthrill81 wrote:...10% in bonds reduces the expected return by .4-.5% every year, which is a big crutch over a nearly 40 year period...
well, over 40 years many things can change and you can't actually predict "expected return" based only on past performance... so that term is inherently misleading as typically used.

VG may have data that indicates that 10% bonds actually increases the expected return in a range of possible future scenarios (even if that wasn't the case based only on past results). Alternatively, it may be that it's just marketing: a TR fund that is only 1 or 2 stock funds for decades wouldn't sell.
rkhusky
Posts: 17763
Joined: Thu Aug 18, 2011 8:09 pm

Re: Vanguard Target Retirement 2065

Post by rkhusky »

Not reblancing will always look good after a long bull market. Rebalancing can provide a bonus if the rebalancing happens to occur at market lows and highs, but this is not possible to predict. Hence, rebalancing bonuses/losses tend to average out over varied market conditions.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Vanguard Target Retirement 2065

Post by willthrill81 »

freebeer wrote:
willthrill81 wrote:...10% in bonds reduces the expected return by .4-.5% every year, which is a big crutch over a nearly 40 year period...
well, over 40 years many things can change and you can't actually predict "expected return" based only on past performance... so that term is inherently misleading as typically used.
If you won't use the relative long-term historical performance of asset classes to craft a portfolio, I don't know what else you would use. Perhaps your crystal ball is functioning better than mine.

At any rate, I can guarantee you that every TDF out there was crafted with long-term historical performance at the forefront of the reasoning.
freebeer wrote:VG may have data that indicates that 10% bonds actually increases the expected return in a range of possible future scenarios (even if that wasn't the case based only on past results). Alternatively, it may be that it's just marketing: a TR fund that is only 1 or 2 stock funds for decades wouldn't sell.
VG's data going back to 1926 are quite clear that adding bonds has reduced a portfolio's long-term historical performance, which is expected by virtually every expert in the industry. I see no reason or point to disputing that.
The Sensible Steward
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Vanguard Target Retirement 2065

Post by willthrill81 »

rkhusky wrote:Not reblancing will always look good after a long bull market. Rebalancing can provide a bonus if the rebalancing happens to occur at market lows and highs, but this is not possible to predict. Hence, rebalancing bonuses/losses tend to average out over varied market conditions.
Unless you consider the last 45 years a long bull market, the data simply do not support the notion of a rebalancing bonus, which is a persistent myth with virtually no real data to support it.

Rebalancing maintains a desired level of risk but has not improved long-term historical returns. As I posted above, in many situations rebalancing can have a high cost.
The Sensible Steward
freebeer
Posts: 2014
Joined: Wed May 02, 2007 8:30 am
Location: Seattle area USA

Re: Vanguard Target Retirement 2065

Post by freebeer »

willthrill81 wrote:
freebeer wrote:
willthrill81 wrote:...10% in bonds reduces the expected return by .4-.5% every year, which is a big crutch over a nearly 40 year period...
well, over 40 years many things can change and you can't actually predict "expected return" based only on past performance... so that term is inherently misleading as typically used.
If you won't use the relative long-term historical performance of asset classes to craft a portfolio, I don't know what else you would use. Perhaps your crystal ball is functioning better than mine.

At any rate, I can guarantee you that every TDF out there was crafted with long-term historical performance at the forefront of the reasoning.
freebeer wrote:VG may have data that indicates that 10% bonds actually increases the expected return in a range of possible future scenarios (even if that wasn't the case based only on past results). Alternatively, it may be that it's just marketing: a TR fund that is only 1 or 2 stock funds for decades wouldn't sell.
VG's data going back to 1926 are quite clear that adding bonds has reduced a portfolio's long-term historical performance, which is expected by virtually every expert in the industry. I see no reason or point to disputing that.
So maybe I used the term "data" loosely when I really meant "information" or "insights" (from VG, not me) - what I meant is that even though, yes, adding bonds has reduced a portfolio's long-term historical performance, that doesn't mean that this is expected to definitely continue into the future. After all, adding international stocks has reduced a portfolio's long-term historical performance over a US-only portfolio over the past decades but that's because of the great secular US bull market not something that can be counted on to persist into the future. Reducto ad absurdum, adding almost anything but AAPL to a portfolio would have reduced performance over the last 37 years but that doesn't mean an AAPL-only portfolio has highest expected returns over the next 37 years. It seems logical to me that VG planners may have rational reasons to favor a floor of 10% bonds for the next 37 years (TR 2065) even if that wouldn't have been optimal over the last 37 years.
rkhusky
Posts: 17763
Joined: Thu Aug 18, 2011 8:09 pm

Re: Vanguard Target Retirement 2065

Post by rkhusky »

willthrill81 wrote: Rebalancing maintains a desired level of risk but has not improved long-term historical returns. As I posted above, in many situations rebalancing can have a high cost.
Right. Rebalancing always looks good at the bottom of a long bear market, as does having more bonds. Not rebalancing always looks good at the top of a long bull market, as does having less bonds. Bonds are insurance, and insurance has a cost, and a benefit when things go bad.
ThrustVectoring
Posts: 771
Joined: Wed Jul 12, 2017 2:51 pm

Re: Vanguard Target Retirement 2065

Post by ThrustVectoring »

There's two overlapping effects - improved risk-adjusted returns through diversifying into a low-correlation asset (bonds), and the overall level of risks and returns through a lower equity exposure.

You can add back in risk and returns through leverage. It is a more complicated strategy than going straight 100% equities, though.
Current portfolio: 60% VTI / 40% VXUS
User avatar
#Cruncher
Posts: 3977
Joined: Fri May 14, 2010 2:33 am
Location: New York City
Contact:

Re: Vanguard Target Retirement 2065

Post by #Cruncher »

willthrill81 in [url=https://www.bogleheads.org/forum/viewtopic.php?p=3452699#p3452699]this post[/url] wrote:Just go to Portfolio Visualizer and compare different allocations between something like TSM and ITT (longer data than TBM but similar performance) with and without rebalancing. At least from 1972 to now, the differences I've seen range from zero to under 10 bps.
Thanks for mentioning Portfolio Visualizer. I hadn't seriously used it before. But now I see it's powerful and easy to use. First with the Backtest Portfolio Asset Allocation I confirmed the same results for VTSMX & VBMFX as in my previous post. Then I used Backtest Portfolio Asset Class Allocation for "US Stock Market" : "Intermediate Term Treasury". Here are the results for Jan 1972 - May 2017:

Code: Select all

                                           ------ 90:10 ------
                       100:0      0:100    No Rebal    w Rebal
                      -------    -------   --------    -------
Initial Balance        10,000     10,000     10,000     10,000
Final Balance         847,476    234,917    786,220    802,147
CAGR                   10.27%      7.20%     10.09%     10.14% 
Stdev                  15.47%      5.91%     14.40%     13.94% 
Best Year              37.82%     31.13%     34.78%     34.77% 
Worst Year            (37.04%)    (4.33%)   (34.92%)   (32.00%)
Max. Drawdown         (50.89%)   (10.70%)   (48.39%)   (45.52%)
Sharpe Ratio             0.41       0.42       0.42       0.43
Sortino Ratio            0.59       0.66       0.60       0.62
US Mkt Correlation       1.00       0.08       1.00       1.00
There was a "bonus" for re-balancing, but it was tiny: 0.05% points (10.14% - 10.09%). But also note that the 100% stock portfolio was only 0.18% points better (10.27% - 10.09%).
willthrill81, continuing, wrote:But with other asset classes, there has actually been a big cost to rebalancing. For instance, with a single $10,000 investment, a portfolio comprised of 50% SCV and 50% ITT rebalanced annually from 1972 to now had a CAGR of 11.26% but 12.70% if no rebalancing was used.
Here is what Portfolio Visualizer shows for "US Small Cap Value" : "Intermediate Term Treasury" for Jan 1972 - May 2017:

Code: Select all

                                           ------ 50:50 ------
                       100:0      0:100    No Rebal    w Rebal
                      -------    -------   --------    -------
Initial Balance        10,000     10,000     10,000     10,000
Final Balance       4,335,721    234,917  2,285,319  1,270,320
CAGR                   14.30%      7.20%     12.70%     11.26% 
Stdev                  17.88%      5.91%     14.33%      9.63% 
Best Year              54.78%     31.13%     36.82%     34.27% 
Worst Year            (32.05%)    (4.33%)   (28.50%)    (9.82%)
Max. Drawdown         (56.13%)   (10.70%)   (51.34%)   (24.27%)
Sharpe Ratio             0.58       0.42       0.58       0.68
Sortino Ratio            0.86       0.66       0.85       1.03
US Mkt Correlation       0.89       0.08       0.87       0.85
The effect of re-balancing is indeed significantly negative in this case as you say. But note how much riskier is the "50:50" allocation without re-balancing. This shouldn't be too surprising because without re-balancing what starts out as 50:50 ends up 95:5 [95% = 4336 / (4336 + 235)].

I'm surprised you chose a 50:50 allocation, willthrill81. Here are the figures for a 90:10 SCV:ITT allocation which is more germane to the subject of this thread:

Code: Select all

                                           ------ 90:10 ------
                       100:0      0:100    No Rebal    w Rebal
                      -------    -------   --------    -------
Initial Balance        10,000     10,000     10,000     10,000
Final Balance       4,335,721    234,917  3,925,641  3,519,598
CAGR                   14.30%      7.20%     14.05%     13.78% 
Stdev                  17.88%      5.91%     17.21%     16.14% 
Best Year              54.78%     31.13%     50.00%     50.68% 
Worst Year            (32.05%)    (4.33%)   (31.63%)   (27.52%)
Max. Drawdown         (56.13%)   (10.70%)   (55.57%)   (50.62%)
Sharpe Ratio             0.58       0.42       0.58       0.60
Sortino Ratio            0.86       0.66       0.86       0.88
US Mkt Correlation       0.89       0.08       0.89       0.89
The effect of re-balancing is still negative but only 0.27% points (14.05% - 13.78%).

Before this thread I assumed that re-balancing between stocks and bonds almost always increased return. Now I see this isn't so. Re-balancing them is still good; but as a way to maintain the risk level, not as a way to increase return.
User avatar
willthrill81
Posts: 32250
Joined: Thu Jan 26, 2017 2:17 pm
Location: USA
Contact:

Re: Vanguard Target Retirement 2065

Post by willthrill81 »

#Cruncher wrote:Thanks for mentioning Portfolio Visualizer. I hadn't seriously used it before. But now I see it's powerful and easy to use.
You're welcome. I really like it. The Monte Carlo simulation is really useful too.
#Cruncher wrote:The effect of re-balancing is indeed significantly negative in this case as you say. But note how much riskier is the "50:50" allocation without re-balancing. This shouldn't be too surprising because without re-balancing what starts out as 50:50 ends up 95:5 [95% = 4336 / (4336 + 235)].
It is indeed 'riskier' to not rebalance, but this is a bit of a 'false' risk in comparison to rebalancing. The additional risk from not rebalancing comes entirely from the added gain that would not have been present had the portfolio been rebalanced. In a very real way, it is only the 'extra' money you get from not rebalancing that is subject to the added risk.
#Cruncher wrote:I'm surprised you chose a 50:50 allocation, willthrill81.
No reason it particular other than the fact that it illustrates that in most situations, portfolios that are not rebalanced have become increasingly stock heavy as the stocks gain more and more ground over bonds.
#Cruncher wrote:The effect of re-balancing is still negative but only 0.27% points (14.05% - 13.78%).
Yes, the difference clearly shrinks as the portfolio becomes very stock heavy, but you know better than most what an additional .27% annual return can do to a portfolio over decades.
#Cruncher wrote:Before this thread I assumed that re-balancing between stocks and bonds almost always increased return. Now I see this isn't so. Re-balancing them is still good; but as a way to maintain the risk level, not as a way to increase return.
Rebalancing is good in the sense that it maintains your desired AA and its associated risk, but it can be very bad in the sense of having a big cost associated with lower returns as well. Sure you can experience more volatility by not rebalancing, but the money being 'risked' is money you would not have had if you had rebalanced.

I think that this persistent myth of a worthwhile rebalancing bonus comes from the idea that "you can buy stocks when they're cheap with your bonds," while not considering that the money in bonds is holding back the portfolio while the stocks are going up. Given that bull markets tend to surpass the bear markets, this strategy just doesn't pan out the way many think it does.

On another note, thank you for being so cordial in your response. The default response for most is to get very defensive and insist that their initial position is right come heck or high water. I respect your openness to new ideas.
The Sensible Steward
Post Reply