Should you invest in DFA funds?

(MoneyWatch) 

Dimensional Fund Advisors (DFA) is a low cost passive mutual fund family. It's based on the work of esteemed finance professors Eugene Fama and Ken French, known as the Fama French Three Factor Model, that shows additional return can be achieved by weighting portfolios toward smaller and value companies.

DFA is celebrated for great performance, but fund researcher Morningstar gives the family mediocre performance ratings. And these funds can only be purchased through an advisor, meaning you get lower returns after advisor fees. Here are my findings and what they may mean to you.

Morningstar's average ratings

Morningstar's famous five star rating system defines the average mutual fund as a three star rating. Remember that the average mutual fund underperforms its appropriate index, so a three star fund is an underperformer.

Morningstar categorizes 99 percent of assets within DFA funds into one of four asset classes, and rates each DFA category as average between a dismal 2.2 star to a more respectable 3.4 star as shown here.

I took a weighted average of the four categories based on each asset class size and calculated a 3.0 average rating from Morningstar. This is in spite of having expense ratios that Morningstar defines as very low. By comparison, Vanguard funds average nearly 3.6 stars.

Now before jumping to the conclusion that DFA has only performed average, I looked at the Morningstar overall performance versus the category average over the past five years. Between 2007 and 2011, DFA funds returned a total of 16.2 percent, which bested the category average of the same asset classes by 5.9 percentage points. Granted, these returns are before the advisor fees charged, but the performance ratings should also be before any fee charged by those advisors, since DFA funds are no load.

Morningstar's response

I spoke to Russel Kinnel, Director of Mutual Fund Research at Morningstar, about the paradox of DFA outperforming its peers yet getting only average star ratings. Kinnell noted the star ratings were based on past performance over a combination of three, five, and ten year periods. He looked at the domestic stock ratings and pointed out several that, over the five year period, had below average performance versus the category average.

Kinnel agreed that, as a whole, DFA had outperformed the category average during the five year period. Kinnel stated, however, that this wasn't true on a risk adjusted basis. First, DFA holds less cash than other active funds, which increases volatility. Second, DFA funds tend to be more volatile, as they hold more smaller cap stocks than other funds in the same category. For example, The DFA US small Cap fund (DFSTX) holds stocks that average only $0.912 billion in market capitalization, while the average for the category is nearly double at $1.761 billion. The smaller companies have greater volatility, and the Morningstar performance is adjusted for this volatility.

DFA vice president Weston Wellington responded "Dimensional balanced strategies have outperformed Morningstar Model Portfolios by an appreciable margin over the last 10 years, so it does not appear that ratings are especially helpful in selecting the best performing funds." He did note that DFA tends to tilt more heavily into small cap and value, which could be consistent with Morningstar's average performance ratings, adjusted for additional volatility.

DFA vs Vanguard
If Everyone Indexed

My take

The extra return from DFA funds was never touted as a free lunch -- only as compensation for taking on more risk.  Morningstar adjusted for this risk and found DFA returns only to be average.  I'm a bit surprised, however, that they didn't beat the category average, even adjusted for risk.  That's because their lower than average expense ratios alone would have predicted above average risk-adjusted returns. 

On Thursday, I'll examine whether advisors timed DFA  funds poorly and lowered returns for investors.

[JerryNA100]

investor4life stated that "The equal weighted average of all 10 has DFA outperforming Vanguard by 0.9% per year." He further speculated on another item, lack of access to some types of international funds in Vanguard, but since that was clearly a guess and not quantitative, I will not use that 'guestimate'. Given that I can buy Vanguard directly and could only buy DFA through a financial advisor, I would be best served to hire a fee-only advisor since an advisor who charges a percentage of assets would absorb most if not all of the difference. Even most fee-only advisors who sell DFA funds take a large amount (expressing that flat fee as a percentage of assets) unless one has at least $1M to $5M. This is similar to a study I read that showed most fund managers who out-performed the market tended to charge high enough expenses that the investor's net gain over the average market was close to zero. Even with DFA's slight advantage, I would be paying someone else to take more risks with my money, and the advisor would always win even if I lost. While I very much like DFA's take on the Fama French model, I'm pretty sure that I do not want to do this. I know that you wanted to be fair, Allan, and compare DFA to Vanguard returns as clearly as possible, but my real-world gains have to take all costs into account. This is, of course, theoretical on my part because I do not have the minimum suggested amount of investable assets of $1M to make a switch to DFA worth the time of an advisor, never mind worth my while.
regards,
Jerry

[Allan_Roth]

JerryNA

I agree. As an advisor, I can buy DFA funds without charging me the advisro fees. While I do own some DFA funds, I have at least 20x more invested in Vanguard funds, which shows where I come out on the argument.

[dolomitibettega]

Allan;

Indeed the DFA performance is not a free lunch, but should not be taken in a vacuum. You have to look at the whole portfolio, and what makes DFA useful is if you have a slightly lower equity allocation, but keep your bonds as safe as safe can be such as treasury, FDIC insured CD, savings bond, or low yield/pre-refunded and/or AA/AAA municipals.

With this strategy, often used by Larry Swedroe, you have the same expected return, but lower the "fat tails". In other words, using this strategy, instead of an investor that would normally, select a 60/40 equity/fixed portfolio, they have the same expected return by using a 40/60 split, but with less risk. The drawback is that if you have a prolonged bull market, say that of the 80's and 90's, then you will have less overall gain. Still, this strategy might be worth considering because if equity does well, you will still realize the gain from small stocks. If equity does not do well, you have less to worry about because the fixed is so safe.

Personally I take a mixed view. I do Vanguard and I have a fairly high equity allocation because I am young and in a high tax situation, but I do keep the fixed as safe as possible, also because my only real choice are municipals due to tax considerations, whose yields are still relatively attractive if you buy them in small lots, even at the upper end of the "safety" curve.

All the best

[Allan_Roth]

dolomitibettega,

I agree you need to look at the whole portfolio. I do think rebalancing is key and and keeping safe money prohibits rebalancing. Much of the return of the safe portfolio of low equity (weighted toward small and value) has come from a soaring bond market. That can't continue unless rates turn negative (on a nominal basis).

I also have a mixed view as I own Vanguard and DFA funds.

[dolomitibettega]

Allan;

Thanks for the reply.

I agree that the returns for any portfolio have been skewed by the generous returns of both stock and bond markets.

While you are mostly right, the fact of the matter is that fixed income of any kind starts correlating with equity past a year of duration, even if using the highest rated. Keeping it all "super safe" does not prohibit rebalancing, in fact it allows you liquidity when you need it most through FDIC insured products (preferrably with low or no early withdrawal penalty), savings bond with same said put option, treasuries or equivalent, or just plain old safe cash. Still, at least historically, you can make out a bit better by using the commodities to extend. One of the things that can be done to offset the fixed income risk is take 5% of the allocation and invest in something like PCRIX which is a commodity futures fund. If you keep a cash like fixed, the real return is not very much. By extending durations, even considering the last 30 years, you only increase real return by 40 basis points. But depending on the portfolio it can be worth it, IE maybe 5 year duration when using commodities, and more if heavily tilted to equity (say 70-80% or more) because in that case the volatility is dominated by the equity part and the investor will then "take home" the increased duration premium. But this doesn't work too well with heavily fixed portfolios where durations should still remain 5y or less because the purpose of such portfolios is to reduce risk, and increasing duration in fixed is dangerous. FDIC insured CD's with put options make this a moot point, but unfortunately expose you to heavy taxation if you live in a high tax state and have high income, the kind you need to start getting taxable assets and maybe needing an advisor like yourself. Add all the taxes and you get close to a 50% ding, maybe more in some states. If you were lucky to buy a noncallable long term CD years ago or even recently like that "Ally" deal back in Nov 2010, maybe it still looks good when compared to current municipal yields. But the tax exposure is no good.

The total portfolio is important. Even at times when the expected return of an investment that is otherwise uncorrelated with the others is negative (like fixed now), you can still improve the performance of the portfolio by adding it and rebalancing.

The only thing that really prevents rebalancing is the odd-lot municipals, even if you keep quality very high. In this case, the higher yields are a liquidity premium. But if you use your retirement accounts for rebalancing w/ treasuries, and keep at least some FDIC stuff, savings bond and cash on hand, that should not be a problem and the municipals can form the core of the fixed.

I think that Morningstar is a joke, I am sad you even try to defend them. I think the reasons why are obvious, they do way more harm than good with their star ratings which should be ignored. The only data useful from morningstar is bid/ask spreads for ETF's, expense ratio, tax consequences/distributions or portfolio composition but most of this in 2012 can be gleaned from other sources in a more detailed fashion. But I also think equity indexed annuities, basically annuities of most kinds except the fixed ones bought by those well into retirement to insure against living too long. This is because all those term and equity indexed "products" are designed to reduce the cost of capital, and whatever "Attractive" protections they might offer against market downturns can be bought more cheaply, usually by employing the small value tilt/safe fixed strategy.

[investor4life]

Also, can you at least make one comment on how much of a joke Morningstar ratings are? No mention of the serious flaws in the Morningstar databases with funds changing categories and brining their records with them, Morningstar ignores the serious impact of survivorship bias and reports only the results of the survivors, etc.

[investor4life]

Alan, my point being, if DFA funds are below average, how bad does that make Vanguard? They underperform DFA in every category.

Further, as investors in PORTFOLIOS, we shouldn't really concern ourselves with individual fund risks, but how they contribute to the overall portfolio. And because small cap and value factors globally have very low correlations, a diversified portfolio of small cap and value stocks globally will not be as risky as the simple average of the individual fund risks.

Finally, before trying to draw conclusions about the timing ability of advisors, you should have just checked with Weston and DFA. They have reems of data showing how consistent their fund flows are in good times and bad thanks to the consistency of advisors and their well educated clients.

They had net positive inflows to their equity funds in Q3 and Q4 of 2008! At the same time, outflows from their bond funds (that had positive returns and didn't get caught up in the subprime stuff) had outflows that matched the equity inflows, clearly illustrating DFA investors continued to rebalance into the teeth of the 2008 bear market, a feat almost no other class of investors pulled off.

[Allan_Roth]

investor4life,

I think you make several errors in the above two posts:

1) You state small cap and value factors globally have very low correlations but don't say to what or over what period. I saw many advisors taking correlations from 2000-2002 to overall US stocks putting their clients into small and value in 2007 telling their clients they were bullet proof. Don't extrapolate on a sample size of one.

2) You noted I should have checked with Weston Wellington. Did you read the article? Don't you think I also would have checked with him on the dollar weighted piece tomorrow?

3) You make personal attacks such as calling Morningstar a "joke." Why would survivorship bias or fund style drift have more of an impact on DFA than any other fund family?

Indeed, I still don't think you read the piece. You sound far more like an insurance agent making a personal attack on a piece I wrote about equity indexed annuities than someone trying to get at an important issue.

I think DFA is an outstanding fund family but I think your comments do them a disservice.

[investor4life]

Alan,

I occasionally read your blog. This isn't one of your better efforts. First of all, investors don't really have the choice of whether to invest in DFA or not. As you probably know, DFA is available only to fee-only investment advisors. So the question is: do you need an advisor. If so, then you probably should look for one with access to DFA.

Now, as to the DFA vs. Vanguard question, lets simply look at 10yr returns through 12/30 for asset classes DFA and Vanguard have in common:

US Large 10YR
DFA enhanced LC = 3.0
Vanguard S&P 500 = 2.9

US Large Value
DFA Large Value = 4.6
Vanguard value index = 3.5

US Small
DFA Small cap 6.7
Vanguard Small index 6.7 0

US Small Value
DFA Small Value = 8.1
Vanguard Small Value Index = 6.1

Real Estate
DFA REIT = 10.1
Vanguard REIT = 10.3

Int'l Large
DFA Int'l LC = 4.9
Vanguard Developed Index = 4.7

Int'l Large Value
DFA Int'l LV = 7.6
Vanguard Int'l Value = 5.9

Int'l Small
DFA Int'l Small = 11.0
Vanguard Int'l Explorer = 8.6

Emerging Markets
DFA EM = 14.2
Vanguard EM Index = 13.3

Short Term Bonds
DFA 5YR Global = 4.5
Vanguard ST Bond Index = 4.1

What we see is that in 9 of 10 asset classes DFA has higher returns, with REITs being the one asset class where DFA has trailed by 0.2%. The equal weighted average of all 10 has DFA outperforming Vanguard by 0.9% per year.

And this ignores the fact that Vanguard doesn't offer International Small Value funds, Emerging Market Value funds, or Emerging Market Small Cap funds (except inside of the FTSE World ex. US Small Cap Index that is very new), whose returns over the last 10 years have been 11.9%, 17.7%, and 17.1%, all of which have higher returns than any Vanguard broad based index fund. Lack of access to these asset classes through Vanguard, that often make up as much as 20% of a diversified equity portfolio, probably cost investors another 1%.

So it looks as though the "cost" of having only Vanguard funds at your disposal is about 2% per year if the last decade is any guide.

[Allan_Roth]

investor4life

I think you missed the key point here. If you re-read the posting, you will see that I noted DFA outperformed their peer group in return. Morningstar's answer to this is that they did so with more volatility, which is why they were only average on a risk adjusted basis.

If you didn't like my posting today, you will really hate Thursdays which examines whether DFA advisors timed the funds poorly.

[140limited]

This article makes me question whether "risk-adjusted performance" as applied to individual stock mutual funds is a valid measurement criteria. It depends on the use of the fund, of course, in connection with an overall portfolio.

Risk-adjusted performance seems more appropriate to apply to an entire portfolio, rather than to particular funds. Especially if such higher-volatility funds are utilized to form a portfolio which has less volatility. This could be done, for example, by including Dimensional's funds for their higher probability of long-term (15 years or greater) out-performance in the portfolio's allocation to equities(due to the value and small cap effects, and the ability of these funds to capture such effect). Then, the investment adviser could lower the overall allocation to equities in the individual client's investment portfolio by 10% to 15%. This would likely achieve a similar long-term return, but with far less PORTFOLIO-LEVEL volatility than a portfolio which has a "balanced" equities portfolio - especially during most stock market downturns.

I would note that Modigliani risk-adjusted performance (M2 or RAP) is a measure of the risk-adjusted returns of investment portfolios. It measures the returns of the portfolio, adjusted for the deviation of the portfolio (typically referred to as the risk), relative to that of some benchmark (e.g., the market).

I would also question the implied conclusion that holding cash in a mutual fund portfolio is a bad thing. Holding cash represents an opportunity cost to investors. I always perceived large cash holdings in stock mutual funds as a negative for individual investors, who should be "fully invested" in my view in equities, fixed income investments, or other asset classes. Settling for funds which often hold 6% to 12% (or more) in money market funds, and settling for the drag on returns resulting from such holdings, seems problematic - or at the minimum a factor which must be taken into account when forming an asset allocation for the overall portfolio. (If this was done, then a greater allocation to equity mutual funds would be undertaken - if those equity funds had high cash holdings. Again, this means the individual funds may have lower risk-adjusted returns, but to compensate for the cash holdings within the funds the overall portfolio could have greater exposure to equities - and hence the same - or higher - risk-adjusted returns at the portfolio level.)

There are some very low-cost (web-based) advisors who provide access to DFA funds. Some charge a very low percentage fee, or even a flat annual fee. Other advisors charge higher fees, but usually throw in a lot of additional services (financial planning, wealth management) for such higher fees and personal service.

I look forward to Thursday's column. However, and regardless of whether the result you ascertain is a positive or negative for Dimensional's funds, I would caution that taking a "snapshot" of returns of funds at any one point of time often comes up with incorrect results. If funds have been around for 20 or more years, why not measure them over rolling 10-year time periods, or rolling 15-year time periods, or over the entire time period? Since funds and indices rarely have the same exposures to book-to-market and market cap-driven factors, "starting points" and "ending points" over any 5-year or 10-year period can lead to poor analyses. For example, did a significant overvaluation or undervaluation of large cap stocks vis-a-vis small cap stocks, or value stocks vis-a-vis growth stocks - exist at either the beginning or the end of the period chosen to be viewed? This could really skew results over a discrete time period - even 10 years.

Since some sector indices have been around for 30+ years (i.e., Russell), comparing long-term returns of a fund (which has been around a long time) relative to indices may be a better indication of the fund's performance.

Lastly, I would note that even Morningstar has admitted that fees and costs are a more significant factor in predicting future returns than its own ratings. Since Dimensional's funds - especially its micro-cap and "core equity" funds - either have very low internal transaction costs, or add to returns through block purchases of (small-cap, mid-cap) stocks at discounts, or add to returns through securities lending practices (possible for a passive and diversified fund), these positive cost/fee attributes should show up in the long-term performance data.

[Allan_Roth]

140limited,

I think all of your points are valid and well thought out. Ultimately, it's the volatility of the entire portfolio that matters, cash is a drag on performance in the long term, and costs really do matter.