Since the 2008–09 financial crisis caused bond yields to plunge to new lows, market forecasters have been predicting a rise in interest rates. And when bond investors think rates will increase, they tend to move to short-term bonds or cash. That has certainly been popular advice over the last three-and-a-half years.
As bond investors now know, the forecasters were spectacularly wrong about the direction of interest rates. Yields have fallen since the crisis, and as I wrote about in a recent feature for Canadian MoneySaver, anyone who moved to to short bonds or cash did far worse than investors who simply held the whole bond market. From 2009 through 2011, the iShares DEX Universe Bond (XBB) delivered an annualized return of 6.9%, compared with 4% for the iShares DEX Short Term Bond (XSB), and about 1.5% for cash.
The next three years
The fear of rising yields is even more palpable today, and the soothsayers may be proven right. But how badly would bond index funds suffer in a rising rate environment?
To get some insight, I ran some hypothetical scenarios to see how things might shake down if rates all along the yield curve climbed 100 basis points (1%) annually for the next three years. Then I asked BlackRock, provider of iShares ETFs, to estimate how that might affect XSB and XBB. The numbers in the table below are based on the following assumptions:
- During the entire three-year period, the profile of the two bond indexes—their duration and average maturity, for example—remain approximately the same as they are today.
- Credit spreads—the difference between the yield on government bonds and corporate bonds of the same maturity—remain where they are now.
- The rate increases occurs at the beginning of the year, resulting in an immediate decline in the value of the fund.
- Over the next 12 months, cash flows from coupon payments and the sale of bonds are reinvested at the new higher rates.
| Estimated returns if rates rise 100 bps annually | ||
| XSB | XBB | |
| Year 1 | 0.30% | -2.80% |
| Year 2 | 1.20% | -1.50% |
| Year 3 | 2.20% | -0.30% |
| Annualized return | 1.20% | -1.50% |
| Total three-year return | 3.70% | -4.50% |
Source: BlackRock Asset Management Canada, Inc. |
||
What’s the takeaway message from this simulation? The first is that bonds probably didn’t perform as poorly as you expected. With our simulated rate increase of 1% annually across the yield curve, XBB suffered three straight years of negative returns, but the losses were modest. Rising rates will hurt certainly bond returns, but the talk of bonds getting “killed” is probably overstated. Remember, too, that rate increases like this are likely to happen only if the economy gets red hot, which would probably lead to higher equity returns on the other side of your portfolio.
But there’s something else to consider: the opportunity cost you would have paid if you sat in short-term bonds or cash during the last three years, waiting for the rate hikes that were “certain” to come. As my Canadian MoneySaver article explains in detail, if bond returns over the next three years turn out to be similar to those in our simulation, XBB would still outperform both XSB and cash during the full six-year period beginning in 2009.
Investors should accept that no one can predict interest rate movements, and it is expensive to try. Rather than making tactical shifts that amount to guessing, bondholders are likely do best when they choose a diversified fund with an appropriate duration for their time horizon, and then hang on for the long term.
Hey Raffael, ironically i have moved to etf’s and a conservative model portfolio, but with my canadian equity component I have ‘destroyed’ the index with thi, enb, trp, and cpg along with xre over the last 5 years. i will probably wait until that streak ends and move to canadian and us etf’s. i think i got lucky and have to leave it at that.
Oldie I acualy hold bonds in RRSP and plan to do the same with TFSA. Equity portion is in non registered accounts. Target split of overall portfolio is 30% bonds, 70% equity.
You are right first 25% of RESP bond portfolio will be cashed in 6 years.
You are correct if interest rates unexpectedly rise XBB would take a hit, likely few percentage points. However, the following year I am using annual contribution and proceeds of sale for portion of XBB (together they would add up to 50% of value) to buy short term bond XSB that is now selling with higher coupons as consequence of raised rates. Similar thing but in different ratio would be repited year after and so on until year 6 at which point I am starting to cash in. The way I understand it I am matching bond duration with my time horizon. In theory that would mean I will get my invested capital back plus some. I can not predict which way interest rate will go nor when but I can structure portfolio in such a way that losses are minimal if the rates go up and return predictable if rates stay as they are or even better drop. There is no risk free investment. Risk can be mitigated but not eliminated.
Hey Raffael, on a final thought would just add with respect that in your effort to run into the arms of safety after getting hammered. you may have created a portfolio that has all the risk of an all equity portfolio. You just moved to the other side. Bond etf’s are not ‘safe’ at all.
When we go to a trend of rising rates, your long bond etf’s will get hammered. They could show 60 even 80, 90% capital losses (on paper). Mid range will get smacked as well. Check out some charts on the Bank of Canada site for interest rate history.
Following the crowd (retail investor) has always been the wrong thing in investing. They always get it wrong due to psychology. They are perhaps the most accurate telltale sign in a biz where predictions are mostly useless. But you can always count on the retail investor to get it wrong. And now they are fleeing equities and going into bonds.
I am personally happy to add more equities at this point. And will do so meaningfully if we get some nice pullbacks.
Dale, my understanding is that bonds (or bond ETF’s) had double digit losses ( on paper as you say) only two times. I believe 1936 and 2009 ( or around there) and loses were in more like 20% range than 60-90% as you seem to think.
@Dale: With respect, I think statements like that are irresponsible. I’m concerned they can scare inexperienced investors into making poor decisions. As Gordon says, long-term government bonds (using US data going back to 1926) have never experienced anything like the losses you’re imagining. They had a few years of single-digit losses in the 1950s and 1960s, even more modest loss in the 1970s, and then about a 15% or 20% loss in 2009 (depending which index you use). That 2009 loss was sandwiched between enormous gains (30%-plus) in 2008 and 2011.
Of course, the past is not the future, and unexpected things happen. But for long bonds to suffer the losses you’re talking about, rates would have to rise from about 2.5% to about 7% or 8% in a very short period. It is difficult to imagine what economic conditions would prompt something like that. I would put that in the category of “the US dollar is going to zero” and “gold is going to $10,000.” Possible, sure, but not a foundation for building an investment strategy.
Investment decisions should be made after careful assessment of the risks, not Armageddon scenarios.
Thanks, I know the post is provocative. Just following the bond ‘rules’ that I’ve read in other places and on your site. If a bond etf with 20-yr average duration sees just a 2% increase in rates, that’s a 40% drop minus income?
I would guess that 40-60% paper losses are not an Armageddon scenario. Rates rising 3-4% is back to ‘normal’ range.
Things have been calm for a while, but go back into other decades and you can see quick and violent spikes such as from 3% to 12%.
Wondering how bond etf’s would only lose 20% in this kind of environment thanks.
http://www.tradingeconomics.com/canada/interest-rate
@Dale: When did interest rates move quickly from 3% to 12%?
late ’93. some other quick spikes of several percentage points around the period as well.
chart is listed as Canada Benchmark Rates.
@Dale: Your data are wrong, or you must be misreading them. Here are the historical rates on long-term Canadian bonds directly from the Bank of Canada:
http://www.bankofcanada.ca/wp-content/uploads/2010/09/selected_historical_page14.pdf
The rate hike actually occurred in 1994, and they went from 7.16% that January to 9.40% a year later. The DEX Long Term Bond Index (the same one tracked by XLB) returned -7.4% in 1994.
There was another hike in 1999, from 5.23% to 6.23%, and the index returned -6%.
I have Canadian index data going back to 1948 and those are the two worst years.
Thanks. They are listing it as Bank of Canada as well. The spike(s) was over several months. You can adjust the time period on the chart linked
Thanks, I will do some research on rate and spikes.
Looks like that is the Bank of Canada rate that I found? Guessing Long Bonds don’t react with very much correlation (that is the actual long bond chart that you found). They move together of course, but LB’s are much more muted in response to bank of Canada moves?
Thanks again.
@Dale: Happy if this could help. The Bank of Canada website is a great source for historical rates.
I think the misunderstanding comes down to which specific interest rates people are referring to. People often talk casually about “rates going up,” but it’s not clear which rates they’re talking about. It’s not that uncommon for short rates to go up in the same year long rates go down. Or for short rates to go up a lot, and long rates to go up a little.
Central banks only have direct control over the shortest of short-term rates (i.e. the overnight rate referenced on the site you linked). Those at the longer end are driven mostly by market conditions: they are likely to rise when the economy heats up and fall during recessions. That’s why it’s so hard to imagine long rates shooting up dramatically in the current climate.
Thanks for that. Does that mean short term bonds would spike more violently (and more immediately) to rapid bank of Canada moves?
Again, appreciate your time and input/expertise.
@Dale: Yes, short-term bonds would be affected almost immediately by a spike in the Bank of Canada rate. But there’s no violence. With bonds that are only one or two years to maturity, the loss would be very modest and the silver lining is that you get to quickly reinvest maturing bonds in new higher-coupon bonds. So you recover quickly. If you happen to have your money in a high-interest savings account, you get a bonus, because the rates are likely to go up and you suffer zero loss.
One other thing you might have noticed is that banks change their rates on lines of credit and variable mortgages instantly when the Bank of Canada changes its overnight rate. But fixed-rate mortgages move in conjunction with five-year bond yields, since they usually have five-year terms. Sometimes these two rates go in opposite directions in the same year, most recently in 2010.
@CCP:
2 questions: firstly in the Bank of Canada Long Bonds chart you linked to, the rubric made a referral to the duration of the “long bonds” listed, but this was to a different page from the one (p 14) you had linked. What was the actual duration of these “long bonds”?
Secondly, I’m trying to wrap my head around the risks of various bond structures (apparently not as severe as some posters had feared) and the potential benefits. Given that a 3% TFSA savings account was proposed by one of the posters in this blog (July 12, 2012, “The Long and Short of Barbell Bonds”) as an alternative for bonds in this economic climate in the diversified portfolio, I wonder if this might warrant a side by side analysis and comparison of these two alternatives.
@Oldie: On the Bank of Canada site, “long bonds” seems to be mean 30 years. See this page, which says the long bond benchmark has a maturity date of 2041:
http://www.bankofcanada.ca/rates/interest-rates/canadian-bonds/canadian-bonds-yields-5-year/
If you can find a savings account that pays 3%, that would certainly be a good alternative to bonds. But the People’s Trust offer seems to be unique in Canada, and there is no guarantee that it will continue. A rate like that is often a teaser, and they can change it at any time. The going rate on cash these days is about half that.
There’s nothing wrong with holding cash. But one downside is that if rates fall, you just get a lower interest rate going forward: you don’t get a boost in the value of your holding like you would with bonds.
@CCP:
I get it. A teaser might be withdrawn any time. But as long as you’re in there till they withdraw it, you haven’t lost any opportunity, you get your advertised 3% till the month they yank the rug from under you, and you have to go back to Plan A, buying bonds again. My question was, if it could be assumed to be available, is it a good alternative to bonds, and, notwithstanding the teaser aspect, considering the guaranteed return of 3%, it appears your answer is yes. Bond yields are, what?… 2-3% and the return is plus/minus whatever capital gain/loss occurs. Given there is a downside possibility, and the upside appreciation is limited by the fact that prevailing interest rates are low enough to squeeze the possible declines to unlikeliness, to my unsophisticated eye, 3% guaranteed seems to beat the bonds. I’m just trying to figure out, what’s the catch?
The major difficulty, as far as I can see, is the fact that it is structured and offered only as a TFSA savings account. But you can still have your $5ooo in this TFSA (I’m talking about 2012 for the time being) and have your $15,000 worth of Equity ETF’s in another TFSA elsewhere, then do your rebalancing in 6 months when your next $5000 of TFSA headroom becomes available. Your rebalancing is only mildly complicated by the fact that it has to be done across 2 separate TFSA accounts. I’m assuming that transferring between different TFSA’s does not trigger all kind of tax difficulties.
@Oldie: I would agree with pretty much everything you said. Clearly 3% on a savings account is an excellent deal these days. The only quibble I would make is that it is not a “guaranteed 3%” for any specific period. It could change tomorrow. And there is a potential opportunity cost: if you were to put your $5,000 in a bond instead and interest rates fall, your bond will go up in value and you will continue to receive the coupon payment. That’s good. Ihereas your savings account will no longer be paying you 3% and there is no capital gain. That’s bad. Yes, you can get out of the savings account and buy the bond, but now it will be more expensive than it was before.
I don’t want to belabor this point: these are pretty small risks. And it is very unusual to be comparing cash at 3% to long-term bonds at 2.5%. This discrepancy just comes about because People’s Trust happens to be offering an extraordinarily unusual rate on cash.
@CCP:
Yes, I think you have given this all the attention it deserves. I see I have spent an inordinate amount of effort trying to parse an unusual, and quite likely in the long run, unhelpful situation. But a useful exercise in the learning situation for me.
would add that maybe the 90% loss is a stretch. Given how artificially low rates are, a 30-50% drop may not be an unlikely scenario. We r in uncharted territory.
Sorry, Oldie, I didn’t mean you were belaboring the point. I meant I was. :)
@ CCP:
Not at all; no offence was perceived. I think you correctly concentrate on the major points, and, also correctly, try to advise what are lesser, perhaps rather inconsequential considerations that beginners like myself have not yet learned to discern as being any different in importance.
Please tell me if I have this table correctly filled out for an example of requiring funds in year 2020. (assuming XBB duration is 7 years, and XSB is 3 years, for this example, just ignore equity amounts)
XBB XSB Cash Duration Year
100% 0% 0% 7.00 2013
75% 25% 0% 6.00 2014
50% 50% 0% 5.00 2015
25% 75% 0% 4.00 2016
0% 100% 0% 3.00 2017
0% 67% 33% 2.01 2018
0% 33% 67% 0.99 2019
The problem I have with understanding this is, remembering the words “stay in for the duration”, so how are you staying in for the duration if you have been contributing to XBB prior to year 2013, and then selling some in year 2014 to buy XSB? Do you think about it as a first in, first out type of scenario?
I’m just trying to actually plan how this would work the best, and having a hard time accepting this method.
Any additional thoughts, or idea would be appreciated.
Thanks, Que
@Que: Yes, this is a reasonable way to shorten the duration of a bond portfolio as you approach the date you need the money. (You might also do the same thing by mixing only XBB and cash.) Although it seems complicated, it’s not much different from how your risk exposure would change if you bought a single bond with a duration of 7 and held it until maturity.
I think the mental stumbling block here is that when you hold XBB for several years, you feel like you’re just “buying and holding.” But inside that fund, the turnover in holdings is quite large. All bonds are sold as soon as they have one year left to maturity, for example. And new bonds are constantly being added at the longer end to track the index and ensure that the duration of the fund remains more or less constant. So it feels like you’re not selling anything, but in reality you are.
Have a look at the target maturity bond ETFs from BMO. They do exactly what
you’re doing in your table: they simply mix two other bond ETFs in the right proportion to achieve the desired duration and they adjust the mix as necessary. The RBC target maturity ETFs work differently: they simply buy a number of bonds with the desired term and hold them to maturity. Two different techniques, but they accomplish the same thing.
https://canadiancouchpotato.com/2011/02/14/bmos-target-maturity-corporate-bond-funds/
https://canadiancouchpotato.com/2011/09/23/ask-the-spud-rbcs-target-maturity-etfs/
Let me know if this helps. It’s a confusing concept, I realize.
@Que: I think there’s another concept that needs to be clarified.
You asked “How how are you staying in for the duration if you have been contributing to XBB prior to year 2013, and then selling some in year 2014 to buy XSB.” You’re not. But it depends on what you are trying to achieve.
Let’s assume that you put $10,000 in XBB today, and the fund has a duration of seven years, and you need the money in seven years. The only assurance you have is that seven years from today (in August 2019), your total return including distributions is not going to less than 0%, no matter what happens to interest rates. (Technically, it’s not a guarantee, since duration is only an estimate, but it’s close enough.) So you will still have at least $10,000.
If you’re making more contributions along the way, the situation changes. Say the fund returns 5% in the first year so it rises to $10,500 in August 2013. That month you also contribute another $2,000, so the fund is now worth $12,500. The duration of the fund is still seven years, so you can now assume that it will be worth at least $12,500 in August 2020. In August 2019, in theory, the fund could be worth less than $12,500 (though it will be at least $10,000).
The same is true if you decide to shorten the duration along the way. Every time you make an adjustment, you are resetting the base value and the time period during which you can be sure you’ll suffer no capital loss.
Does that make sense?