A Portfolio Makeover

Last September, I wrote a piece for Canadian MoneySaver called An ETF Portfolio With Added Dimension. The article looked at the strategies used by Dimensional Fund Advisors, “the best money management firm that most people have never heard of.” The portfolio I put together for that article became the Über-Tuber — as in “the ultimate Couch Potato” — which you’ll find on my page of Model Portfolios.

If you’re not familiar with Dimensional Fund Advisors, that’s not surprising: the firm keeps a low profile, and its funds are offered only through a select group of advisors who must first undergo extensive training. DFA’s investing strategies are based on the academic work of Eugene Fama and Kenneth French, whose research demonstrated that value stocks and small-cap stocks have historically delivered higher returns than the overall market. (For more details, see the DFA website and my own post, Where Do Returns Come From?) The point of my article was to suggest a way for do-it-yourself investors to use Dimensional’s passive strategies with ETFs.

I later learned that the article made the rounds among DFA advisors, who gave it a mostly positive review, though with some criticisms. I recently spoke about this with Justin Bender, an advisor with PWL Capital in Toronto who uses both DFA funds and ETFs with his clients. Using Morningstar’s PALTtrak research tool, Justin drilled into the Über-Tuber to compare its asset class exposure to what you would get in an actual Dimensional portfolio like this one:

20% DFA Canadian Core Equity Fund (DFA256)
16% DFA US Core Equity Fund (DFA293)
20% DFA International Core Equity Fund (DFA295)
4% DFA Global Real Estate Securities Fund (DFA391)
20% DFA Five-Year Global Fixed Income Fund (DFA231)
20% DFA Investment Grade Fixed Income Fund (DFA449)

Then he made a number of suggestions to bring the ETF portfolio much closer to being a DFA clone. Here’s what he came up with:

10% iShares S&P/TSX Capped Composite (XIC)
5% Claymore Canadian Fundamental (CRQ)
5% iShares S&P/TSX Small Cap (XCS)
11% PowerShares FTSE RAFI US 1000 (PRF)
5% Vanguard Extended Market (VXF)
4% PowerShares FTSE RAFI Dev Mkts ex-US (PXF)
5% Vanguard MSCI EAFE (VEA)
6% PowerShares FTSE RAFI Dev Mkts ex-US Small-Mid (PDN)
5% Vanguard Emerging Markets (VWO)
4% Claymore Global Real Estate (CGR)
30% iShares DEX Short-Term Bond (XSB)
10% iShares US IG Corporate Bond (CAD-Hedged) (XIG)

Peeling back the Potato

Justin provided some commentary to explain the changes:

  • “The original Über -Tuber had 40% of its Canadian equity allocation in small cap stocks, but the DFA Canadian Core Equity Fund has just 15% to 20% in small cap stocks. I reduced the small-cap allocation to reduce risk in the portfolio. The original portfolio also had an underweight in energy and materials stocks and an overweight to financial stocks. I reduced exposure to CRQ and added XIC to better mimic the DFA fund and also to lower the overall MER.”
  • “I found that 11% PRF combined with 5% VXF [a fund of US mid- and small-cap stocks] gave the appropriate small-cap and value tilts relative to the DFA US Core Equity Fund.”
  • “The original Über -Tuber had too much international large cap and not a sufficient amount of small cap, relative to the DFA International Core Equity Fund. I replaced the Vanguard All World ex-US Small-Cap (VSS) with PDN to gain more value exposure while also increasing the exposure to mid and small-cap companies. I replaced the iShares MSCI EAFE Value (EFV) with VEA, which helped reduce costs as well as remove the overallocation to large value and spread it more evenly across the other size and style categories. The sector allocations also aligned more closely to the DFA fund after these changes.”
  • “Using equal amounts of the two DFA fixed-income funds results in an allocation of approximately 25% to US bonds (with currency hedging). We can get a similar exposure by allocating 10%  to XIG and the other 30% to XSB. The approximate maturities, as well as the mix of government and corporate bonds, also mirror the DFA model portfolio more accurately than a single position in XSB.”

Too many moving parts?

The Über-Tuber started off as an intellectual exercise more than anything else. As a couple of DFA advisors have told me, it has a lot of moving parts, which will make it difficult and potentially expensive to maintain: the original was already a behemoth with 11 ETFs, and the above version uses an even dozen. Is it too unwieldy?

If you’re going to build a portfolio like this, a low-cost brokerage is a must. Your account would have to be quite large before you even consider it: $100,000 is the bare minimum, and unless you’re investing at least $300,000 or so there is probably no meaningful benefit over the Complete Couch Potato. Rebalancing could be something of a nightmare — I would suggest doing so only when the allocations are way out whack.

In fact, the more I speak with experienced advisors who understand investor behaviour, the more I have come to value simplicity. The urge to tinker with a complex portfolio is just too great. I’m not going to make any changes to the Über-Tuber on the Model Portfolios page for now, and I’ll be giving some thought to how it might be streamlined without losing too much of its small-cap and value tilt.

In the meantime, I’d love to hear from readers who are currently using the Über–Tuber or something like it — I know you’re out there. How are you managing?

31 Responses to A Portfolio Makeover

  1. Neil May 30, 2011 at 9:33 am #

    Do you ever assess the correlation between these asset classes? It seems to me they are all just variations on the basic bond/equity/international diversification strategy. What is the real advantage to adding so many different components?

  2. Canadian Couch Potato May 30, 2011 at 9:51 am #

    @Neil: Great question. The short answer is, yes, there, has been a lot of work done on asset class correlation, and there is a diversification benefit to splitting domestic and international stocks, tilting to value and small stocks, adding real estate, etc. The most accessible book on this topic is Rick Ferri’s All About Asset Allocation. That said, you can certainly reach a point where further diversification has no benefit and only adds complexity. I would think that 11 or 12 ETFs is right at that upper limit.

  3. DM May 30, 2011 at 10:39 am #

    Dan my current portfolio generally resembles the Uber Tuber, with a bit more devoted to fixed income and Canadian dividend (CDZ). I’ve had it in place for about 2 years and it has performed reasonably well. My return for calendar year 2010 was in neigbourhood of 9%. The problem is it drives me crazy. I have it spread out over 5 accounts: my RSP, my wife’s RSP, my TFSA, my wife’s TFSA and a joint non-registered account. So this set up imposes many constraints that I need to manage in addition to periodic rebalancing. For example, foreign equity should be help in RSPs, to the extent possible Canadian dividend paying ETFs should be in non-registered, etc. Maybe it’s me getting lazy, but I find the value added of maintaining such as complex portfolio is not worth the burden it imposes on me in terms of managing it. I am inclined to shed many of the more esoteric positions (e.g. VSS) and return to something more closely resembling the Complete Couch Potato. Maybe you’ve posted it already but I’d love to see how a simple and easy to maintain portfolio like the CCP stacks up against the Uber Tuber in terms of total return over last 5 years.

  4. Canadian Couch Potato May 30, 2011 at 10:47 am #

    @DM: Thanks for sharing — you’ve confirmed my suspicion. A portfolio that drives you crazy is not one you’re going to stick to over the long term. I wouldn’t blame yourself for being lazy: laziness is is good quality in investors, so long as they have set up a good plan.

    It’s not possible to get a five-year performance report on the model portfolios, because most of the ETFs have not been around that long. But I would guess that any difference in performance would show up to the right of the decimal point.

    Next time you add money to your accounts or need to rebalance, I would agree you’re better off trimming things back and making the whole process simpler.

  5. Charles Chase May 30, 2011 at 11:20 am #

    This is very timely since I’m getting ready to switch into an ETF portfolio and took the uber-tuber as a starting point, trying to simplify it somewhat. I’m just waiting for funds to transfer before I begin buying.

    If you want to reduce the number of funds I don’t think you can get to less than eight if you want to keep all the elements. For Canadian, American and International equity you’ll want a big cap (fundamental or otherwise) and a small cap fund. So that’s six right there. Then you have bonds and real estate. I don’t think you can find any ETFs that will consolidate any of these eight categories. I might consider dropping the real estate fund, leaving seven funds to deal with.

    However, is a portfolio of eight to twelve funds really all that “difficult and potentially expensive”? If you have $100k you should qualify for trades of $7 or less. To rebalance the whole thing should therefore cost less than $100 (0.1%). As for difficulty, I don’t think rebalancing that number of funds should be difficult for anybody who has ever used a spreadsheet. For each fund you should probably keep an eye out in case they stop tracking correctly or start putting out too many divestments but I don’t see that being a big problem. Are there other potential difficulties I’m missing?

    Thanks for all the good work Dan.

  6. Canadian Couch Potato May 30, 2011 at 11:32 am #

    @Charles: Everything you say is correct in theory. It’s certainly possible to manage a complex portfolio if you have discipline and some skill with a spreadsheet. However, I think the reality is that many people get caught up in the minutiae and lose sight of the big picture. For all but the largest portfolios, adding 4% to a secondary asset class is probably not worth the trouble or the added expense.

    Do let me know how you make out with your own portfolio!

  7. Charles Chase May 30, 2011 at 1:16 pm #

    I had not read DM’s comment before I posted. He makes a good point about managing one portfolio across three accounts, which most of us do in order to take advantage of RRSP and TFSA tax savings. That issue does increase the difficulty and potential cost of using multiple funds. In the worst case scenario after a few years of rebalancing you could end up with a small amount of each fund in every account, which certainly isn’t ideal.

  8. qasimodo May 30, 2011 at 6:39 pm #

    Dan

    It is common to be concerned about transaction costs, but I never hear any concern about the impact of odd lot sizes. Any thoughts?

    (I manage 2*RSP, 1* spousal, 2*TFSA, 1*non-sheltered and use TD e-funds to manage both txn costs and lot sizes, until they accumulate efficient sums)

    q

  9. Steve in Oakville May 31, 2011 at 8:22 am #

    @CCP – “But I would guess that any difference in performance would show up to the right of the decimal point” – Yes Dan, I totally agree. When I started putting together my portfolio, I fretted about all these percentages and allocations and questions. But the truth is that the model portfolios all have the same basic asset allocation breakdown, and that is going to be the better predictor of return than how each asset class is constructed and which ETFs are used for each asset class. And therefore, I’ve come to value simplicity…

    @Qasimodo – I use a version of the Cheapskate portfolio as the core of my RRSP, with XIC for the CDN content instead of HXT. All of the ETFs in that portfolio trade with enough volume that the spreads aren’t significant. I’ve never had an issue buying, say, 48 shares instead of 100. On thinly traded ETFs, maybe limit orders would solve the odd lot problem.

  10. Canadian Couch Potato May 31, 2011 at 8:44 am #

    @qasimodo: I have never had a problem with odd lots either. I think once or twice my orders were filled in two steps, but not with any significant difference in price. As Steve suggests, a limit order would ensure you don’t have any surprises.

    @Steve: Thanks for the insight. Designing the “perfect” portfolio is a fun exercise, but once you get down to actually implementing and maintaining it, things become less entertaining. I think all of us would be better off devoting our energy to saving more money, which is likely to have a much larger effect on our wealth.

  11. Greg May 31, 2011 at 11:14 am #

    I came to the comments and found DM had written what I planned to write.

    I, too, have our family investments in 5 accounts. But, being retired and less than 65, we also have a sizeable chunk of cash in two High Interest bank accounts as this year’s working capital. (One is for Ally’s higher interest, the second is for ING’s speed of transfer to RBC.

    For me, that cash component has to fall outside the normal asset allocation. I do track the income on the high interest as “investments” in Quicken.

    Re-balancing has not a high priority as I have been fortunate to have “new cash” arrive in the form of inheritances and income of sale of former home and so have used that “new cash” (which fortunately didn’t arrive as a single lump sum) to get my asset allocation closer to what I want, with closer being the operative word. Initially, I wanted it “right” to 2 decimal places, but soon realized that was not the best strategy for me.

    Most important for me is to track performance relative to “reality”. I found Croft’s benchmarks ( http://croftgroup.com/indexes/calculator.asp ) and use them with my Quicken performance data as a measure of how I am actually doing.

    It will never be perfect. I just want it as close as I can get, while maintaining sanity.

  12. Pinery May 31, 2011 at 11:21 am #

    Here is an interesting article by Andrew hallam, author of Millionaire Teacher explaining why he keeps things simple and not “Slice and Dice” to garner excess returns. Right now my portfolio is similar the “Uber-Tuber”.
    http://andrewhallam.com/2011/01/where-did-my-money-go/#comment-1929
    When i rebalance, i’m seriously considering keeping things simple by moving to a portfolio similar to the complete couch potato.

  13. Eric May 31, 2011 at 12:16 pm #

    Hi Dan,

    I believe it’s important to have exposure to value & small-cap stocks. It can reduce volatility & increase returns (as confirmed by many studies). I do it using dividend ETF’s. My portfolio is an hybrid between your yield-hungry couch potato & portfolio with added dimension (Canadian MoneySaver).

    If an investor wants to keep it simple and add a strong value & small-cap tilt to his/her portfolio, 8 ETF’s is all you need:

    35% XSB

    14% CRQ
    6% XCS

    14% PRF
    6% VBR (Vanguard small value)

    12% PXF
    8% PDN

    5% CGR or VWO or both

    I use Instant X-Ray at Morningstar for my assets allocation.
    If you get into a more complex portfolio, a spreadsheet is essential and new money (RSP contributions or dividends) should be use for rebalancing.

  14. Canadian Couch Potato May 31, 2011 at 12:34 pm #

    @Eric: Nice portfolio, thanks. Eight funds are certainly easier to manage than 12, and you’ve got all the major pieces in there. I’m surprised that you find this portfolio less volatile than a more basic one — with all that small and value exposure, I would think you’d be more volatile.

  15. Eric May 31, 2011 at 1:40 pm #

    @Dan: less volatile beacause more different asset classes with less correlation.

    3 year total return NAV (05/27 ) from morningstar:

    CRQ: 4.4% XCS: 5.9% vs XIC: 0.7%
    PRF: 4.9% VBR: 5.4% vs VTI: 2.12%
    PXF: -2.4% PDN: 4.3% vs: VEA: -4.3%

    as you can see, adding value & small-cap can help increase your returns. I don’t think this portfolio will win every year but it should give you a ”smoother” ride.

  16. Eric May 31, 2011 at 2:02 pm #

    @Dan: This portfolio was just to illustrate my point. I don’t use RAFI based indexes. I find them over concentrated. CRQ – 45% financials. PDN – 33% in Japan.

  17. Pinery May 31, 2011 at 3:46 pm #

    @Eric
    We are all on this forum to learn. You suggested a theoretical portfolio with Rafi fundamental indexes and then later said you don’t use Rafi indexes in your portfolio because they are too concentrated which is a very valid point because of the 4 parameters Rafi uses to constitute the Rafi indexes( ie sales , dividend, Cash flow and book value) . This will explain why financials and Japan are overweight in Rafi indexes. So which index do you use in your portfolio that closely resemble DFA portfolios.

  18. Simon June 1, 2011 at 12:18 am #

    Eric,

    You mentioned that you use Instant X-Ray at Morningstar for my assets allocation. This is a really neat tool, but how do you get it to analyze CAD ETFs such as XIU.TO and XSB.TO?

  19. Canadian Couch Potato June 1, 2011 at 12:23 am #

    @Eric and Simon: I was going to ask the same question about X-Ray. Do you need to pay for a subscription to get full access to this service?

  20. Simon June 1, 2011 at 9:28 am #

    @CCP: I believe if you want to see more detailed stats on a portfolio, you’ll need to register and sign up for the pay-premium version. However, the free version already seems pretty detailed, giving enough information to go on for most users.

    The biggest problem that I have seen is that it does not work for Canadian-based ETFs.

    Here are two links:
    http://portfolio.morningstar.com/RtPort/Free/InstantXRayDEntry.aspx
    http://portfolio.morningstar.com/NewPort/Free/InstantXRayDEntry.aspx

  21. Eric June 1, 2011 at 11:08 am #

    @Dan & Simon: I use X-Ray only for US ETFs. For more infos on Can ETFs , I look at VTI (or any US ETFs), go to Portfolio, than enter the Can ETF you want.

  22. Eric June 1, 2011 at 12:15 pm #

    @Pinery: I don’t know much about DFA only that they follow Fama-French research, which is that value stocks have a much higher rate of return without much additionnal risk.

    As for your link to Hallam’s post, it always depends on the period you are looking at:
    from 1975 to 1995, for a US investor: 40% short bond , 30% S&P 500 & 30% EAFE = 14.05% average return.
    Now ”slice & dice” stocks: 7.5% each (S&P500-large value-small-small value) +
    7.5 % each EAFE & large foreign value & 15% foreign small = 17.15% average return.
    (datas from : Invest. strategies for the 21st century by Frank Armstrong)

    For my portfolio I use broad market index ETFs and dividend ETFs to add a value tilt.
    In Can: CDZ & XDV USA & Intern: CVY & DWX
    My portfolio is still a work in progress…

  23. Charlotte September 3, 2011 at 9:37 am #

    I am just beginning to learn about investing (I am one of those who blindly accepted the underperforming mutual fund with a 2% MER that my RRSP agent gave me years ago) so please bear with me. Why is the recommended stock portion usually split about 30% Canadian, 35% US and 35% international? I understand the reasoning behind allocations of stock vs bond vs cash/gold etc, but why would the stocks not be, say, 50% canadian and 50% international?

  24. Martin January 26, 2012 at 7:12 pm #

    Dan,

    First time posting, but I look at the site every day…thanks for all the great information provided by you and the other readers. I’ve been building a Couch Potato portfolio over the past year. One of the challenges I face is that as I read and learn more about passive portfolio construction, my thinking evolves and I tweak the portfolio in spite of the knowledge that asset allocation should remain consistent over time.

    For example, my Canadian equity component is split among XIU, CDZ and more recently Vanguard Canadian equity VCE. I don’t feel that this is a big problem in and of itself…except that the portfolio is a little more Frankenstein than I want.

    Lately I feel I’m missing out by not applying a value/small cap tilt – hence posting on this thread. In the interest of avoiding more tweaking (and transaction costs) than necessary, I don’t want to change my core holdings eg. VCE/VTI/VXUS in favour of the Powershares and Claymore RAFI ETF’s etc.

    I’d appreciate any opinions on achieving a small-cap/value tilt by adding 5% holdings of XCS, VBR and VSS to my Canadian, American and International allocations. I realize that it woudn’t work as well as the made-over Uber Tuber, but I’m hoping it would capture some of the benefit of a small cap/value tilt the without completely revamping the portfolio.

    Thoughts?

  25. Canadian Couch Potato January 29, 2012 at 9:12 am #

    @Martin: Thanks for the comment. I sympathize with your plight. I too used to be a tinkerer, always looking for ways to improve the asset mix and find the perfect asset mix. After a lot of reading, I eventually made peace with the idea that there is no such thing. There are a lot of excellent portfolios, but none is perfect. I wrote about it here:
    http://www.moneysense.ca/2010/12/14/whats-on-your-potato/

    I would be wary of adding too many ETFs to any portfolio, especially if it is moderate in size. This is especially true if you are adding 5% positions in value or small-cap funds. Remember that one of these holding would have to increase by 20% in order to move the portfolio 1%. Is this added complexity really worth it? Remember, too, that after you add these positions you will probably have the urge to tweak things again.

    Personally, I think the Complete Couch Potato is more than adequate for the vast majority of investors, especially those with portfolios under $200,000 or so. Bottom line: try to settle on an allocation you’re comfortable with, and then stick to it. Good luck!

  26. Martin January 29, 2012 at 2:56 pm #

    Thanks Dan,

    I think I’ll add a 10% small cap/value component and then try to just leave it alone!

    Portfolio is in the high six figures, so I believe it should be a worthwhile addition; with a negligible percentage lost to transaction costs.

  27. Chris February 4, 2012 at 11:30 am #

    Hi

    First time posting. After my 70/30 split equity/fixed income mutual fund portfolio only came up with around a 2% return over the past 15 years, I have given up on it. I am chalking it up to fees, hence the shift to ETFs. Thanks so much for this site, it has been very helpful.

    I have been selling the mutual funds and switching out to ETF for both my wife and my own accounts. I have felt fairly comfortable on the fixed income side but I feel a bit stumped on the equity side. Two issues, one is the RAFI versus MSCI. I know no easy answer there, but if you have some new insights please share.

    The second which I would like your opinion on is whether it would be even easier to just go with a iShares portfolio builder (if I side with MSCI) or Claymore core portfolio (if I side with RAFI), and just go to bed—don’t even bother with the couch? You get broad diversification without the headaches of rebalancing. BUT I see the MER on each is 0.6 to 0.7. They are all composed of ETFs, so is the MER on top of the ETFs in the fund? I am wondering what the total MER cost really is.

    Thanks

  28. Canadian Couch Potato February 5, 2012 at 12:24 pm #

    @Chris: Welcome to the blog, and thanks for posting. To address your first question, I am agnostic about the RAFI indexes. I think they are based on a sound methodology that increases exposure to value stocks, which have indeed outperformed over the very long term. My concern has always been the execution: the indexes have been shown to outperform in backtests, but the funds themselves have been uneven in their ability to match the performance of those indexes.

    Rgarding the ETF wraps available from Claymore and iShares, don’t worry, the stated MERs of these funds include the MERs of the underlying funds. There is no double-charging. I am not a fan of the iShares Portfolio Builders ETFs because they use an active tactical asset allocation strategy and appear to be designed for US investors, with a lot of exposure to fixed income denominated in US dollars. The Claymore CorePortfolios are better and fin fore small portfolios, but I think you can get the costs down considerably if you build a portfolio yourself, as long as you’re willing to do the maintenance.

    These posts are out of date, but may still be useful:
    http://canadiancouchpotato.com/2010/07/26/under-the-hood-claymore-coreportfolios/
    http://canadiancouchpotato.com/2010/03/16/under-the-hood-ishares-core-portfolio-builders/

  29. Sampson September 27, 2012 at 12:18 am #

    After having come back to this thread (several times actually), I have a thought/question.

    Does one need exposure to the conventional indices at all? Or could the Uber-Tuber be simplified if only the small/value indices are held? You could cut out 3 of the holdings, and get down to something more manageable.

  30. Canadian Couch Potato September 27, 2012 at 8:39 am #

    @Sampson: I think a properly diversified portfolio also needs growth and large-cap (which are the dominant factors in broad-based indexes). The small and value premiums can take a long time to show up: there have been periods of decades when they underperform. So to minimize volatility (and cost) it makes sense to have the core indexes in there as well. In fact, I would go so far as to argue that if you want to keep things simple, just use the core and forget about the small-value tilt.

Trackbacks/Pingbacks

  1. This and That: Investor mistakes, Interest rates and more… | Canadian Capitalist - June 2, 2011

    [...] Canadian Couch Potato explains how to build DFA-like portfolios with ETFs. [...]

Leave a Reply