Last Thursday, the Macdonald-Laurier Institute released a report warning Canadians that several provinces have a high probability of defaulting on their debt payments over the next 20 years (although they acknowledged that the risk of default for all provinces over the next 5 to 10 years was essentially zero).
Although this Greece-like crisis is only one possible outcome over many, it is still recommended that investors take the time to understand the underlying risks of their fixed income investments. We can start by popping the hood on the DFA Investment Grade Fixed Income Fund (DFA449).
Long Term Provincial Bonds
The fund is underweight long term provincial bonds and overweight long term federal bonds (relative to the iShares DEX Universe Bond Index Fund (XBB)). For investors who are concerned with the long term sustainability of our provincial finances, this fund may be more suitable than a broad market ETF.
Short Term and Long Term Corporate Bonds
Adding short term corporate bonds to a portfolio has historically resulted in higher returns and higher risk-adjusted returns, relative to an all short term federal bond portfolio. Adding long term corporate bonds to a portfolio has actually resulted in lower returns and lower risk-adjusted returns, relative to an all long term federal bond portfolio. DFA has a significant overweight to short term corporate bonds and an underweight to long term corporate bonds – I doubt this allocation is merely by accident.
The DFA fund uses the variable credit approach to fixed income investing. When credit spreads (the difference between corporate bond yields and government bond yields) are wide (narrow), the fund will overweight (underweight) corporate bonds, relative to the DEX Universe Bond Index. When intra-corporate spreads (the difference between lower quality A/BBB bonds and higher quality AAA/AA bonds) are wide (narrow), DFA will also overweight (underweight) these credits within their corporate bond allocation.
The DFA fund includes a number of issuer and credit quality constraints that control the risk of the variable credit strategy.
Credit Quality Constraints:
For example, if the DEX Universe Bond Index has 22% exposure to A-rated issuers and 8% exposure to BBB-rated issuers, the DFA fund can have a maximum of:
As with all fixed income investments, it helps to understand the underlying risks inherent in any strategy. As longer term provincial, municipal and corporate bonds are riskier than longer term federal bonds, I prefer to favour products that have a low allocation to these issuers.
Larry Swedroe, research director at Buckingham Asset Management in St. Louis, has suggested that including emerging market bonds (EMBs) within the fixed income allocation of a portfolio is like tucking sticks of TNT inside it. Although Swedroe is exaggerating to some extent, let’s take a look at a couple of portfolios to understand why his comment does in fact make sense.
This is a typical balanced portfolio that does not include an allocation to EMBs. From 1994 to 2011, it had an annualized return of 6.94%, a standard deviation of 8.06%, and a Sharpe ratio of 0.45.
EMBs Included in the Fixed Income Allocation
By underweighting investment grade fixed income by 5%, and allocating the same amount to EMBs, the annualized return increased to 7.02%. The standard deviation of the portfolio also increased to 8.30%, resulting in lower risk-adjusted returns of 0.44.
EMBs Included in the Equity Allocation
If we replaced emerging market equities (EMEs) with EMBs, the annualized return would have increased to 7.07%. The standard deviation of the portfolio would have decreased to 7.61%, while the risk-adjusted returns would have increased to 0.49. It is important to note that EMBs outperformed emerging market equities by about 4.57% during the measurement period – this type of outperformance would not be expected going forward.
Asset Location – for those considering including emerging market bonds in their portfolio, they should be held in your tax-deferred account (in order to avoid the negative after-tax effects of holding premium bonds in your taxable accounts)
Foreign Withholding Taxes in Tax-Deferred Accounts – By holding a Canadian domiciled ETF, which holds a U.S. domiciled ETF (such as the iShares J.P. Morgan USD Emerging Markets Bond Index Fund (CAD-Hedged) (XEB)) in your tax-deferred account, foreign withholding taxes of 15% will be lost forever. With a coupon of 6.58%, approximately 0.99% (6.58% × 15%) of your return will be lost to foreign withholding taxes.
Note: If you invest in a Canadian domiciled ETF that holds a basket of emerging market bonds directly, you can avoid these foreign withholding taxes. An example would be the BMO Emerging Markets Bond Hedged to CAD Index ETF (ZEF). You could also hold an un-hedged U.S. domiciled ETF in your tax-deferred account, such as the iShares J.P. Morgan USD Emerging Markets Bond Fund (EMB), in order to avoid foreign withholding taxes.
Tracking Error – the actual performance of an emerging market bond ETF can lag the performance of the underlying index by more than the expense ratio (this may be caused by other factors, such as additional trading expenses due to a lack of liquidity). Over the past 3 years, the iShares J.P. Morgan USD Emerging Markets Bond Fund (EMB) has lagged its underlying index by an additional 31 basis points.
Future Expected Returns – our best estimate of the expected return on an emerging market bond ETF is its average yield-to-maturity (YTM) minus estimated foreign withholding taxes minus the estimated management expense ratio (MER). For example, the iShares J.P. Morgan USD Emerging Markets Bond Index Fund (CAD-Hedged) (XEB) currently has an average yield-to-maturity of 3.95%, estimated foreign withholding taxes of about 0.99%, and estimated MER of 0.77%, resulting in an expected return of 2.19% (before-tax). An investor could instead purchase a less risky 5-year GIC and earn about 2.50% (before-tax). Whether or not an investor feels that the risk of investing in emerging market bonds has been adequately compensated for with this low yield is at their discretion.
As it currently stands, I would not recommend emerging market bonds for most investors. If you decide to allocate a portion of your portfolio to this asset class, ensure that you include it in the equity portion (preferably the emerging markets portion) of your portfolio (not the fixed income). Also ensure that you hold it in a tax-deferred account and use the most tax-efficient ETF.
Investors who have worked with me know that I prefer to buy guaranteed investment certificates (GICs) instead of bonds in taxable accounts. The reasoning is fairly straightforward, but not widely understood. Most individual bonds that currently trade in the marketplace are sold at a premium to their par value. What this means is that you initially pay more than $100 to purchase a bond, but only receive $100 at maturity (resulting in a capital loss on your investment). If our tax system allowed Canadians to offset their regular income with their capital losses, this would be a non-issue. As it currently stands, this is a significant issue that can negatively impact the portfolios of many Canadians who hold premium bonds (or bond ETFs and mutual funds) in their taxable accounts.
To better illustrate this concept, consider a taxable investor who has the option of purchasing a Government of Canada (GOC) bond or a GIC, each yielding 1.50% and maturing in three years. As the annual interest of the GOC bond is higher than the yield-to-maturity (3.22% versus 1.50%), we know that this bond is trading at a premium to par. As we can see, the initial cost of the GOC bond is $105,000 and it matures at $100,000 (resulting in a capital loss of $5,000).
The after-tax absolute dollar return of the GOC bond is only $177, relative to the GIC return of $2,532. If the premium bond investor has the ability to offset capital gains with his $5,000 capital loss, his after-tax return would increase to $1,337 (still well below the after-tax GIC return of $2,532).
As a general rule of thumb, investors should avoid purchasing premium bonds, bond ETFs, or bond mutual funds in their taxable accounts. If you are unsure if the average underlying bonds of a pooled product are trading at a premium to par, just visit the company website and look to see if the Weighted Average Coupon (%) is larger than the Weighted Average Yield to Maturity (%) – if it is, you have yourself a basket of premium bonds.
Source: BlackRock Canada
With arguably the most diversified Canadian stock portfolio of any fund in its category (nearly 500 holdings), the DFA Canadian Vector Equity Fund is an ideal choice for those investors seeking the higher expected returns of value and small cap companies. The price is also reasonable, with a management expense ratio (MER) of just 0.45%.
I Gotta Have More Value
Whenever I hear the term “value company”, I picture the scene from Rudolph where he stumbles across The Island of Misfit Toys. In the classic animated TV special, a group of unwanted and defective toys go to live there until the island’s ruler, King Moonracer, can find a suitable home for them.
In the media, RIM’s BlackBerry appears to be the misfit toy of the day, tossed aside for Apple’s sexier iPhone. Value companies are believed to have a financial distress premium built into their price, resulting in higher expected returns.
DFA accepts these out-of-favour companies in the same way that King Moonracer accepted the mutant toys. They currently hold over twice as much RIM than their more discriminating predecessor, the S&P/TSX Composite Index (“The Index”).
I Wish I Were Small
Who didn’t drink RC Cola as a kid?! My cost-conscious father was always stocking the fridge full of cheap alternatives to “The Real Thing”. RC Cola is a division of Cott Corporation, a smaller and relatively unknown beverage company with locations throughout the world. Smaller companies are generally considered riskier than bigger, more established companies (and we should therefore expect a higher return for investing in them). DFA’s fund has over three times the allocation to Cott Corporation relative to the index.
Higher Expected Returns and Higher Risk
If we assume that value and small cap companies will each return 2% more than their growth and large cap counterparts, how much would the DFA fund be expected to outperform the index by (before fees)?
In order to venture a guess, we could first run a 3-factor regression on the historical monthly returns of the Dimensional Canadian Vector Equity Index (this index performance is a back-test of the vector strategy and is available in the Dimensional Returns 2.0 software).
The model is a fairly good fit, explaining 98% of the fund’s monthly returns (indicated by an adjusted R2 of 0.98). The beta (β) of the fund is 1.00, which indicates market risk similar to the index (this comes as no surprise, as broadly diversified equity funds normally have a beta of close to 1.00). The large and significant positive small cap (s) and value (h) coefficients indicate that this fund has substantial small cap and value tilts (0.39 and 0.24 respectively). In comparison, the index would be expected to have small and value coefficients equal to 0.
SMB = 2%
HML = 2%
s = 0.39
h = 0.24
Expected Outperformance = (s × SMB) + (h × HML)
= (0.39 × 2%) + (0.24 × 2%)
In other words, if the Index returned 7% over the long term, and the small cap and value premiums were each 2%, the DFA Canadian Vector Equity Fund would be expected to return 8.26% (assuming DFA was successful at capturing the premiums). Keep in mind that if the small cap and value premiums turn out to be negative in the future, DFA’s vector strategy will also be expected to underperform the index.
Although tilting towards value and small cap companies is not for everyone, the strategy has historically resulted in higher returns (and higher risk), relative to a simple Couch Potato strategy. For investors who are obsessed with maintaining a tight tracking error relative to the index, this fund is not for you.
* Methodology for Construction of Canadian Regression Factors:
Mkt-TBill= MSCI Canada IMI Index – DEX 30 Day T-Bill Index
HML = ½(MSCI Canada Small Cap Value Index + MSCI Canada Value Index) – ½(MSCI Canada Small Cap Growth Index + MSCI Canada Growth Index)
SMB = ⅓(MSCI Canada Small Cap Value Index + MSCI Canada Small Cap Index + MSCI Canada Small Cap Growth Index) - ⅓(MSCI Canada Value Index + MSCI Canada Index + MSCI Canada Growth Index)