The Tuber Gets Tested

Monday’s post comparing one of my model ETF portfolios to one created by Dimensional Fund Advisors attracted a lot of interest from readers—and from some DFA advisors, too.

Shortly after that post went live, I was contacted by a DFA advisor whom I’ll call John (his compliance department keeps him on a short leash). John ran some numbers to compare how the original Über-Tuber might have performed alongside a DFA portfolio over the last 15 years.

Of course, none of the ETFs in my suggested portfolio were around in 1996. So John used index data to estimate the returns of the Über-Tuber, subtracting the MERs of the funds to account for costs. For the DFA portfolio, he subtracted a 1% advisory fee from the fund returns. Here’s the breakdown he used:

Index Über-Tuber DFA
Canadian One-Month T-Bills 20% 20%
DEX Universe Bond 20% 20%
S&P/TSX Composite 4.8%
Barra Canadian Value 7.2%
DFA Canadian Core 20%
Barra Canadian Small 8%
DFA U.S. Vector 18%
Russell 2000 Value 8%
Russell 1000 Value 10%
DFA International Vector 14%
MSCI World (ex. U.S.) 4%
MSCI EAFE Value 10%
S&P Global REIT 4% 4%
MSCI Emerging Markets 4% 4%
100% 100%

Here are the results from July 1996 through April 2011:

Annualized Return Total Return Growth
of $1
Standard Deviation
Über-Tuber 6.36% 149.53% $2.50 7.82%
DFA Portfolio 6.42% 151.57% $2.52 8.14%

And here’s how the path of those returns would have looked during the period:

As you’ll see, the returns of the two portfolios over the 15 years were almost identical. The Über-Tuber trailed the DFA portfolio by just half a dozen basis points annually, and it accomplished that result with a lower standard deviation—which means lower volatility.

What’s it all mean?

So what conclusions can we draw from these data? First, it seems the ETF portfolio would have done a respectable job of mirroring DFA’s small-cap and value strategies. However, the simulation assumed that trading costs were zero, that the investor rebalanced religiously, and that the ETFs all tracked their indexes very closely. It’s safe to say that a real-world investor would have been hard-pressed to accomplish all of that.

What’s more, assuming our ETF investor did manage to do everything right for 15 years, the DFA portfolio still delivered slightly better results, even after deducting a 1% fee for the advisor. An experienced advisor would certainly have added value if he or she handled all of the heavy lifting during this turbulent period, which included three roaring bull markets (1996–2000, 2003–06 and 2009–10) and two swift kicks in the groin (2001–02 and 2008–09).

There’s one more interesting thing to add. John also ran the same simulation with the trusty old Global Couch Potato—which includes just three or four funds—and the results fell right in the middle: an annualized return of 6.40%. There was a bit of divergence between the three portfolios during some periods, but after 15 years it was essentially a wash, no matter which one you used.

Maybe it does pay to keep things simple after all: diversify widely, keep your costs low, rebalance and stick to your plan. If you can do that, you’re at least 90% of the way to the perfect portfolio.

16 Responses to The Tuber Gets Tested

  1. Eric June 1, 2011 at 1:44 pm #

    Great post !

    If the global couch potato does as well as the DFA portfolio, it means that as Canadians we don’t benefit from value & small-cap exposure ……..

  2. Canadian Couch Potato June 1, 2011 at 1:53 pm #

    @Eric: I wouldn’t make any sweeping conclusions from these data. The value and small premiums don’t show up over every period, even every 15-year period.

  3. Canadian Couch Potato June 1, 2011 at 3:59 pm #

    Got an email from a DFA advisor who had some comments on this post:

    – “If anyone is reviewing those two portfolios versus a plain-vanilla Couch Potato, then the length of time used is not long enough. DFA states that in order to get any effect from size and value premium, one may need 20 to 30 years of data — anything less than that is just noise.”

    – “No mention is made of hedging the Canadian dollar, which had a great impact form 1996-2011. Any good portfolio should have at least 25% to 50% hedged.” [Note that all the portfolios above are unhedged. In general, DFA does not recommend currency hedging in equities, although in Canada the US and international funds are available with that option.]

    – “The report reminds me of two diet programs that a person could follow:
    the results would just depend on what day the person gets weighed, and not on which program was the best for the client re: a lifestyle change. But you are right: both mentioned are much better than 90% of active portfolios out there.”

    – “Note that an advisor’s fee, if embedded in a Class A fund, is always a tax deductible expense, but if done as a Class F version it may not be. If you’re working with an advisor who uses Class F, make sure they have an opinion from their firm stating that they have clearance from CRA re: their fees and tax deductibility. At a 40% tax rate, that net cost of a 1% fee is only 0.6%.”

  4. Chris June 1, 2011 at 5:33 pm #

    The DFA advisor’s comment that: “…the length of time used is not long enough. DFA states that in order to get any effect from size and value premium, one may need 20 to 30 years of data — anything less than that is just noise” is statistically untrue.

    With 15 years of data, there is *much* more than enough data to overcome the statistical noise threshold. In fact, given the standard deviation above and basic statistics, you get a 95% confidence interval that the DFA portfolio will not beat the Tuber portfolio by more than 0.1% annualized over the long term.

    20 years of data would just narrow the range within the 95% confidence interval to 0.09%; 30 years of data would not narrow it any further (to two significant digits of precision).

    There is enough data to predict that the two portfolios will perform in an essentially identical manner over the long term.

  5. Andrew Hallam June 1, 2011 at 6:39 pm #

    I think a simple couch potato portfolio would be fine for investors. Simplicity often makes it easier to keep our emotions in check as well….and emotions are an investor’s biggest enemy, or friend. The meticulously chosen moving parts of a back-tested portfolio are probably the noise we need to be careful of. After all, it creates a psycholigical quest for Alpha that will lead far more people astray than anything else: based on the temptation to tinker and always look for that “next best thing”—whether it happens to be a complicated passive strategy or not.

  6. Superior John June 1, 2011 at 10:07 pm #

    Dan whether its the Uber Tuber, Super Duber etc. etc., articles like this and the like on the couch potato,( including the insightful thoughts/ comments like Chris, the DFA advisor, and others offer) is worth way more than 1% of a portfolio loss or gain. No one wants to loose 1% to any fund, however, the broad scope of your articles and the great scope of input by your readers would cost at least 1% on an annualized basis with so called expert financial advisors.
    Dan keep up the great articles, and train of thought! I sure appreciate those out there that have great business insight, feeding readers like myself with well researched comments in a comprehendable language. Hello readers keep up this great forum. I am an appreciative reader.

  7. Canadian Couch Potato June 1, 2011 at 10:40 pm #

    @John: Thanks so much for the kind words, and for adding your own comments. I can always rely on my readers to contribute a lot of wisdom and experience to the discussion. Cheers!

  8. John the DFA advisor June 2, 2011 at 10:37 am #

    @Chris: I agree with you: it is not a battle for ALPHA, it is a battle for BETA. The point of all of this is that the Uber-Tuber should mirror the BETA of the risk factors: that is what it is designed to do. So Dan has done an excellent job showing readers that there is no BLACK BOX to these returns: they are BETA-driven.

    I was very happy that the Uber performed the same. It shows that Fama and French are right and that it is the BETA of the risk factor that is creating the return. No special manager is needed.

    Whether it is DFA or the Uber Tuber, have at it. Just don’t think it is easy, or simple, or effortless to get it: it is not. Give it a try, just don’t believe the HYPE of anyone (be it a DFA advisor etc.). Do what you think is best for your family. If you don’t have the confidence to do this yourself, you get the same result and a coach for 1% with DFA. That’s it, that’s all. No magic.

    This is all about pulling back the curtain on the Wizard of Oz, it is not about replacing one wizard with another.

  9. Raman June 2, 2011 at 2:24 pm #

    Thanks for your humility, John. It’s refreshing to see a fund advisor speak in such simple terms for once 🙂

  10. My Own Advisor June 2, 2011 at 4:43 pm #

    Nice post Dan!

    I really enjoy these comparative posts. I’m with Andrew, I think? this proves simple is better or at least, a simple, well-selected indexed portfolio is almost as good as anything else out there. I’m not only learning this from reading, writing, reviewing, analyzing and interacting with other DIY investors, I’m seeing the results through my own experience. Incidentally, I like what I’m seeing…

    Keep up the great work!

  11. Pinery June 2, 2011 at 5:37 pm #

    Thank you all posters for the wealth of knowledge and information we are getting on this blog. More proof that simplicity is a virtue. I’ve always battled with managing my portfolio of over 10 etf’s.

  12. Jas January 9, 2013 at 7:38 pm #

    It’s interesting to see that Mawer’s balanced mutual funds, which has almost the same asset allocation (40% bonds, 60% equities 1/3 US 1/3 Can 1/3 Int), including a small cap tilt, has approximately the same returns than both the Uber-Tuber and the DFA portolio during this time period (june 1996 to April 2011). As you can see below, the graph is quite similar to the one provided by the DFA advisor.

    http://www.theglobeandmail.com/globe-investor/funds-and-etfs/funds/summary/?compareBench=&FromMonth=7&FromYear=1996&ToMonth=4&ToYear=2011&id=17966&symbol=MAW104&style=globe_bal&profile_type=ROB

    http://www.mawer.com/mutual-funds/fund-profiles/mawer-balanced-fund/

  13. Francis June 2, 2013 at 3:51 pm #

    By reading that post, I was really interested to look forward to get my money with DFA but I see on your web page that the fee are 1,5% not 1%, it a huge difference, I wonder if they just raise their fee since then?

  14. Canadian Couch Potato June 2, 2013 at 4:28 pm #

    @Francis: The 1.5% is the maximum at PWL and typically applies only to small accounts. Most clients qualify for the 1% fee level.

  15. Jas June 12, 2014 at 9:47 pm #

    @CCP:
    I was looking at your 20-year model portfolio performance (1994-2013):
    http://canadiancouchpotato.com/wp-content/uploads/2014/01/CCP_Model_Portfolio_Performance_1994-2013.pdf

    Are you still convinced that small cap and value tilt is worth the trouble and/or added expenses?

  16. Canadian Couch Potato June 12, 2014 at 11:24 pm #

    @Jas: It really depends on the individual circumstances. I would say if you have a very large portfolio (at least mid six figures) it is worth it, though only if you can manage it efficiently.

    At PWL Capital we use a combination of DFA funds and ETFs to get that tilt at quite low cost: a typical portfolio MER is about 0.35% or less. But these portfolios would be difficult for a DIY investor to manage, and of course they could not access the DFA funds anyway. So for a DIYer using a complex ETF portfolio it would be a challenge: you would have to be a pretty experienced investor.

Leave a Reply