Q: I was surprised to see a Vanguard infographic pointing out that international [non-US] bonds are the largest asset class in the world. Do you have any thoughts on why Canadians have not embraced international bonds in their portfolios? – A.M.
While stocks grab all the headlines and dominate the conversation among investors, the bond market is vastly larger. Yet while a diversified index portfolio can include 10,000 stocks from over 40 countries, chances are your bond holdings are entirely Canadian.
There are some good reasons for a strong home bias in bonds. The main one is currency risk. Exposure to foreign currencies benefits an equity portfolio by lowering volatility (at least for Canadian investors), but taking currency risk on the bond side is unwise. Because currencies are generally more volatile than bond prices, you’d be increasing your risk without raising your expected return. That’s a bad combination.
It also gets to the heart of why few Canadians have international bonds in their portfolios: there just aren’t many good products offering global bond exposure without currency risk. iShares and BMO have a number of ETFs covering US corporate and emerging markets bonds. But if you wanted to build a low-cost, diversified, currency-hedged portfolio with high-quality bonds from around the world, it was hard to do so until this summer. The wait ended in June with the launch of the Vanguard U.S. Aggregate Bond (VBU) and the Vanguard Global ex-U.S. Aggregate Bond (VBG), both of which are hedged to Canadian dollars.
Now that international bonds are available to all Canadian investors, should you rush to build a globally diversified fixed income portfolio? For most people, the answer is probably no. The benefits of holding non-Canadian stocks is enormous, as equity returns vary widely from country to country. But the diversification benefit of international bonds is likely to be much more modest. Here’s how the broad Canadian bond market stacked up against a global index over various periods ending in July 2014:
Canadian | Global | |
---|---|---|
1 year | 5.80 | 5.95 |
2 year | 4.33 | 5.07 |
3 year | 5.19 | 5.11 |
10 year | 5.52 | 4.89 |
Since February 1999 | 5.73 | 5.18 |
Source: FTSE TMX Canada Universe Bond Index (formerly the DEX Universe Bond Index) and Barclay’s Global Aggregate Bond Index (Hedged to CAD). Data from Dimensional Returns 2.0.
If you’ve got $100,000 and a 30% allocation to bonds, a global bond holding is probably not worth the added complexity and cost. But the argument for adding global bonds is might be stronger for those with very large portfolios and a higher allocation to fixed income. Say you have $2 million and a target asset mix of 70% bonds and 30% equities. Now your fixed income allocation is a hefty $1.4 million, and a shock to the Canadian bond market would deal a swift kick to your net worth. In that situation it might make sense to divide that among Canadian, US and international bonds to reduce your risk.
The only global bond I hold is the TLT up +17.09% YTD
@gil: TLT combines long-term bonds with currency risk, which is just about as volatile a bond fund as you can get. It’s worked out well this year, but it’s a not a strategy I would recommend for any long-term investor.
In short, if someone were willing to take more risk to (hopefully) get more return, they should lower their bond allocation and raise their equity allocation, rather than go for the risk-without-reward of international bonds, right ?
@Paul G: Yes, that’s right. Adding international bonds isn’t likely to increase your expected return, though they may lower your risk level somewhat (assuming you hedge the currency risk).
I’m having trouble finding the allocation by country for VBG. Vanguard doesn’t seem to post it with their fund information. Anyone know where one could find it? Interested to see the country-by-country allocation (especially allocation to Canadian bonds).
@Scotty: You can see the allocation for BNDX, which tracks the same index but is USD-hedged, here: https://personal.vanguard.com/us/funds/snapshot?FundId=3711&FundIntExt=INT#tab=2
Scroll down to see diversification by country. As of July 31, 2014:
Japan 23.1%
France 11.6%
Germany: 10.2%
UK: 8.4%
…
Canada: 5.1%
Was wondering about maybe preferred shares for the fixed income part of a portfolio, these can be complex to understand maybe we can have a article on preferred shares for boosting yields as they pay about 2-3 percent more? What is the risk with these, they are mostly all financial company’s I assume?
Thanks
Mark.
As always Dan your blog is enlightening. You suggest that a 70/30 (FI to Equity) investor should consider diversifying the Canadian Bonds to US and International Bonds but in your review of Share Owner portfolios you criticized that approach – see https://canadiancouchpotato.com/2014/06/09/shareowner-canadas-first-etf-robo-advisor/ . Is it the size of the portfolio in question?
I am considering the Share Owner Conservative Portofolio as the the overall fee of .49% or so on my $780,000 in investments would save me $8346.00 in fees over what I pay TD/Waterhouse.
Is there a better solution?
@Scotty,
Tricky one, I found the full holdings, ~2,700 of them, mix of corporate and government, no nice way to slice and dice. Couple highlights from a quick pivot:
2.3% Australia Gov
1.9% Belgium Gov
4.3% German Gov
1.5% Cad Gov
7.0% France Gov
7.2% Italy Gov
19.6% Japan Gov
2.0% Korea Gov
1.8% Netherlands Gov
3.7% Spain Gov
5.3% UK Gov
I only looked for stuff >1%, so some of these might be a little on the low side, but other than Japan, there is no real concentration.
@Marko: The portfolio size is definitely a factor, but the issue with Shareowner portfolios is that the US bonds are unhedged. Remember that you can tweak the Shareowner model portfolios however you want, so if you agree with my assessment you could just drop BND.
Hi CCP, I have a general question about bond ETFs. Any reason why you would recommend (or not) for young investors with a long time horizon to go for long duration/maturity bond etf ?
Thanks.
It looks like these funds, assuming their holdings roughly match the benchmark, will be slightly weighted toward premium bonds – so not great in a taxable account. I’m curious what you think about the tradeoff for a high net worth investor like the 2M with 70% in FI example you used. Such an individual would almost certainly have a large percentage of their fixed income holdings in taxable accounts. Does the diversification benefit outweigh the added tax expense? What factors would you consider when deciding between these ETFs and more tax-efficient Canadian options, like HBB or a GIC ladder?
@Marc-André: The main reason to avoid going with all long-term bonds is the volatility. My sense is that most people accept and tolerate volatility in stocks, but they are rattled by it on the fixed income side. Long-term bonds can and do suffer double-digit annual losses (and gains), whereas a broad-based bond fund is much more stable. My sense is that an average term of 10 years or less is right for most people, and short-term bonds (less than five years) are an even better choice for those who really want to dampen volatility.
@Nathan: The advice about not holding premium bonds in a taxable account applies here, too. In my example of the investor with $2 million, it would likely be possible to hold at least a portion of the bonds in registered accounts. If it’s not, then it would make more sense to go with more tax-efficient Canadian options such as GICs. For PWL clients the Dimensional fixed income funds are very useful here, because they are global and they use primarily low-coupon (discount or close to par) bonds and so are relatively tax-efficient and can be held in non-registered accounts.
@CCP, RE: “If it’s not, then it would make more sense to go with more tax-efficient Canadian options such as GICs.”
Or, if the investor is confortable using using swap-based ETF, simply use HBB for bonds holding in non-registered account and use RRSP space to hold foreign equities and avoid the withholding tax :)
@CCP, @mark, that’s an interesting question about preferred shares. Maybe an idea for a future post? In some ways preferreds straddle the fixed income/equity border. I know there was a run up in interest in preferred shares after the financial crisis but I don’t know if they deserve a place in a long term passive portfolio.
I understand your logic from a historical performance approach, but is it possible to make the argument that International bonds could hedge against unforeseeable issues with Canada’s economic credibility on an international scale?
I know it’s a doomsday argument, but is it worth noting?
@Sky: I think that’s a perfectly reasonable argument. I don’t think it’s so much about the doomsday scenario (Canada defaults on its debt and your bonds go to zero). But look at Japan, where interest rates have hovered close to zero for many years: Japanese investors who diversified with global bonds would have enjoyed higher returns with very little added risk.
Hi, thanks for another interesting article. One question, though: does foreign withholding tax apply to global bonds? Thanks.
@Tyler: There is no withholding tax on bonds: it only applies to dividends.
can someone explain the number of years the yield to maturity for bond funds is for ?
for example if i put $10,000 in a 2.5% 5 year GIC today and $10,000 in a core bond etf fund like VAB which has a yield to maturity of 2.4 % showing on their website how would the total value look after 5 years if there was no change in interest rates ???
VAB shows average maturity of 10.6 years
@Jake: This is great question, but the answer may not be as clear as you had hoped. :) The issue here is that a five-year GIC has a fixed maturity date, so you know exactly what your return will be. A bond ETF is very different: it is managed in a way that ensures its exposure remains more or less the same Indefinitely.
In simplified terms, if VAB has an average maturity of 10.6 years, it will likely have a very similar average maturity next year and the year after. Compare that with a five-year GIC. Next year it will effectively be a four-year GIC, and the year after a three-year GIC and so on.
The upshot is that it is impossible to predict the return of a bond ETF the way you can with a GIC. If all interest rates stay absolutely unchanged, then the total return on the bond ETF would be very close to its yield to maturity. But in practice, that will never happen. So if what you are looking for is stability and predictability with your fixed income investments, bond funds are not a great choice: GICs or a bond ladder would be more suitable.
By the way, the reason the GIC is offering a higher yield for five years than VAB is offering for 10 years is that you’re being compensated for the GIC’s liquidity, You can always sell VAB, but the GIC is locked in for the full five years.
great reply and written in a way I can understand, thanks
still learning investing, lately trying to read up on bonds and gic’s. bonds are part of fixed income asset class however with the reading about how interest rates affect the unit value i question if bond funds are actually fixed income? to me fixed income would mean if i put $10,000 in to something i expect to get a known return, stocks on the other hand are opposite, have no idea what the return will be. bond funds seem to be a little of both, bond unit values indeed can go down in value which can affect the total return while you get some interest as well..
to me gic’s are a great investment but aren’t advertised as that way. i guess companies promote about investments that they can make money on.
money market funds seem like one of the worst ripoffs in investing today, 5 year returns of less than 0.5% after paying 1% a year give or take to get it. mind blowing that this is even allowed by gov’t
@Jake: Bonds are one of the most difficult assets to understand. The math can be complex, and it’s not always intuitive (“why do my bonds fall in value when interest rates go up?”) so you’re not alone.
An individual bond typically does pay a “fixed income” and does have a predictable return if held to maturity. But you’re right: once you hold a basket of bonds in a fund and then trade those bonds to maintain a constant exposure, well, nothing is “fixed” anymore.
Couldn’t agree more about money market funds. I can’t think of any good reason to use one anymore when so many better alternatives are available, including investment savings accounts with no fee and CDIC protection:
https://canadiancouchpotato.com/2010/10/12/parking-cash-in-your-portfolio/
@CCP, I think your summary (last) paragraph is spot on.
My portfolio approach is to have a target for equities vs. fixed income and an independent target for home currency vs. foreign currencies. I do it this way to keep it simple in my own head and not mix the two ideas.
The reason why I want foreign currency exposure is simple… history shows us to not put all ours eggs into one basket. Throughout history all countries have seen their currencies or local government debt substantially devalued or very volatile at some point. For example, most recently, in 2008 think what happened to investors in Iceland that had a home country bias. Closer to home, reviewing the Canadian, Australian and New Zealand dollars around 2008/2009 you can see how wildly they swung. Or think of Argentina now. They examples never stop and never will.
This is what I think of when I hear the phrase “currency risk.” It thus puzzles me immensely why some genuinely smart people argue the opposite and that foreign currency exposure (especially foreign bonds) is risky.
I’ve held a 20% US dollar exposure for about 15 years now and never once questioned it. I’ve always felt that it reduced true purchasing power volatility over the years. If anything, I have questioned that 20% is too low.
It seems so clear to me that I want to have foreign currency exposure that I am actually starting to worry that I may be naïve or missing something.
Am I?
@Brian G: “It thus puzzles me immensely why some genuinely smart people argue the opposite and that foreign currency exposure (especially foreign bonds) is risky.” I think the important idea here is that a diversified portfolio needs different ingredients: one is foreign currency exposure and another is fixed income to reduce volatility. The problem comes when you mix them. When you add currency exposure to fixed income, you make the portfolio more volatile, not less, so you’ve eliminated one of the essential ingredients. If you instead take your currency risk on the equity side (which can actually reduce volatility even further) and keep the fixed income in your home currency you can accomplish both goals.
So if you plan on keeping 20% of your portfolio exposed to the US dollar that’s a good thing, but it makes more sense to get that through US stocks rather than unhedged US bonds.
@CCP, thank you for your attempt to explain this to me, and I do respect what you are saying but I’m afraid I am no closer to deeply understanding why holding only domestic bonds is less risky than holding a currency unhedged basket of bonds. I will attempt to explain my hang up…
To me holding a basket spreads risk and thus lowers it. Risks like: currency devaluation, single issuer default, etc. But to simplify the discussion and only focus on the currency aspect; consider the case of holding the lowest volatility asset: cash. Given the choice of only holding Canadian dollars or a basket of liquid currencies, the basket still seems more attractive. I can never see how the basket will be higher risk than holding only Canadian dollars alone because of the simple fact that the basket will lower currency devaluation risk caused by any Black Swan events. In a global economy the basket surely will hold it’s purchasing power better no matter what happens, won’t it? To back up this intuition, I consider all the cases in history (recent and not so recent) where currencies took a large shock because of one thing or another. Iceland being a great example, but there are many, many others.
Look at Iceland and three investors:
1. had all her/his wealth in cash in the bank
2. had all her/his wealth in a global basket of short term bonds
3. had all her/his wealth in a global basket of stocks
I’m sure #1 would have though he had the lowest risk portfolio if he had of listened to the common wisdom. #2 and #3 recovered and did fine if they stuck to their plans. #1 never recovered.
One might try to argue that Iceland was a special case and it would never happen to Canada. However, history shows that what happened in Iceland is actually more common than not.
Maybe the risk misunderstanding comes when one looks at a basket of currencies and watches it’s value in Canadian dollars alone. In that case, then yes, I agree that a basket will appear to be more volatile than holding Canadian dollars alone. But I guess my argument is that this is an illusion because the Canadian dollar alone doesn’t represent purchasing power. To really track volatility I would think one would need to compare each portfolio against a measure of purchasing power such as the Big Mac index.
Hi
Not sure under which section to ask this question. My query is regarding tax saving. I follow the couch potato strategy and invest in TD eseries funds in both RRSP and TFSA. I have maxed out on both for this year. I googled for how much tax would I be paying and it showed a little over 40K as my tax obligation. I would like to understand what are the avenues to save tax other than RRSP. Appreciate any inputs. Thanks
Adding to my above query-I am relatively a new investor in Canada. Have been investing for the past 2 years in RRSP and TFSA. I don’t have any other investments in non registered accounts. I am single so no RESP etc.
@Satuk: In most cases, once you run out of RRSP and TFSA room your only choice is a non-registered account. Sorry!
Thanks for the reply.
1.Is it wise to follow the same investment strategy(e-series mutual funds) for non registered account as well? I find it easier to invest in e-series.
2. if I max out on RRSP, does it mean that I have met my tax obligation or will I get a big shock when I file my returns and will I owe more tax?
@Satuk: e-Series funds are perfectly appropriate in non-registered accounts, though the bond fund will be very tax-inefficient.
Your RRSP contributions should lower the amount of tax you owe when you file your return, but many other factors also affect the amount of income tax you will eventually owe.
@CCP: Thank you very much for your response.
So with 30K in bonds you would recommend just Canadian, with 1.4 million global, US, and Canadian.
There’s a pretty big gap here. When would you say its worth getting into global bonds, 100K in bonds? 200K?
VBG and VBU are hedged with a MER of .2% so there not really that expensive and have protection against currency volatility, or am I missing something?
@Tim: I don’t know if there’s a simple cutoff, but my inclination would be to say somewhere around $500K, and even then, only if the fixed income allocation is at least 40% or 50%.
@CCP, has your opinion of holding non-hedged bonds changed since I last wrote on Sept. 15? If not, let me try to persuade you with some updated information.
Here are the returns I calculated for some popular ETFs for 2013, 2014 and annualized rate of return respectively.
Reporting in CAD
Symbol, 2013, 2014, ARR
———————————–
XBB -1.3%, 8.3%, 3.4%
BND (hedged) -2.1%, 5.3%, 1.5%
BND (un-hedged) 6.7%, 19.3%, 12.8%
50% XBB + 50% BND (hedged) -1.7%, 6.8%, 2.5%
50% XBB + 50% BND (un-hedged) 2.7%, 13.8%, 8.1%
Note that the returns of XBB and BND (hedged) are correlated whereas XBB and BND (un-hedged) are not. Therefore as we would expect from modern portfolio theory, a 50/50 mix of XBB and BND (un-hedged) has been less volatile than a 50/50 mix of XBB and BND (hedged).
This is just as I suspected from my logic argument given in a pervious post. I don’t know where the Canadian dollar is going but with my approach I don’t need to know. In simple terms a XBB and BND (un-hedged) mix will allow you to sleep at night and not worry about what direction the Canadian dollar (or US dollar) is going. A XBB and BND (hedged) portfolio offers no such diversification protection and will always be more volatile.
Given that fixed income is supposed to be the part of your portfolio that is less volatile, I still don’t understand why anyone wouldn’t take the mixed currency approach.
If one looks at the returns in USD the argument is even more compelling. Given that much of our purchasing power for things like food or anything imported is usually priced at the source in USD; using USD to measure performance is not a bad measure even for a Canadian portfolio.
As 2014 yet again pointed out… every year some country (or countries) in the world have their currency devalued so holding all your currency risk in one basket is a bad idea…. just talk with a Russian investor. What country will be next… nobody knows.
@Brian G: I think you need to run your numbers as part of a broader portfolio, likely containing at least 50% equities, most of which will not be in CAD. If your portfolio is 100% bonds, then certainly there is a diversification benefit to having some in unhedged foreign currency. That benefit decreases significantly (possibly into the negative) once you add a certain amount of foreign currency exposure from equities.
@Nathan, I don’t disagree but it doesn’t take away from the fact that when just looking at the fixed income component, the mixed approach benefits from modern portfolio theory/diversification.
Clearly any two funds that share a common currency will be more correlated than if they didn’t. E.g. XBB and XSP will be more correlated because they share the same currency than XBB and ZSP. This is why in my September posts, I advocated my approach of just setting an asset allocation target, then a currency allocation target and then constructing your portfolio.
For a hypothetical example: suppose one wanted a 60% equities, 40% bonds asset allocation and a 50% / 50% split between CAD and USD currencies (and no exposure to other currencies.) To achieve this, one could use:
20% ZSP – US Equity (un-hedged)
20% XIN – International Equity (hedged)
20% VCN – Canadian Equity
30% BND – US Bonds (un-hedged)
10% XBB – Canadian Bonds
If you wanted a different asset or currency allocation it’s clear to see how that can be done it will always work; even when you have a large or small fixed income component.
My sense is that everyone believes bonds should only be domestic and have no currency exposure because most academic financial literature is written from a US centric view where holding US currency may be the best option in that case but the best approach for Canadians may be different. If anyone has any Canadian academic papers that have studied this, I would be grateful if you point me to them.
After pressing send, I realize that an even better approach for the above hypothetical portfolio may be:
10% ZSP – US Equity (un-hedged)
10% XSP – US Equity (hedged)
20% XIN – International Equity (hedged)
20% VCN – Canadian Equity
20% BND – US Bonds (un-hedged)
20% XBB – Canadian Bonds
Reason being that it gives you a better diversification of interest rates between US and Canada. E.g. it shares interest rate risk 50/50 between the two. The asset allocation and currency risks remain the same. So, on the whole it seems better.
For anybody still following this… here are the events that happened since my last post:
– The Swiss bank shocked the market by letting their currency float against Euro and it shot up
– The Bank of Canada cut their overnight rate and the CAD dropped even more
– The ECB proposes 50 billion euros per month bond buybacks; if done this will drive down Euro even more
Since my first post here in September, I have personally heavily weighted up my portfolio towards USD and it has paid off handsomely. Today was a watershed for me and it really paid off. But frankly, I don’t think it is over and I am weighting up even more. I am now even most convinced that currencies are the largest risk in any portfolio now. How can one trust the CAD when the government has policies that will and have made it fall? Same for ECB/Euro and Japan. At this point, most currencies are not to be trusted.
I am increasing my USD exposure via USD denominated stocks, USD hedged exposure to European stock market and Japanese stock market, USD cash/short term bonds and gold.
Any bond (or cash) heavy portfolio that is all in CAD (or Euro or Yen) is likely suicidal going forward. You’ve been warned.
I’m putting this and my previous posts out there to be on the record, so that I can demonstrate the folly of not managing currency risk in a portfolio and specifically the folly of having a CAD centric portfolio like the new Couch Potato Portfolio on here is.
I’ve put on the record and now time will tell if I am right. So far it is in my favor.
@CCP:
Is there a tax efficient option for global bonds inside taxable accounts? Are premium bonds also a problem with global bonds funds/etf?
Thanks!
@Jas: According the webpage for VBG, the index has an average coupon of 3.1% and yield to maturity of 1%, so the fund is full of premium bonds. The only tax-efficient global bond funds I am familiar with are those from Dimensional Fund Advisors, but they are not available to DIY investors.
@CCP:
How do Dimension Fund Advisors’s bond funds manage to be more tax efficient (ie avoiding premium bonds) if they follow a passive indexing approach?
@Jas: DFA does not use a passive approach. They follow rules-based strategies, but they don’t track an index.
@CCP: I also notice that wide spread between the average coupon for BNDX and the 30 day SEC yield (which I think is similar to YTM, if i am not mistaken). Is that because of the European Central Bank and Bank of Japan’s monetary decision to use quantitative easing is causing interest rates to decline overseas?
With that in mind, I thought adding BNDX might be a useful diversification strategy for my bonds portfolio since we know the FED will be raising rates sooner or later, but this seem very unlikely for Europe or Japan, in the near term anyway. I am using BNDX instead of VBG because some of my portfolio resides in the US.
I noticed in another discussion that hedging doesn’t really work as advertised:
https://www.pwlcapital.com/en/Advisor/Toronto/Toronto-Team/Blog/Justin-Bender/September-2012/The-U-S-Currency-Hedging-Decision
With low interest rates, won’t the tracking error and hedging costs dominate over the returns? Is it really worth it?
… in particular I am thinking short term US corporate bonds, if it makes a difference.
@Jack: Thanks for the comment. Hedging often doesn’t work as advertised when it comes to equities, largely because of the volatility. With bonds it is much easier to hedge a portfolio accurately. But I would agree: unless an ETF is able to hedge effectively and keep costs very low, the diversification benefit offered by short-term US bonds may not be worth it.
Ah, interesting. So would that not be true for say a mid term bond fund such as ZMU which has a better yield to maturity of 3.54%. Do you think that extra 1% yield over similar CAD mid term corps would be enough to overcome the modest hedge drag if one wanted some diversification on the bond side?
@Jack: Yes, I would think any drag from the hedging cost would be significantly less than 1%.
@CCP Thanks for the helpful article! I have a question about my portfolio. The bond portion (30%) is around $400,000. Do you think its worth diversifying at this level? Almost all my holdings have to be in taxable accounts, so I was thinking 20% HBB and 10% HTB. Any thoughts?