Why has the iShares DEX Real Return Bond (XRB) dropped so dramatically this year? I thought this asset class was protective in times of rising interest rates (which are correlated with inflation), but perhaps I misunderstood. I also see the yield to maturity is almost zero. Please set me straight about the role of real return bonds in a portfolio. – K.T.
Let’s begin with a refresher on real return bonds, or RRBs. They have a lower coupon than traditional bonds, but their principal gets adjusted every six months according to the current rate of inflation, as measured by the Consumer Price Index.
For example, let’s say an RRB has a face value of $1,000 and a coupon of 3% annually (1.5% semi-annually). This bond would initially pay you $15 in interest every six months. However, if inflation rises by 1% before the next interest payment is due, the RRB’s principal will be adjusted upwards to $1,010. Now the 1.5% semi-annual coupon applies to this larger amount, and your next interest payment would be $15.15.
The coupons on federal RRBs today range from 1.5% to 4.25%, but the weighted average yield to maturity of XRB is 0.30%, which sounds dismal. But you need to understand that this is the real, inflation-adjusted yield. If you buy a 20-year RRB with a quoted yield of 0.30% and you hold it to maturity, you’ll earn a return of 0.30% plus the inflation rate over that period. A nominal bond of the same maturity might boast a yield of 2.8%, but if inflation is 2.5% during the life of these two bonds, their total return will be the same.
In fact, if you want to learn the market’s expected inflation rate, you can look at the difference between the yield of an RRB and the yield of a conventional bond of the same maturity. As of February 27, the Bank of Canada lists the current yield on benchmark long-term bonds as 2.53%, while the yield on benchmark RRBs is 0.52%. That suggests the market’s best forecast for inflation is just over 2%.
Time to get real
The primary role of RRBs in a portfolio, then, is inflation protection. Investors often consider gold, commodities, real estate and dividend stocks as inflation hedges, and these asset classes may indeed offer some protection. But no asset class is more directly tied to the inflation rate than real return bonds. If you hold an RRB to maturity, your principal is guaranteed to maintain its purchasing power.
RRBs also play an important diversification role, since they are not highly correlated with equities, real estate, or even traditional bonds.
However, like all asset classes, real-return bonds are exposed to several risks, and investors need to understand these so they’re not surprised when the asset class performs poorly, as it has in the past year or so.
First, the average term to maturity of the DEX Real Return Bond Index is over 20 years, and the average duration is about 16. This makes RRBs vulnerable to rising interest rates. If long-term rates rise just half a percentage point, a fund such as XRB can be expected to decline about 8%. Long-term yields are currently about 20 basis points higher than they were last May, which explains a good part of this ETF’s recent price drop.
RRBs will also perform poorly (relative to nominal bonds) if inflation turns out to be lower than expected. And if the CPI declines during a six-month period, the principal of RRBs will be adjusted downward, which would causes XRB and similar funds to fall in value.
Remember, a diversified portfolio will always have short-term winners and losers: indeed, that is the whole point of diversifying. From 2009 through 2011, RRBs were double-digit performers: in the last of those years they returned over 17% while Canadian and international equities declined significantly. (I should note that investors were being told not to buy them at time.) Now the tables have turned: in the last 10 months or so, RRBs have declined while stocks have soared.
It would be wonderful if we could predict the next outperforming asset and reposition our portfolios appropriately. Unfortunately, we cannot. So the sensible strategy is to diversify across many asset classes with low correlation, hold them all the time, and rebalance when appropriate.
I was considering purchasing XRB in January when I re-balanced. I decided against it as I wanted to make it as simple as possible for my first major re-balance and I was already switching VTI to VXUS and had to do FOREX. Plus that account is <$50,000. I'm definitely going to incorporate XRB in the future.
Well it took me a year but I’m now convinced sticking to the plan of a diversified portfolio is the way to go (and have a life to boot). As long as you have a predetermined stocks to bonds ratio and stick with it and rebalance regularly you’ll do great. It is really hard to do though and I’m living proof… last year I thought I could outsmart the market. Stocks “looked” too high to me based on research and I was convinced they would drop, so I decided to be overweighted in bonds. If I had only invested the 60:40 ratio of stocks:bonds and rebalanced once every few months my return for 2012 would have been several percentages higher. I’ve learned that a simple, boring plan is good in long term investing. This site has really helped too. Thx CCC.
@ CCP: I wonder about the posted inflation rate i.e. how accurate it is, or is the govt. just telling us what sounds good, so people don’t get mad. My informal, unsubstatiated belief from actually spending money on gas, electricity and groceries tells me that inflation is way worse than the 2-3 % advertised. Have you ever seen any evidence to verify posted inflation rates?
@madMike: I’ve heard others make that same claim, and I’m not sure what to make of it. The methodology used to calculate the CPI is pretty transparent (see this link). Note that some inflation measures specifically exclude energy and food.
Thanks for the detailed discussion.
One additional question that came to me is the appropriateness of RRBs for portfolios reaching a retirement horison of under 20 years (the average time to maturity of XRB). This sort of thinking goes into choosing whether to invest in short, mid, or long-term bonds. Does this carry over to RRBs?
Hi Dan,
I’ve always thought that if I were going to hold a lot of bonds (I don’t) then I’d probably mix half-and-half real-return and nominal to minimize my exposure to inflation surprises in either direction. Does that sound reasonable to you?
Wow! Great post. Bonds and bond ETFs confuse me the most and this was a great explanation. However, what do you mean that – if inflation goes up the “principle will be adjusted upwards”? Do you mean the underlying holdings in the fund? If so, where does this money come from?
@JPalmer – I’m with you, bonds and bond ETFs confuse me as well. What doesn’t sit right with me is that you can actually lose money on bonds/bond ETFs (depending what interest rates do), when they are supposed to be the low-risk portion of the portfolio. That’s why I decided to set up a 5-year ladder of GICs at good rates (Achieva) to replace what would be the bond portion of my portfolio. The rest of my portfolio follows the Complete Couch Potato formula.
@JAH: It’s a good question. There is some value in having that inflation protection during retirement, since you are no longer earning an income. (Younger people might expect their salaries to increase and offset inflation somewhat.) Remember, too, that if you retire at 65 and are in good health, you mat be planning for a 30-year retirement, so a duration of 16 may not be a problem.
@Patrick: The Complete Couch Potato devotes one-quarter of the bond component to RRBs. One-third might be OK, but half is probably too much, especially if the portfolio is heavily weighted to fixed income overall. (If you’re just 10% RRBs and 10% nominal bonds, that’s probably fine, but I don’t think I would go 30% RRBs and 30% nominal bonds.)
@JPalmer: The bond’s principal increases in value, just like any other security might go up in price. (In the case of an ETF, it too would rise in price to reflect the value of its underlying holdings.) That’s how the inflation protection is built in to the bond: remember, the coupon always stays the same, so the interest payments only rise because that coupon is paid based on the principal amount.
Think of it like this: imagine you’re a landlord and you charge your tenant a monthly rent equal to 1% of the house’s value. If house prices in your area go up, so does your rent, and so does the price you’d get if you sold the home.
@KJF: There’s nothing wrong with using a GIC ladder in place of a bond fund if it makes you more comfortable. Overall, however, the risks are not much different. This post may be of interest:
https://canadiancouchpotato.com/2010/03/29/bonds-v-bond-funds/
@CCP: Can you elaborate on your replies to JAH & Patrick in regards to an example of say, 80% bonds & 20% equities, how much would you max out on the Real Return Bonds?
Thanks,Que
@Que: There’s no magic number, but I don’t think I would want more than one-quarter to one-third of the fixed income in RRBs, and I wouldn’t want them to make up more than about 15% of the portfolio overall. So in your example, you might go with 15% RRBs, 65% nominal bonds, 20% stocks.
As described, with interest rates presently low and expected to rise, it seems like the expected return on any investment in RRBs will only beat inflation in the event of another stock market crash.
Though I can appreciate the value of ‘diversification’ above all else, surely almost any other fixed income option is a better idea for the foreseeable future, right? I mean we’re not due for another crash for at least 5 years. :)
Both my husband and I are retired and have defined pension plans that cover all our living costs and savings. My pension also has a cost of living adjustment. Therefore our investment portfolio has 10% allocation for bonds. I wonder what the correct mix should be? Can you direct me to CCP archives or elsewhere to ensure we have a balanced portfolio. Thanks, Dan
@CCP: The Complete Couch Potato sample portfolio has been a valuable reference point as a reasonable investment template for us students. At the risk of redundancy, may I ask if the 10%-30% allocation of DEX Real Return Index Bonds-DEX Universe Bond Index would be appropriate for a retiree with a 15-20+ year horizon (assuming for the moment one is comfortable with the 60% Equity component). What I am getting at is whether the DEX Real Return portion, with an average term of 20 years, and the DEX Universe portion, with a range of terms averaging 9+ years covers the territory adequately between now and the horizon, without the need for further diversification in the fixed income category, term-wise or otherwise.
A discussion of the tax treatment of RRB would be useful. My understanding is they are best held in a tax sheltered account?
Since interest is paid by RRBs and interest has the worst tax treatment compared to other investment returns, they would be best kept in a tax-sheltered account.
@Claire: I wish it were simple, but there is no “correct” asset allocation. There is only an allocation appropriate to your situation, and that depends on many factors.
@Oldie: With a 15- to 20-year horizon that mix should be just fine. There’s no need for further diversification, but it may make sense to shorten the overall duration as the horizon gets shorter. You probably wouldn’t want to be funding your living expenses with 20-year bonds.
@Shorty and Brian: All bonds should be held in a tax-sheltered account whenever possible, but this is particularly important for RRBs. The coupon payments are fully taxable as income, as you would expect. But in addition, the principal adjustments are fully taxable as they accrue, even though you don’t actually receive any income. (You will receive this money when the bond matures and the principal is returned to you.) In this way they are a little like strip bonds.
@CCP: Great post. Timely as well, since I’ve been considering adding real return bonds to my portfolio. I don’t doubt that RRBs provide diversification, but I am confused as to how they offer inflation protection. As you mention, if inflation increases then the RRB principal will increase and the RRB will increase in value. OK, but if the inflation increase is accompanied by an increase in interest rates (as usually happens) that would decrease the value of the RRB since they are long bonds. Wouldn’t these two things offset each other? Or am I missing something?
@CCP: Many thanks for answering questions that have been vexing me lately. I moved into ETFs in 2010, have 10% in XRB and 40% in other bond ETFs. A large part of my portfolio is in the taxable account, including nearly all of the XRB. While recently comparing my transaction history with tax-related information that came with the 2010 and 2011 T3s, I was surprised to find that the total XRB distribution reported in the latter was about double that reported in the former. I had assumed that the column reporting distributions “Paid to You” referred to cash that had been paid to me. I asked my brokerage (BMO Investorline) about the discrepancy. They explained that the T3 included both cash and Reinvested Distributions (RD). They also provided a link to the relevant iShares page, which includes a footnote indicating that the RD amount should be added to the adjusted cost base (ACB). It appears that RD is related to Return of Capital much like antimatter is related to matter: RD increases ACB, provides no cash, but is taxed like interest, while ROC reduces ACB, provides cash, but is not taxed. (Curiously, the RD from XRB includes a portion designated as ROC that is apparently annihilated during the transaction.) Anyway, I was left with two questions: Why does RD exist? and Should I get XRB the hell out of Dodge? Your comment to Shorty and Brian provides the answers. I am still left wondering why the Investorline tax statement provides several reminders to subtract ROC from ACB but says nary a word about RD or the need to add it to ACB. Their ommission could result in naive investors (like myself) paying too much capital gains tax. Is Investorline in cahoots with the Canada Revenue Agency?
@Smithson: It is possible that an increase in inflation might coincide with rising rates, and that these two forces would act against each other. But the relationship is not so tidy that they would cancel each other out. It’s also important to remember that RRBs should not be considered in the abstract, but rather in comparison to nominal bonds, which are vulnerable to rising rates and have no inflation protection.
@PeterH: A fund needs to report reinvested dividends because these are taxable to the investor even if they are not paid out in cash. This is perfectly legitimate, since fund investors did receive the benefit of the distribution: it’s just that its reflected in a price increase rather than as cash in your pocket. I’m not sure I understand your other questions, and I’d suggest you talk with an accountant or tax adviser to make sure you report everything accurately.
@CCP: Thanks, I think I understand!
“but it may make sense to shorten the overall duration as the horizon gets shorter.”
I take it this means, gradually selling off DEX RRBs down to zero and increasing DEX Universe to the full portion of fixed income as horizon winds down to 10 years, then starting the purchase of DEX Short Term Bond Funds to replace DEX Universe gradually as the horizon winds down to fewer years, and finally conversion to GIC’s or cash, is this correct?
What about protection from inflation once you sell off DEX RRBs? Are there intermediate and short term equivalents of DEX RRBs that can be used as the horizon winds down?
This post is hugely instructive! I realize now that I understood inflation incompletely before; getting this right is yet one more essential interlocking brick in my investment education.
How is the current price of an ETF like XRB determined? Based on your descriptions above it sounds like it is determined (somehow) strictly by its internal value. However, I also understand that an ETF at least in in some respects is similar to a stock whose value is determined (as I understand it) by market forces, i.e. the perceived value by buyers and sellers of the stock.
@RH: A stock or bond’s “internal value” and its market value are one and the same. Interest rates (and therefore bond prices) are impacted by supply and demand, and those effects are what determine bond prices. So a bond ETF’s price is simply based on the total value of all the bonds in the portfolio.
CCP: Is there any particular reason that you would target 25-33% of the bond component of a portfolio for RRBs? Is it just because the duration is so long and it doesn’t make sense to increase the duration too much because of the interest rate risk?
@CCP: I should clarify the last part of my earlier comment. You have explained why XRB generates large reinvested distributions, making XRB a poor choice for taxable accounts. I also understand how to update the ACB. My main point is that Investorline fails to report crucial information needed for the ACB update. The transaction history includes only cash distributions. The T3 “Summary of Trust Income” reports only the combined amount of cash and reinvested distributions. To calculate the amount of reinvested distributions (which are added to the ACB), you need information from the iShares website. Investorline provides no warning about this matter. I have asked Investorline to provide a more complete report. Perhaps other brokerages already do this. An accountant would know all of this. Those of us who do our own accounting need to be wary of things that we don’t know that we don’t know.
@PeterH: Sorry I misunderstood. Yes, as you’ve discovered, brokerages are not required to provide all the information you need to manage taxable account, so they don’t usually do any more than necessary. Their book values are also notoriously unreliable, so you need to keep your own careful records.
@CCP: Looking up the particulars of the 2 RRB’s you list on your ETF page, I find that XRB holds only 13 component RRBs and ZRR only 7. I have reviewed your general comments in a previous post about unbundling as it applied to REIT index ETFs; in spite of lower MERs of RRB ETFS, compared to the REIT ETF’S, given the lower number of component holdings in the two RRB ETFs, is their any specific reason why one should not consider unbundling XRB or ZRR if the math indicates a worthwhile saving?
@Oldie: Unbundling is fine in theory but there are the practical problems of finding inventory, high commissions, difficulty investing small amounts and rebalancing.
@CCP: From a retirement planning perspective, what are your views on planning to hold RRBs to maturity vs planning with “cashing in early” in mind
(Obviously the former requires some coordinated timing and the latter entails market risk/reward)
thx
@qas: Holding RRBs (or any other bonds) to maturity is fine if that’s what you prefer. But there are so few RRB issues in Canada that it may be difficult to to build a portfolio that neatly matches your time horizon.
@CCP: Still trying to understand Index Fund basics, I reviewed XRB and ZRR, and both claim to “to replicate, to the extent possible, the performance of the DEX RRB Non Agency Bond Index…” and yet ZRR, while holding the same 6 Govt of Canada RRBonds as XRB, has totally omitted the 7 Provincial RRBonds that is included in XRB. I am a little disconcerted, but maybe that is a good thing because I learned something — it is the onus of the investor to verify that any Index fund actually covers the territory that it claims to do.
Having understood that, could I ask your opinion as to how much the omission of the Provincial RRBs affects the “RRB” worth of ZRR, or even if it causes any meaningful deviation from the DEX RRB Non Agency Bond Index.
@Oldie: The two funds do not claim to track the same index: XRB tracks the the DEX Real Return Bond Index. I would not expect the two funds to differ much, since the six federal RRBs make up about 85% of XRB.
@CCP: *Doh* (DEX RRB Non Agency Bond Index is not the same as DEX Real Return Bond Index)… so much to learn!! Thanks again.
Incidentally, I dug up
http://howtoinvestonline.blogspot.ca/2010/06/which-way-is-best-to-invest-in-real.htm
which gives a nice overview of RRBonds vs XRB vs ZRR which helps to answer some of my and qas’s concerns above. One nugget I found useful was that ZRR will automatically invest the interest if requested; XRB won’t. Also, I didn’t realize that 1% commission is buried in the price when buying Bonds directly; I believe this was actually mentioned in a previous CCP post, but I forgot; and also availability of (directly purchased) actual RR Bonds is sometimes so limited that you can wait weeks for your order to be filled — Dan, you alluded to this difficulty, of course, but I had no idea it could get so awkward to fill your order
I would like to invest in real return bond but in a mutual fund because my allocation to it will be only around +/- 6500 and I would prefer the distribution reinvest automatically without fee.
Does this fund is a reasonable option : PHILLIPS, HAGER & NORTH INFLATION-LINKED BOND FUND SERIES D(RBF1650:CAD) ?
It new from this year, I would like your opinion on it.
Thanks
@Francis: That fund is not actually new: it has been around since 2009, but all of the Phillips Hager & North funds recently changed their codes to “RBF” to reflect that they have been bought out by Royal Bank.
There are only a tiny number of real return bonds available in Canada, so every fund is going to hold almost exactly the same thing, though in slightly different proportions. With that in mind, this fund is a reasonable alternative to a real-return bond ETF if you are investing a small amount, would like all interest payments reinvested, and want to avoid trading commissions.
If you look at the three-year performance of the PH&N fund, it is about 30 basis points lower than XRB, which is explained by the difference in MER.
@CPP: I’m wondering if you could shine the light on something I don’t understand regarding the YTM of XRB. About a year ago the YTM of XRB was around about 0.6%. That is, the bonds were quite expensive. Over the next few months, the price of XRB fell as interest rates rose, and the YTM went up over 1%. Today the price of XRB is about 7% higher than a year ago, yet the YTM is 2.66%. Why is that? Many thanks!
@Tristan: This is a great question. RRBs used to report their YTM in real terms: that is, 0.6% meant 0.6% plus inflation. With the inflation expectation at 2% these days, that effectively meant a 2.6% yield. It seems they have stopped reporting this way and now just give the nominal yield. I guess this makes sense, since reporting real yields was more confusing for investors.
@CCP: Aha! Thanks. BTW, I love the new search feature on the site. From time to time, someone says something I know is wrong, and I vaguely remember you wrote a post in it and want to find that post. Sometimes it was not obvious which of the previous sections it would be in, so it was often difficult to find. Now it’s a cinch.
Hello CCP, would it be better to get exposure to RRB through an ETF (such as ZRR) or is it preferable to buy the actual bonds directly from a discount broker (and pay the applicable spreads)? There are some experts who suggest it is better to hold them to maturity to avoid exposure to changes in the market value of the etf when the funds are needed, which exposure would not exist if the RRBs are held to maturity (except, of course, the inflation adjustment). This is for a long-term horizon (i.e. late 30s). Any views or articles on the subject would be helpful. Thanks.
@Phil: It’s important to remember that the risks are the same whether you hold the ETF or the underlying holdings directly. The individual bonds will also fluctuate in value every day, just like the ETF. And while it is true you can hold the bonds to maturity and get back the face value (plus inflation adjustments) you can also hold the ETF for a period equal to its duration and also eliminate the risk of loss of principal. Remember that RRBs in Canada have extremely long terms, so holding to maturity is a very long commitment. An ETF is likely to provide better liquidity (the retail spreads on RRBs are very high) and make it easier to rebalnce.
Hi CCP
Can you comment on whether it makes sense for someone hoping to retire relatively soon (5 years best case, 10 years worst case) to hold RRB ETFs given their long duration (15.5 years for XRB)?
My (naive?) thinking is that such a long duration could result in some serious volatility when interest rates rise, which would be detrimental to retirement funding (only RRSP and CPP, OAS to rely on).
For comparison I looked at Vanguard US TIPs ETF, VTIP, and it has a duration of 2.5 years. Is there a reason why lower duration RRB ETFs are not available in Canada?
Somewhat off topic, but could you comment on the appropriateness of these mixes for the fixed income portion of my portfolio (fixed income would be 40% of total portfolio – all percentages below expressed in terms of total portfolio):
15% CLG, 15% VSC, 5% XRB, 5% ATL5000, or
17.5% CLG, 17.5% VSC, 5% ATL5000
I’m worried I may be being too conservative, but I have a hard time seeing any upside to holding longer duration bond ETFs.
[Converting the US and foreign side of the portfolio to ETFs was significantly easier than deciding on what to do on the bond side]
TIA
@Jim R: Great questions. It’s true that RRBs in Canada have very long durations, and they can be very volatile. In the context of a portfolio they can be a good diversifier, but they are certainly not an essential ingredient. There’s nothing wrong with dropping them if you would prefer a more stable ride with your fixed income.
I can’t comment on whether the fixed income portfolio you’ve presented is right for you, but I don’t see any problems with it in principle (whether or not you include the RRBs). You’ve covered both government and corporate bonds and kept your duration relatively low, so you should not expect much volatility, which is probably what you want from bonds. Try not to agonize over the decision!
@CPP, I’m wondering if Jim’s question was also about having to withdraw some of an asset (minimum RRSP withdrawal and need to maintain allocations) in 5 years time and buying XRB now with it’s 15 year duration. Unless you were willing to withdraw something else and allow XRB to increase it’s allocation if a price drop occurs, wouldn’t it be safer to buy shorter duration bonds instead?
@Tristan: It is always safer to buy shorter-duration bonds. You should never buy a bond fund with a duration longer than your time horizon, so RRBs are not a suitable investment for someone who expects to withdraw the money in five years.
That said, you shouldn’t have money in equities if you need it in five years either. Once you get to the stage where you’re drawing down your RRSP/RRIF the usual strategy is to keep an amount equal to the minimum withdrawals in cash and a ladder of GICs etc.
You make mention at the end of your article that you can not time the market. Actually, yes, you can – if you understand cycles. I have made a lot of money since I discovered the study of cycles.