Another Promise of a Free Lunch

Earlier this month, Mackenzie Investments produced this advertisement for the Mackenzie Sentinel Corporate Bond Fund. The banner headline reads “Think high yield means high risk?” The rest of the ad reads, “Think again. High-yield corporate bonds have produced equity-like returns with less risk.”

Two graphs then compare the performance of the fund with the S&P/TSX Composite Index from November 2000 (the fund’s inception date) through October 31 of this year. During that period, the Mackenzie fund delivered annualized returns of 5.7% with a standard deviation of just 5.8%. The Canadian stock market, by comparison, delivered just 4.5% with much more volatility: a standard deviation of 15.5%.

Those are compelling numbers: high returns with lower volatility is what every investor wants. But it’s always important to scrutinize comparisons like this. There’s nothing incorrect in the data per se, but one needs to ask whether the fund really did deliver superior risk-adjusted returns, and whether it is likely to do so in the future.

To begin with, the start and end dates examined here are crucial. If you look at the table under the two graphs, you’ll see that the S&P/TSX Composite returned 8.5% annualized during the 10 years beginning in October 2001, outperforming the Mackenzie fund by 2.8% a year. So by moving the start date by just one year, you get a completely different result. It may be true that “high-yield corporate bonds have delivered equity-like returns,” but only during specific periods when equities dramatically underperformed their historical averages.

Were you rewarded for taking more risk?

The ad also argues that “it takes rigorous company and credit analysis to find bonds with long-term growth and income potential. When you do, it adds up to superior returns and a smoother market ride.” In the case of the Mackenzie fund, it also adds up to an MER of 1.7% a year.

There are only two ways that a bond manager can deliver superior returns than a broad-market index. The first is by adjusting maturities—that is, by selecting a portfolio of bonds with shorter or longer terms than the benchmark. The second is by varying credit quality: by selecting bonds of lower quality, and therefore higher yields. With that in mind, did the Mackenzie Sentinel Corporate Bond Fund add value?

Justin Bender, a CFA and portfolio manager at PWL Capital in Toronto, supplied the following analysis. He compared the Mackenzie fund’s performance to a blended benchmark consisting of two indexes of federal government bonds: 60% DEX Mid Term Federal Bond and 40% DEX Short Term Federal Bond. This 60-40 mix creates a benchmark with an average term to maturity of about 5.7 years, which is identical to that of the Mackenzie fund’s portfolio. In this way, we’ve taken maturity out of the equation so we can focus only on credit quality.

Since high-yield bonds have far more credit risk than government bonds of the same maturity, investors should naturally expect higher returns. Would they have achieved them with the Mackenzie fund? Here are the numbers:

1 yr. 3 yr. 5 yr. 10 yr. Since 11/03/2000
Mackenzie Sentinel Corp. Bond 3.4% 11.6% 4.3% 5.7% 5.7%
Benchmark 5.6% 6.2% 6.0% 5.6% 6.4%

Sources:  PC-Bond, BMO Financial Group, Mackenzie Investments, Dimensional Returns 2.0. Returns as of October 31, 2011.

As you can see, since its inception, the Mackenzie fund has not even outperformed government bonds of similar maturities. Over the last 10 years, it managed a mere 10 basis points of outperformance. Over shorter periods, the results are mixed. This hardly a compelling argument for the fund’s “rigorous company and credit analysis.”

And what about volatility? The Mackenzie fund’s standard deviation of 5.8% does indeed look low when compared with the S&P/TSX Composite. Yet our 60-40 benchmark of government bonds not only achieved higher returns since November 2000, but it did so with a much lower standard deviation of just 3.5%, according to Bender’s analysis.

“Finding equity-like returns with less volatility is something many market weary investors are looking for,” the Mackenzie ad says. No argument there. But the data suggest they need to keep looking.

22 Responses to Another Promise of a Free Lunch

  1. Chad Tennant December 1, 2011 at 9:52 am #

    This is worthy of a ’60 Minutes’ segment. I thought only consumer packaged goods (CPG) companies played the time period manipulation game as to highlight favourable results. Good catch on the time horizon and fund comparisons.

  2. Canadian Couch Potato December 1, 2011 at 10:03 am #

    @Chad: A few people have told me that I look like Andy Rooney. Actually, this is a favourite technique in the investment industry. There are rules, of course, about how firms market their performance, so they’re not saying things that are untrue. But they can be sneaky, and it’s very difficult for most investors to spot their tricks.

    One book that I highly recommend on this subject is How to Lie with Statistics, by Darrell Huff. It’s 67 years old, but still useful today. You can get it as an eBook for $6.49.

  3. Chris December 1, 2011 at 10:15 am #

    I think you’re mostly off-base with this post. There’s no point in comparing high yield bonds to standard bonds. They’re much more correlated to equities. The relevant comparison is to the corresponding equity benchmark. And in general for the last 20 years high yield bonds have outperformed large cap equities, both on an absolute basis and on a risk-adjusted basis.

    Take a look at JNK versus SPY over the last five years (including the height of the financial crisis). JNK has an annualized total return of 4.7% versus -2.2% for SPY, but JNK also has much lower beta. JNK also had a maximum drawdown of -38.6% during the financial crisis, versus -53.9% for SPY. Both of those measures of risk are very significantly in favour of high yield.

    It’s not wrong in any sense to say that junk bonds *have* produced equity-like returns with less risk.

  4. Canadian Couch Potato December 1, 2011 at 10:28 am #

    @Chris: I don’t dispute that high-yield bonds have produced equity-like returns over some periods, including the last five years (though not the last 10, as discussed above). Where are you getting your data about high-yield bonds outperforming equities over the last 20 years? Is there a high-yield bond index with useful data going back that far? I was not able to find one.

  5. Andrew F December 1, 2011 at 10:55 am #

    Your benchmark should be a high yield bond index. Then you can see whether the active manager added value over a passive strategy. Comparing one asset class to another is apples to oranges.

    I looked for a 10 year old+ relatively passive high yield fund and found the Vanguard High Yield Corporate Fund (MER 0.25%):

    (1 yr, 3yr, 5yr, 10 yr, inception 1978)
    Returns before taxes 2.86% 10.76% 5.59% 6.60% 8.66%

    By those measures, it looks like the Sentry fund at least earned its fee. In other words, they basically took any alpha they generated for themselves. Quelle surprise.

  6. Canadian Couch Potato December 1, 2011 at 11:10 am #

    @Andrew F: Comparing one asset class to another is precisely what the Mackenzie ad does. My point here was that fixed income, in general, outperformed equities over the period being considered, and you were not rewarded for taking additional credit risk. In fact, you were penalized for it, because government bonds delivered even higher returns with even less volatility.

    Using the logic of the ad, one could argue just as strongly that government bonds have delivered “equity-like returns with less risk” since late 2000. But should we expect this to continue going forward?

  7. Chris December 1, 2011 at 11:20 am #

    One source with 20+ year data is page 12 of this document:

    http://advisorshares.com/system/files/funds/public_docs/wp_peritus_case_for_high_yield_82011.pdf

    I actually have a slightly older version of that document, calculated with a different end date where high yield has a fractional advantage in absolute terms. In the linked version, with June 2011 as an end date (i.e. before the recent drop in the S&P 500), the S&P 500 has a small advantage in absolute terms but is at a significant disadvantage in risk-adjusted terms.

    In any case, the evidence is pretty clear that high yield has historically produced equity like returns with less risk.

  8. Andrew F December 1, 2011 at 11:23 am #

    Oh, I agree that the ad is misleading. I was just defending high yield debt as an asset class. I think it can be a useful component of a diversified portfolio.

  9. LateStudent December 1, 2011 at 12:16 pm #

    The S&P/TSX Composite is a price index, not a total return index, right?

    Is it fair to compare returns of *any* fund to an index that doesn’t include total returns?

    …is this a naive question??

  10. Michael Davie December 1, 2011 at 1:31 pm #

    Great post. Here are another few excellent books about how we can be misled with numbers:
    Proofiness by Charles Seife: http://www.amazon.com/Proofiness-Dark-Arts-Mathematical-Deception/dp/0670022160
    The Drunkard’s Walk by Leonard Mlodinow: http://www.amazon.com/Drunkards-Walk-Randomness-Rules-Vintage/dp/0307275175
    Fooled by Randomness by Nassim Nicholas Taleb: http://www.amazon.com/Fooled-Randomness-Hidden-Chance-Markets/dp/1400067936

  11. Dan Hallett December 1, 2011 at 6:20 pm #

    Dan B – As always, a great post. I had planned to write something about this because of the ad’s striking ‘headline’. I probably won’t – since you’ve hit the main points – but can add a couple of points.

    There is a ‘suite’ of global bond indices tracked by Bank of America Merrill Lynch – most notably their corporate and high yield indices for Canada, the US and globally. Their data starts in 1986-87 if memory serves. So contrary to one comment above, the most appropriate index is NOT and equity index but a corporate and high yield index.

    I agree that comparing corp/HY bonds to Canadas ‘mixes’ added value/detracted by the manager and the risk premium available from taking credit risk over the measurement period. But your point is taken – in the context of their headline claim – that reaching for yield via credit risk (or by taking other types of risk) will not ensure higher returns even over longer periods of time.

    LateStudent – While the index named in the add doesn’t specify, the numbers line up with the TOTAL RETURN version of the index. So they are at least using the right version of the index.

    Finally, it’s interesting to note that while Mackenzie ‘owns’ the track record, the two main portfolio managers behind this funds history have left for other oppotunities over the past couple of years. So the people who made big contributions to that longer term track record are no longer running the fund.

  12. Justin Bender December 2, 2011 at 9:43 am #

    For anyone interested, here are the oldest returns I could find for high yield bonds versus equities (September 1986 to October 2011):

    Bank of America Merrill Lynch High Yield Master II Index:
    Annualized Return = 7.31% (in CAD)
    Annualized Standard Deviation = 8.28%

    S&P 500 Index:
    Annualized Return = 7.62% (in CAD)
    Annualized Standard Deviation = 14.09%

  13. Canadian Couch Potato December 2, 2011 at 10:49 am #

    @Dan H: Many thanks for your comment. As you point out, one always has to be wary of choosing a mutual fund based on past performance. Even if those managers did have genuine skill, they may no longer be around to make the decisions.

    @Justin: Thanks for these data. I don’t have monthly returns, but the broad Canadian bond market returned 8.6% annualized from 1986 through 2010. I await the Mackenzie ad pointing out that investment-grade bonds can provide equity-like returns with less risk. 🙂

  14. Superior John December 2, 2011 at 3:42 pm #

    Another great post Dan. Excellent feedback by your audience. Keep it coming..

  15. Chris December 3, 2011 at 2:32 pm #

    @CC: To be fair, investment grade bonds are unlikely to repeat their past performance well into the next decade (it’s next to mathematically impossible, at this point). The case for high yield continuing its past performance is much stronger.

  16. Tom Bradley December 5, 2011 at 11:28 am #

    Dan (and Justin), thx for doing this. This ad makes my blood boil. I hadn’t seen the numbers before this post, but the ad itself also should be called out. Comparing two different asset classes (I know HY is highly correlated to stocks) is very deceptive. There must be a 1000 ads we could create by do cross class comparisons. Hopefully your blog gets the attention of the OSC.
    Keep well.

  17. Dale December 14, 2011 at 8:49 am #

    But unfortunately there is no etf for Canadian high yield debt. There’s US that’s currency hedged, but no Can to my knowledge. Not liquid enough according to folks at Claymore.

    Anyone know of a second best? Is there a mutual fund with reasonable fee?

  18. Canadian Couch Potato December 14, 2011 at 8:59 am #

    @Dale: The iShares DEX HYBrid Bond Index Fund (XHB) invests in Canadian high-yield bonds:
    http://ca.ishares.com/product_info/fund/overview/XHB.htm

    The cheapest mutual fund I’m aware of is the PH&N High Yield Bond Fund, with an MER of 0.87%:
    https://www.phn.com/Default.aspx?tabid=532

  19. Dale December 14, 2011 at 9:05 am #

    Thanks for that. I had investigated xhb and thought it still held a lot of US, but it shows Canadian as its top 10 holdings. And yes PH&N is a great company compared to most who sell mutual funds and advice. Thanks for both.

  20. Dale December 14, 2011 at 9:08 am #

    Actually their high yield xhy is US as well. This one is hybrid bond?

  21. Canadian Couch Potato December 14, 2011 at 9:21 am #

    @Dale: Yes, XHY is just the US-listed iShares high-yield bond ETF with currency hedging added.

    XHB holds only Canadian high-yield bonds. Don’t get distracted by the the term “hybrid.” It’s mostly a play on words (“High Yield Bond” = HYB) rather than a meaningful description. Here’s the index methodology:
    http://www.canadianbondindices.com/pdf/DEX_HYBrid_BI_Methodology.pdf

  22. Dale December 14, 2011 at 9:42 am #

    Dang. PH&N is closed. Great returns and managment.

Leave a Reply