Here's a statement of the obvious: The opinions expressed here are those of the participants, not those of the Mutual Fund Observer. We cannot vouch for the accuracy or appropriateness of any of it, though we do encourage civility and good humor.
FYI: 3 Is Company Three colors: red, white, and blue. Three funds: Vanguard Total Stock Market Index (VTSAX), Vanguard Total International Stock Index (VTIAX), and Vanguard Total Bond Market Index (VBTLX). According to author Taylor Larimore, those are all that an American investor needs. Regards, Ted https://www.morningstar.com/articles/871546/the-3fund-portfolio.html
The article reviews Taylor Larrimore's "The Bogleheads' Guide to the Three-Fund Portfolio." Rekenthaler concludes that Vanguard should simply offer one fund comprised of a US equity index, international equity index and bond index. JR writes in the essay, "it would be simpler yet if it were a one-fund portfolio. It strikes me that Vanguard might do its investors a favor by bundling Taylor's three funds into a single offering." I wrote John today, asking why the Fidelity Four-in-One Index (FFNOX) doesn't satisfy his desire for a simple index of indexes.
He notes, by the way, that Taylor is 94 now, having been born in the same year as the first mutual fund.
I'll let you know if I hear anything interesting back.
Well, every Vanguard Target Retirement fund is a combination of the three funds in the article plus Total International Bond so... they already kind of do?
Well, what about Vanguard Developed Market Index? I don't see anyone ever mentioning it and wonder why it exists. Basically looks like Global fund without Emerging Markets?
Then how about using above fund with VEIEX and Vanguard Total Bond Market Index? 3-funds again.
Better question - who is going to write article mentioning above 3-fund portfolio?
I believe the difference between the three fund portfolio as proposed by Taylor and the Fidelity fund cited isa matter of weighting.15% bonds as in the Fidelity fund is probably fine until sometime in the 40s but is probably too aggressive for somewhat older investors like Taylor
msf has already pointed out the lack of advantages of the AO_ family when I noted their simplicity, but they too (AOA, AOR, AOM, AOK) address some of these concerns. Maybe. e.g., https://www.ishares.com/us/products/239756/
And I believe others have pointed toward various target-date and glidepath offerings.
The idea behind cap weighted indexing is that you duck the risk (and benefit) of issue selection by buying everything.
I've never seen anyone advocate choosing a stock/bond ratio to match the ratio of worldwide equity to worldwide debt. So implicitly at least, people tend to regard asset allocation differently from issue selection.
Hence the suggestion that asset allocation be left to the investor. But to what degree? For example, there's VT, that covers all the stock in the world. That can certainly take the place of two funds, VTSAX and VTIAX. It also takes away a degree of flexibility, but is that flexibility needed?
Maybe so; people may want finer granularity of their asset classes. So the first question is how does one define one's asset classes (finely or coarsely)? After that comes the question of picking the preferred mix.
The most obvious way to construct a portfolio having answered these questions is to pick broad based funds, one for each asset class, and then mix them in the desired proportions.
The difficulty in using all in one funds is that they must roughly match your preferred allocations. FFNOX, with its 85% share in equity just isn't going to do it for you if you want a 40/60 stock/bond mix.
Fund families address this by offering a suite of funds, each with a different allocation. Vanguard has its LifeStrategy funds. If one of those works for you, great.
But there's still the matter of what classes of assets they're allocating. The Vanguard funds appear to be offering various admixtures of US stock, US bond, foreign stock, and foreign bond. But in reality, all their LifStrategy funds have the same ratio of US stock to foreign stock, so it's as if they simply used a single global stock fund. Likewise, all the funds have the same US/foreign bond ratio, so it's as if they used a single global bond fund.
Rather than offering a four-dimensional choice of allocations, you've just got a two dimensional offering - the only thing that varies is the ratio of global stocks to global bonds. 20/80, 40/60, 60/40, 80/20.
It's just impractical to offer more. Thinking about US stock/Int'l stock/US bond/foreign bond combinations, there could be: 10/10/10/70, 10/10/30/50, 10/10/50/30, 10/10/70/10, 10/30/10/50, 10/30/30/30, 10/30/50/10, 10/50/10/30, 10/50/30/10, 10/70/10/10, 30/10/10/50, 30/10/30/30, 30/10/50/10, 30/30/10/30, 30/30/30/10, 30/50/10/10, 50/30/10/10
That's 17 funds right there. This is why you're not likely to find an all in one fund with your desired asset allocation. Thus all in one funds don't work unless you give up a large chunk of control over your asset allocations.
I had one family member, whom I advised for free (talk about getting what you pay for), who ultimately decided upon $5k each into AOA, AOR, AOM, and AOK, rebalancing every few months (free, at ML). Now each account worth more, of course; this was a while ago. They felt (rightly or wrongly ) they had good control over allocations. It sounded a little sketchy to me, but turned out to be pretty simple and straightforward, or was on paper, since it was not ever fully clear to me exactly how they decided to adjust and shift funds. Timing, iow. I guess they were being their own dynamic index fund manager, like so many in other etfs. Worked fine, but better than other approaches?
Give or take rounding error, Barclays is doing the same thing as Vanguard. All the sibling funds hold the same underlying funds (though unlike Vanguard, these may vary from time to time).
If one looks only at the underlying equity funds, these are held in the same proportions across all sibling funds. For example, 80% of AOA is in equity funds. Of those equity holdings, 48% are in an S&P 500 fund, 38% are in a developed market fund, etc.
60% of AOR is in equity. Of those equity holdings, the breakdown is the same: 48% in the S&P 500 fund, 38% in the developed market fund, and so on.
The only difference among all these funds is the stock/bond ratio. They are 80/20, 60/40, 40/60, and 30/70.
So your family member was getting a 52.5/47.5 stock/bond portfolio by allocating evenly across all the funds. (Just average the percentages.) Rebalancing simply maintained this stock/bond ratio. Had Barclays offered a variant with that particular ratio, all your family member would have needed would have been that single fund. Or if a 60/40 split rather than a 52.5/47.5 had been acceptable, then buying a single fund - AOM - would have sufficed.
Regardless, since there is only one degree of freedom (the stock/bond ratio), the same portfolio could be achieved with a mix of just two funds. This is a simple linear algebra problem, or could even be viewed as an 8th grade math mixture problem:
You have $100 to spend. You want to buy $52.5 of stock (and the rest in bonds).
For every $1 you spend on AOR, you'll get $0.60 in stock (and $0.40 in bonds). For every $1 you spend on AOM, you'll get $0.40 in stock (and $0.60 in bonds).
How much should you spend out of $100 on each fund to wind up with $52.5 in stock?
Let x = amount spent on AOR, so ($100 - x) is the amount spent on AOM. So 0.60 * x + 0.40 * ($100 - x) = $52.5 0.20 * x + $40 = $52.5 0.20 * x = $12.5 x = $62.5
We can check this: $62.5 in AOR buys 60% x $62.5 = $37.5 of equity (and $25 of bonds) $37.5 in AOM buys 40% x $37.5 = $15 of equity (and $22.5 of bonds).
Together, that's $52.5 of equity, and $47.5 of bonds.
A common complaint against 401k plans offering too many funds is that because participants don't know what to do with all the choices, they tend to simply buy the same amount of each fund. With these iShares, the result was reasonable - something close to a 50/50 allocation. But with 401k plans, where you might have 17 equity funds and 3 bond funds, a participant could wind up with a portfolio that was 85% in equity (looks like we're back to FFNOX), whether that was suitable or not.
>> only one degree of freedom (the stock/bond ratio),
In their early days I do not believe (memory) they were as fixed and quite as formulaic on the 7 noncash classes as they are now, as I suspected then what you say and found the p/e to vary and the proportions likewise (a little). I think there was a smidgen of RE or something else at the outset, although I don't know how to find holdings makeup history. Even now, if you drill down, p/e varies (not over a wide range), as do the proportions trivially (AOM is slightly different from the others in spots, for some reason). I am not clear why the p/e would differ now. But the four now comprise the same 7 etfs, yeah, no point in faux rebalancing. There's not enough SC and MC in any case for many tastes, and too much international for mine.
Regardless, rest assured I shan't mention the AO_ group here again.
You have a good memory. Earlier prospectuses (e.g. this one from 2013) didn't imply there were fixed stock/bond ratios, but the current prospectus does:
[For AOM] As of July 31, 2017, the Underlying Index included a fixed allocation of 60% of its assets in Underlying Funds that invest primarily in equity securities and 40% of its assets in Underlying Funds that invest primarily in bonds. As of July 31,2017, the Fund invested approximately 63.57% of its assets in Underlying Funds that invest primarily in equity securities, 36.24% of its assets in Underlying Funds that invest primarily in bonds and the remainder of its assets in Underlying Funds that invest primarily in money market instruments.
The 60/40 seems to be a target, since the next sentence gives the actual allocations.
Looking at the funds' allocations for April 30, 2014, AOA had 3.77% in Cohen & Steers REIT ETF (ICF) (see page here), while AOK had none (see page here). So the funds used to include different underlying funds, including RE.
Though as you suggested, not really enough to make much of a difference.
Well, they all started in late fall 2008. A good time for anything to start, without question. But that means there was a healthy stretch of slightly variant asset allocation from the current state.
Making any worthwhile difference? Dunno, but not much the last decade has made for true diversification in the usual senses, has it?
Back when I was recommending them for simplicity (many folks very much like to get a piece of paper with a single entry or small handful of entries on it that they think are a good representation of something or other), I too believed in the virtues of international, RE, and SC/MC in countering LC.
Whatever. It is instructive (for me anyway) to graph all of them from 11/08 on against various stellar balanced funds, just for kicks, to see how much is (is not) added with active management. I just now compared FBALX, DODBX, FPACX, OAKBX, JANBX vs AOR and AOA.
Is there data on actual amounts invested in different investment categories globally? For example size of total US equity market, total international equity market (broken into classes such as developed vs. emerging, for example), total US bond market, total global bond market. It is harder to draw a line regarding other investing options, but we can stay with 'conventional retail investor liquid options' for lack of better wording to exclude markets for currency, commodities, physical assets (gold, houses, properties), art, etc. Idea was to invest according to size of market (cap weighted).
Comments
He notes, by the way, that Taylor is 94 now, having been born in the same year as the first mutual fund.
I'll let you know if I hear anything interesting back.
David
Then how about using above fund with VEIEX and Vanguard Total Bond Market Index? 3-funds again.
Better question - who is going to write article mentioning above 3-fund portfolio?
And I believe others have pointed toward various target-date and glidepath offerings.
I've never seen anyone advocate choosing a stock/bond ratio to match the ratio of worldwide equity to worldwide debt. So implicitly at least, people tend to regard asset allocation differently from issue selection.
Hence the suggestion that asset allocation be left to the investor. But to what degree? For example, there's VT, that covers all the stock in the world. That can certainly take the place of two funds, VTSAX and VTIAX. It also takes away a degree of flexibility, but is that flexibility needed?
Maybe so; people may want finer granularity of their asset classes. So the first question is how does one define one's asset classes (finely or coarsely)? After that comes the question of picking the preferred mix.
The most obvious way to construct a portfolio having answered these questions is to pick broad based funds, one for each asset class, and then mix them in the desired proportions.
The difficulty in using all in one funds is that they must roughly match your preferred allocations. FFNOX, with its 85% share in equity just isn't going to do it for you if you want a 40/60 stock/bond mix.
Fund families address this by offering a suite of funds, each with a different allocation. Vanguard has its LifeStrategy funds. If one of those works for you, great.
But there's still the matter of what classes of assets they're allocating. The Vanguard funds appear to be offering various admixtures of US stock, US bond, foreign stock, and foreign bond. But in reality, all their LifStrategy funds have the same ratio of US stock to foreign stock, so it's as if they simply used a single global stock fund. Likewise, all the funds have the same US/foreign bond ratio, so it's as if they used a single global bond fund.
Rather than offering a four-dimensional choice of allocations, you've just got a two dimensional offering - the only thing that varies is the ratio of global stocks to global bonds. 20/80, 40/60, 60/40, 80/20.
It's just impractical to offer more. Thinking about US stock/Int'l stock/US bond/foreign bond combinations, there could be:
10/10/10/70, 10/10/30/50, 10/10/50/30, 10/10/70/10, 10/30/10/50, 10/30/30/30, 10/30/50/10, 10/50/10/30, 10/50/30/10, 10/70/10/10, 30/10/10/50, 30/10/30/30, 30/10/50/10, 30/30/10/30, 30/30/30/10, 30/50/10/10, 50/30/10/10
That's 17 funds right there. This is why you're not likely to find an all in one fund with your desired asset allocation. Thus all in one funds don't work unless you give up a large chunk of control over your asset allocations.
They felt (rightly or wrongly ) they had good control over allocations. It sounded a little sketchy to me, but turned out to be pretty simple and straightforward, or was on paper, since it was not ever fully clear to me exactly how they decided to adjust and shift funds. Timing, iow.
I guess they were being their own dynamic index fund manager, like so many in other etfs. Worked fine, but better than other approaches?
If one looks only at the underlying equity funds, these are held in the same proportions across all sibling funds. For example, 80% of AOA is in equity funds. Of those equity holdings, 48% are in an S&P 500 fund, 38% are in a developed market fund, etc.
60% of AOR is in equity. Of those equity holdings, the breakdown is the same: 48% in the S&P 500 fund, 38% in the developed market fund, and so on.
The only difference among all these funds is the stock/bond ratio. They are 80/20, 60/40, 40/60, and 30/70.
So your family member was getting a 52.5/47.5 stock/bond portfolio by allocating evenly across all the funds. (Just average the percentages.) Rebalancing simply maintained this stock/bond ratio. Had Barclays offered a variant with that particular ratio, all your family member would have needed would have been that single fund. Or if a 60/40 split rather than a 52.5/47.5 had been acceptable, then buying a single fund - AOM - would have sufficed.
Regardless, since there is only one degree of freedom (the stock/bond ratio), the same portfolio could be achieved with a mix of just two funds. This is a simple linear algebra problem, or could even be viewed as an 8th grade math mixture problem:
You have $100 to spend. You want to buy $52.5 of stock (and the rest in bonds).
For every $1 you spend on AOR, you'll get $0.60 in stock (and $0.40 in bonds).
For every $1 you spend on AOM, you'll get $0.40 in stock (and $0.60 in bonds).
How much should you spend out of $100 on each fund to wind up with $52.5 in stock?
Let x = amount spent on AOR, so ($100 - x) is the amount spent on AOM.
So 0.60 * x + 0.40 * ($100 - x) = $52.5
0.20 * x + $40 = $52.5
0.20 * x = $12.5
x = $62.5
We can check this: $62.5 in AOR buys 60% x $62.5 = $37.5 of equity (and $25 of bonds)
$37.5 in AOM buys 40% x $37.5 = $15 of equity (and $22.5 of bonds).
Together, that's $52.5 of equity, and $47.5 of bonds.
A common complaint against 401k plans offering too many funds is that because participants don't know what to do with all the choices, they tend to simply buy the same amount of each fund. With these iShares, the result was reasonable - something close to a 50/50 allocation. But with 401k plans, where you might have 17 equity funds and 3 bond funds, a participant could wind up with a portfolio that was 85% in equity (looks like we're back to FFNOX), whether that was suitable or not.
In their early days I do not believe (memory) they were as fixed and quite as formulaic on the 7 noncash classes as they are now, as I suspected then what you say and found the p/e to vary and the proportions likewise (a little). I think there was a smidgen of RE or something else at the outset, although I don't know how to find holdings makeup history. Even now, if you drill down, p/e varies (not over a wide range), as do the proportions trivially (AOM is slightly different from the others in spots, for some reason). I am not clear why the p/e would differ now.
But the four now comprise the same 7 etfs, yeah, no point in faux rebalancing. There's not enough SC and MC in any case for many tastes, and too much international for mine.
Regardless, rest assured I shan't mention the AO_ group here again.
Looking at the funds' allocations for April 30, 2014, AOA had 3.77% in Cohen & Steers REIT ETF (ICF) (see page here), while AOK had none (see page here). So the funds used to include different underlying funds, including RE.
Though as you suggested, not really enough to make much of a difference.
Well, they all started in late fall 2008. A good time for anything to start, without question. But that means there was a healthy stretch of slightly variant asset allocation from the current state.
Making any worthwhile difference? Dunno, but not much the last decade has made for true diversification in the usual senses, has it?
Back when I was recommending them for simplicity (many folks very much like to get a piece of paper with a single entry or small handful of entries on it that they think are a good representation of something or other), I too believed in the virtues of international, RE, and SC/MC in countering LC.
Whatever. It is instructive (for me anyway) to graph all of them from 11/08 on against various stellar balanced funds, just for kicks, to see how much is (is not) added with active management. I just now compared FBALX, DODBX, FPACX, OAKBX, JANBX vs AOR and AOA.