Risking Less and Prospering using Real Return Bonds? [RRBs]

Asset allocation, risk, diversification and rebalancing. Pros/cons of hiring a financial advisor. Seeking advice on your portfolio?
longinvest
Veteran Contributor
Veteran Contributor
Posts: 3956
Joined: 10 Sep 2012 17:26
Location: QC

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by longinvest »

8Toretirement,
8Toretirement wrote:With RRB's there is no set par value as with nominal bonds.
An RRB has a set value in inflation-adjusted terms (CPI-adjusted terms, if you prefer).
8Toretirement wrote: The holder could receive lesser amounts due to deflation.
The holder will receive exactly the RRB par value adjusted to CPI. The par value is set in inflation-adjusted dollars.

If there is deflation, the par value will have preserved its buying power. If there is inflation, the par value will have preserved its buying power. That's exactly what I want from an RRB investment: exact CPI-adjusted future amounts.

A nominal bond is deceitful; it promises you a precise par value in future dollars. If I buy a nominal bond which promises me $1,000 in 30 years, I take a huge risk as I have no idea whatsoever what the purchase power $1,000 will be in 30 years, because the Bank of Canada's (BoC) current monetary policy is only set for the next 5 years. In 5 years or later, the BoC could change its inflation target or abandon it altogether. It's unlikely, I know, but it is not impossible.

What I am saying is that, between an RRB and a nominal bond, the more speculative investment is the nominal bond, not the RRB. I may not know the par value of the RRB in future dollars, but I actually know the par value of the RRB in current dollars, which is way more important!
8Toretirement wrote: The problem you get, especially with the secondary market when trying to match short term obligations within a portfolio, is RRB's are very volatile compared to short term bonds and the face value at maturity could be less than you paid.
The RRB will pay me back exactly its par value in CPI-adjusted terms on the day it matures. The purchase power of the principal will be preserved regardless of inflation, deflation, hyper-inflation, or hyper-deflation. That's exactly what I want from an RRB: preserving purchase power.
8Toretirement wrote:RRB's are not understood very well
Right. :wink:
8Toretirement wrote: "Determining the value when RRB matures: When an RRB matures, you'll get the face (par) value times the then current index ratio. In other words, you'll get the face value indexed for inflation. This will probably be more than the original face value, but if there's been a period of deflation, it could be less."
Yep, exactly what I want. I don't care what the par value will represent in future dollars; I care about the purchase power of the par value as expressed in current dollars.

The RRB will deliver exactly its par value in inflation-adjusted terms at maturity; it will have preserved its purchase power.
8Toretirement wrote:An entirely separate issue: If you are forced to withdraw during a 13% loss in value, as has been achieved by RRB's in recent history you lose purchase power from an asset that is supposed to provide safety. I believe I read the worst year for nominal short bonds was 4%. Big difference.
That's a different issue. Now we're discussing a change of plans.

If I keep bonds within a liability-matched non-rolling RRB ladder until maturity, I'll get exactly its expected payments in CPI-adjusted dollars (coupons and principals). If I sell any bond before maturity, I am exposed to interest rate risk; if an RRB is far from maturity, the potential loss (or gain) is bigger, if it's close to maturity, the potential loss (or gain) is smaller.

Historically, a total-market RRB ETF (such as XRB) has been less effective as ballast for short-term stock volatility than a nominal bond ETF (such as XBB). This is not surprising as the average duration of the RRB market is 16 years whereas the average duration of the nominal bond market is 8 years leading to significantly higher RRB volatility.

In the US, Vanguard offers a short-term TIPS ETF (VTIP) which has a low duration and, as a result, very little volatility. It would be difficult to build an equivalent RRB ETF in Canada because maturities are spread 5 years apart.

While less volatile than RRBs, nominal bonds carry a huge inflation risk. In contrast, RRBs have no inflation risk but are more volatile. But, let's be clear: RRBs remain much less volatile than stocks!
8Toretirement wrote:If the investor has not received OAS and CPP they might be better served by delaying these pensions that are inflation protected while also receiving higher payouts and associated additional inflation protection on the additional payout.
You're preaching to the choir. :wink:

Here's a post I made in August: Delay OAS to 70, spend 8.8% more at 65!
Variable Percentage Withdrawal (finiki.org/wiki/VPW) | One-Fund Portfolio (VBAL in all accounts)
longinvest
Veteran Contributor
Veteran Contributor
Posts: 3956
Joined: 10 Sep 2012 17:26
Location: QC

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by longinvest »

Quebec wrote:You may have mentioned this already, but why did you opt for a fixed percentage of RRBs in your portfolio, rather than a target dollar amount?
I might have given (or not) an explanation in this thread: Tinkering with my asset allocation? [Real Return Bonds]
Quebec wrote:Also, why choose an ETF as opposed to a customized liability-matched ladder of RRBs?
We live in a rich country where no one is left to die from hunger on the streets. All on my life, I have adapted to my income level. When I had very little income, I lived on what I had (I didn't save, but I didn't accumulate much debt, either). When my income grew, I spent more and I started saving and investing.

As I wrote in an earlier post (http://www.financialwisdomforum.org/for ... 63#p585520) the average Canadian worker can easily get $19,000 annually in inflation-adjusted dollars by retiring at 65 and waiting until age 70 to claim OAS and CPP/QPP. I lived on my own on less than that (in inflation-adjusted dollars) in my early 20s. Of course, I don't look forward to going back to such a situation, but I have always been very adaptable. As long as I have something to eat and a shelter, I can survive.

The only way my current 50/50 (stocks/bonds) portfolio wouldn't be sufficient to fill the gap between retirement and age 70 is if Canada's economic system was destroyed, in which case RRBs wouldn't save the day.

So, I simply don't worry about targeting a precise inflation-adjusted income in retirement. I plan to remain flexible in retirement. I'll adapt my spending to what my income will allow for. I'll have a non-indexed DB pension, OAS and QPP at age 70, and a 50/50 stocks/bonds portfolio where the stocks are half domestic, half international, and the bonds are half nominal, half inflation-indexed. It's a good enough portfolio which will be combined with a robust withdrawal method (VPW). I'm not worried; I'll survive on that even if inflation roars its head and I happen to live a very long time.
Quebec wrote:In other words, why treat RRBs as ''just another asset class'' when it can serve almost as a home-made inflation-indexed annuity to cover basic needs in retirement, or a bit more than that, if you can afford it?
First, a portfolio of individual RRBs is not much different from an RRB ETF. An RRB ETF is just a collection of RRBs. Each RRB will eventually mature at par (in inflation-adjusted terms). Of course, the marked-to-market value of the ETF changes everyday. But, the marked-to-market value of a portfolio of individual RRBs fluctuates similarly. What differs is the exact targeting of maturities. But, unfortunately I don't know the exact date of my death. So, if I was to build an RRB ladder, I would have to build a rolling ladder that I could only start unrolling in very old age, 30 years before the highest possible human survival age. I remember reading that some scientific researchers have estimated it around age 125. So, I could only safely start unrolling it at age 95. I might as well let BMO do all the hard work (within ZRR) until then, don't you think? :wink:

Second, I believe that VPW withdrawals from a diversified balanced portfolio combined with OAS, QPP, and a nominal pension, will be sufficient for a flexible person like me. Anyway, if the worst was to happen, I would apply to get GIS; I sincerely don't expect to ever get there, but it's a safety net.
Quebec wrote:With a fixed percentage of RRBs in your portfolio, and regular rebalancing, you are in part letting your future equity returns decide how much RRBs you will have (in dollars) on your retirement date.
As you said, I treat RRBs as just another asset class (or, more precisely, as a distinct sub-class of domestic bonds). I simply don't care about having a precise value of RRBs (or even a precise portfolio value) on my retirement date. My retirement income will be determined by the total of my pensions and flexible VPW portfolio withdrawals. I'll adapt my spending accordingly.
Variable Percentage Withdrawal (finiki.org/wiki/VPW) | One-Fund Portfolio (VBAL in all accounts)
User avatar
adrian2
Veteran Contributor
Veteran Contributor
Posts: 13333
Joined: 19 Feb 2005 08:42
Location: Greater Toronto Area

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by adrian2 »

longinvest wrote:But, unfortunately I don't know the exact date of my death. So, if I was to build an RRB ladder, I would have to build a rolling ladder that I could only start unrolling in very old age, 30 years before the highest possible human survival age. I remember reading that some scientific researchers have estimated it around age 125.
Memento mori: does human lifespan have a hard upper bound?
Imagefiniki, the Canadian financial wiki
“It doesn't matter how beautiful your theory is, it doesn't matter how smart you are. If it doesn't agree with experiment, it's wrong.” [Richard P. Feynman, Nobel prize winner]
longinvest
Veteran Contributor
Veteran Contributor
Posts: 3956
Joined: 10 Sep 2012 17:26
Location: QC

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by longinvest »

adrian2 wrote:
longinvest wrote:But, unfortunately I don't know the exact date of my death. So, if I was to build an RRB ladder, I would have to build a rolling ladder that I could only start unrolling in very old age, 30 years before the highest possible human survival age. I remember reading that some scientific researchers have estimated it around age 125.
Memento mori: does human lifespan have a hard upper bound?
I was thinking about this: Humans Won't Ever Live Far Beyond 115 Years
For years, people have been saying that the first human who will live to 150 has already been born. That’s unlikely, say Jan Vijg, Xiao Dong, and Brandon Milholland, from the Albert Einstein College of Medicine. After looking at demographic data from the last century, they think that human lifespan has a hard ceiling at around 115 years. A few rare individuals like Calment may surpass that limit, if only slightly, but on average, our species will not.
But, there is some criticism: Human age limit claim sparks debate
The limit is surprising, says Vijg, given that the world’s population is increasing — supplying an ever-increasing pool of people who could live longer — and that nutrition and general health have improved. “If anything you would have expected more Jeanne Calments in recent years, but there aren’t."

But not everyone agrees with his team's interpretation. The age experiencing the greatest increase in survival may have plateaued in many countries, says James Vaupel, founding director of the Max Planck Institute for Demographic Research in Rostock, Germany. But it has not yet plateaued in some that are particularly relevant to this research, namely Japan, which has the world’s highest life expectancy — 83.7 years for those born in 2015, nor in France or Italy, which have large populations and high life expectancies.
...
Researchers also cite future developments in medicine that could further increase maximum lifespan, which the paper does not address.
Variable Percentage Withdrawal (finiki.org/wiki/VPW) | One-Fund Portfolio (VBAL in all accounts)
User avatar
Quebec
Veteran Contributor
Veteran Contributor
Posts: 1645
Joined: 24 Oct 2009 16:49
Location: Quebec City

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by Quebec »

@Longinvest: very clear explanation a few posts above, thank you. It's simple, it's easy, it makes sense.

BUT:
I don't know when I'm gonna die either, but my spreadsheets have us dead at 95. So in a RRB ladder, the last rung could mature at that age, not at age 125. In the very unlikely event we're not both dead then, there would still be OAS + QPP + pensions + other non-RRB assets. A person aged 90 does not really need to own RRBs maturing 20-30 years later. ZRR currently has less than 30% invested in issues maturing in the next 10 years (the 01Dec21 and the 01Dec26), so IMO is not an ideal choice for a 90 year old male. I grant you that until one reaches maybe 70 or 80, the ETFs are probably OK. They could then perhaps be converted into an inflation-indexed annuity, if such things are available then.

- - - - - - -
Thanks again for recommending the book, it has lead me to examine different scenarios and strategies in detail, and thinking more about inflation, sequence of returns risk, etc. especially in early retirement scenarios. I am not quite done yet with this exercise, and I plan to have a professional validate my calculations when I'm done. But so far, given a manageable savings rate, a paid-for house upon retirement, assumptions about OAS, QPP and workplace pensions, and a conformable but not luxurious retirement budget, I arrive at two extreme scenarios:

1. We can retire at age 55 with $500-600k in personal investments (in today's dollars), an aggressive portfolio, and all fingers crossed. If stocks crash just before retirement, work longer. If stocks crash a few years after retirement, come back to work or reduce the budget. I don't like this scenario at all.

2. We can retire at age 60 (or maybe even 59) with $0 in personal investments! This is also stupid. What if govt or workplace pensions are cut, etc.

Not having a clearly defined target retirement age makes it very difficult, obviously, to define a suitable savings rate, or asset allocation. With retirement at age 57, there are still many options in terms of savings rates and asset allocations. One of these options is 100% RRBs and maintaining our current savings rate for another 12 years. But that leaves little room for error, I'd like to have a larger pot of money.

I get a feeling I'm going to get sick of my spreadsheets in a few days and just continue to save as much as possible, with a balanced portfolio.
Imagefiniki, the Canadian financial wiki: a knowledge base of financial subjects written from a Canadian perspective
8Toretirement
Contributor
Contributor
Posts: 225
Joined: 26 Nov 2014 14:55

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by 8Toretirement »

I don't think we are going to agree on deflationary pressure on RRB's so lets turn this around to volatility which feeds into my overall decision to limit RRB's in my portfolio.

I am assuming withdrawal phase decisions. I am also assuming when withdrawing funds to finance retirement then the retiree would likely prefer limiting volatility in preference of predictability.

Look at the 10 yr charts for RRB's and Short Term Bonds. You can project the index on the chart also. I also use short term bonds with less than one yr to maturity with duration of less than 0.5yrs which has almost no volatility. Very stable in this bond market.

RRB

https://www.blackrock.com/ca/individual ... -index-etf

Short Term Bonds

https://www.blackrock.com/ca/individual ... -index-etf
ockham
Veteran Contributor
Veteran Contributor
Posts: 2214
Joined: 04 Apr 2006 21:50
Location: The Prairies

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by ockham »

Quebec wrote:
I get a feeling I'm going to get sick of my spreadsheets in a few days and just continue to save as much as possible, with a balanced portfolio.
I'm an early retiree (4 yrs). I've been reading this thread with interest, notwithstanding my natural aversion to both spreadsheets and deep investment theory.

I made the retirement decision based on the following back-of-an-envelope analysis.

1. My fixed income allocation (as a $$ amount) was large enough such that my FI plus OAS plus CPP plus my partner's modest partially indexed DB pension was sufficient to fund our retirement needs.
2. A serious threat to the "sufficient to fund" conclusion is unexpected inflation. RRBs are a FI product that addresses that threat.
3. RRB ETFs have a long duration that does not decrease over time. The duration of my liabilities does, however, decrease over time. Therefore, RRB ETFs are simply not suitable. Individual RRBs will have to do.
4. Maturity dates on individual RRBs are lumpy. Building an RRB ladder to match liabilities is clumsy.
5. My solution: Three RRB rungs ('26, '31, '36), with GIC ladders and a strip bond ladder to reach '26, and (prospectively) fill the gaps between the RRB rungs. Not perfect, in fact quite klugeish, but "good enough", to use longinvest's phrase.

This leaves equities to fund retirement wants, and as a cushion against errors in the analysis.
User avatar
ghariton
Veteran Contributor
Veteran Contributor
Posts: 15954
Joined: 18 Feb 2005 18:59
Location: Ottawa

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by ghariton »

8Toretirement wrote:I am assuming withdrawal phase decisions. I am also assuming when withdrawing funds to finance retirement then the retiree would likely prefer limiting volatility in preference of predictability.
I still don't understand this.

Nobody is recommending a 100% RRB portfolio. In my case, I hold a portfolio with RRBs and (mostly) U.S. equities, with the allocation varying over time. Since RRBs are negatively correlated with equities -- estimates of the magnitude of the correlation vary, but their significance is pretty well established -- then a combination of RRBs and equities actually reduces the volatility of the portfolio.

In any case, if there is an emergency need for cash, it is most unlikely to involve selling off more than, say half your portfolio -- if that -- when you are middle-aged or older. With a portfolio containing both RRBs and equities, you can choose which securities to sell in times of need. It doesn't have to be your RRBs. In my case, when my brother-in-law needed money to save his business in April 2009, I sold equities, and didn't touch the RRBs. That was in line with my portfolio allocation (a fixed amount of RRBs) and so volatility of RRBs was not a factor.

To my mind, RRBs are an instrument to build up a retirement nest egg, accumulated over the long term as a part (not all) of your periodic investments. They are supposed to provide the money at age 65 (or whenever) which will permit you to retire even though you don't have a DB pension, or the one you have is inadequate. Emergencies before then should come out of your other investments, the ones intended to fund the nice-to-have rather than the necessities. That is consistent with maintaining the ability to cover your basic needs in retirement. To my mind, everything fits together very nicely.

_________________________________________________________________________________________

The main drawback to RRBs is that their rungs are spaced five years apart, as longinvest has noted. That shouldn't be a big problem in the accumulation stage. In the withdrawal stage, one can use nominal fixed investments, with an average duration of 2.5 years or less, to plug the gaps. Unless one is looking at very high inflation, there shouldn't be too much damage. Of course, if you are really worried, there are strips with which to fill in the gaps. But as I've said before, these are highly illiquid and hard to acquire and, to my mind, not worth it.

Finally, as a fairly extreme measure, one could use VTIP, an ETF of short duration U.S. inflation-indexed treasuries. Not much problem with duration or gaps there, but there is currency risk. While I believe in holding a large part of my investments in USD, I am not ready to hold all of them in USD. Diversification in currencies, as in everything else.

George
The juice is worth the squeeze
longinvest
Veteran Contributor
Veteran Contributor
Posts: 3956
Joined: 10 Sep 2012 17:26
Location: QC

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by longinvest »

Quebec wrote:my spreadsheets have us dead at 95. So in a RRB ladder, the last rung could mature at that age, not at age 125.
Managing a rolling RRB ladder until age 65 (or 70, if you update the death age to 100, as I would highly recommend) is a pain. I just don't see my wife trying to calculate bond yields and multiply face amounts by index ratios to figure our if she is being skinned alive by the broker's bond desk. I'm sure she'll never be willing to learn bond math, even less real-return bond math. And, I haven't included the "fun" of setting up a 4-rung GIC ladder when an RRB rung matures, as well as dealing with all the coupons!

All hope is not gone, though. It is quite easy to simulate an non-rolling ladder without sacrificing liquidity, by combining cash and and an ETF. It won't be perfect and might introduce a small inflation risk, but it should be good enough.

Let's say that I wanted to set up, at age 50, a non-rolling RRB ladder that lasted from age 55 to 94 paying exactly $10,000 (inflation-adjusted) each year (at the beginning of the year). Let's assume an average 0% real return on our RRB ladder (close to current average real yields) to simplify calculations. Any additional return will be gravy. What would be the duration of the payments? It would be:
  • (5 + 6 + ... + 44) / 40 =
    (4+1 + 4+2 + ... + 4+40) / 40 =
    ( (40 X 4) + (1 + 2 + ... + 40) ) / 40 =
    (40 X 4) / 40 + (40 X 41 / 2) / 40 =
    4 + 820 / 40 =
    24.5 years.
That's a higher duration than ZRR's 16 year duration! So, I would just drop (40 X $10,000) = $400,000 in ZRR at age 50.

Every year, I would recalculate the duration of my payments and make sure that the duration of my ZRR+cash portfolio is no more than the duration of my payments minus one year (because a normal ladder would lose one year duration during the year, but the ZRR+cash combo won't).

So:
Age 50: payment duration 24.5 => 100% ZRR, 0% cash (that's what we did).
Age 51: payment duration 23.5 => 100% ZRR, 0% cash
Age 52: payment duration 22.5 => 100% ZRR, 0% cash
Age 53: payment duration 21.5 => 100% ZRR, 0% cash
Age 54: payment duration 20.5 => 100% ZRR, 0% cash

Age 55: We need to make a withdrawal. We take an inflation-adjusted $10,000 withdrawal. Remaining payments have a 20 year duration; we don't want our ZRR+cash to have a higher duration than 20-1 = 19 years. => 100% ZRR, 0% cash
...
Age 65: We need to make a withdrawal. We take an inflation-adjusted $10,000 withdrawal. Remaining payments have a 15 year duration; we don't want our ZRR+cash to have a higher duration than 15-1 = 14 years. => 14/16 ZRR => 87.5% ZRR, 12.5% cash
...

This won't be perfect, but will be good enough. If real yields go down (below zero?), remember that after 16 years, the ETF will have returned to the point of indifference; future returns will be lower than expected, so you'll either have to top-up your RRB holding with new money, or somewhat reduce withdrawals according to the new yields. If yields go up, on the other hand, don't increase withdrawals.

The ZRR/cash ratios can be conveniently precalculated, and the ZRR holding set up as a DRIP investment. So, a single transaction per year will be sufficient. The only annoyance will be to chase the best high-interest savings account (https://www.highinterestsavings.ca/chart/) so that the cash remains competitive with short-term real yields. Alterna Bank currently pays 1.95% nominal; that's currently higher than the real yield on the shortest RRB based on the Bank of Canada's 2% inflation target, but it doesn't include unexpected inflation protection.

It's easy to setup and manage, but the outcome is somewhat more uncertain than with a pure RRB ladder. The return of ZRR + cash relative to an RRB with exactly targeted maturity can differ causing some losses (or gains). But, it's good enough. If I was to adopt the separate safe income floor idea, that's how I would implement it.

Another advantage is that, at age 70, the ZRR holding could be conveniently sold and the proceeds invested into an inflation-indexed annuity (assuming they exist at that point), eliminating longevity risk. The fluctuations of ZRR's market price are likely to be correlated to the price of such an annuity. For example, if real yields go up, ZRR's price will drop, but inflation-indexed annuities will get cheaper too. (Allowing for this possible RRB to annuity conversion is one reason why I recommend to use age 100 instead of 95, just in case insurers update their mortality tables by the time you get to age 70, as age 95 would probably work today).
Variable Percentage Withdrawal (finiki.org/wiki/VPW) | One-Fund Portfolio (VBAL in all accounts)
Flaccidsteele
Veteran Contributor
Veteran Contributor
Posts: 4523
Joined: 06 Mar 2014 12:52
Location: Retired Gen Xer somewhere on the planet earth

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by Flaccidsteele »

A good reminder of why I don't invest in bonds. I don't have the brainpower! :lol:
User avatar
Quebec
Veteran Contributor
Veteran Contributor
Posts: 1645
Joined: 24 Oct 2009 16:49
Location: Quebec City

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by Quebec »

Quebec wrote:I grant you that until one reaches maybe 70 or 80, the ETFs are probably OK. They could then perhaps be converted into an inflation-indexed annuity, if such things are available then.
longinvest wrote:...at age 70, the ZRR holding could be conveniently sold and the proceeds invested into an inflation-indexed annuity (assuming they exist at that point)....
OK we agree here.
longinvest wrote:I just don't see my wife trying to calculate bond yields and multiply face amounts by index ratios to figure our if she is being skinned alive by the broker's bond desk. I'm sure she'll never be willing to learn bond math, even less real-return bond math
Yes, that's probably the ''fatal'' issue (bad pun intended) with the rolling ladder. But OTOH, if she buys and indexed annuity at age 70, how does she know she's getting an acceptable deal?
longinvest wrote:And, I haven't included the "fun" of setting up a 4-rung GIC ladder when an RRB rung matures, as well as dealing with all the coupons!
Oh, come on. Anyone can deal with the GIC ladder. Let's say the maturing RRB is worth $50k. Divide the available funds into five with a calculator (if needed). Withdraw one part ($10k) to spend during the next year. Call the discount brokerage and ask them to buy $10k of the highest yielding one year GIC, $10k of the highest yielding two year GIC, etc. The whole process takes no more than 10 minutes (every five years) and involves one mathematical operation. And the RRB coupons, you say? Yeah OK that's messy. Got to go eat breakfast now. :)
Imagefiniki, the Canadian financial wiki: a knowledge base of financial subjects written from a Canadian perspective
longinvest
Veteran Contributor
Veteran Contributor
Posts: 3956
Joined: 10 Sep 2012 17:26
Location: QC

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by longinvest »

Here's an illustration of how managing the duration of the ZRR/cash portfolio works.

I'll do the calculations in constant dollars (e.g. inflation-adjusted dollars).

Let's say that I want an annual $10,000 withdrawal over the next 10 years, starting with $10,000 next year. The average duration of my payments is:
  • (1 year + 2 years + 3 years + 4 years + 5 years + 6 years + 7 years + 8 years + 9 years + 10 years) / 10 payments = 5.5 years
According to my rule, we want the duration of our RRB/cash portfolio to be 4.5 years (one year less than our payments).
Assuming that the duration of ZRR is 16 years, we should invest 4.5/16 = 28.125% of our money into ZRR, and the rest (71.875%) into high-interest cash.

Assuming that ZRR's YTM is 0% real and that our cash is yielding as much as inflation (Alterna Bank's 1.95% is close enough to the Bank of Canada's 2% target), we need 10 X $10,000 = $100,000 to fund our 10 years of $10,000 payments.

So, we invest $28,125 into ZRR and put $71,875 at Alterna Bank. Now, we all know that ZRR is volatile. We invest our money in the morning at 9:30AM, and at 10:00AM, the Governor of the Bank of Canada makes a statement which inflames markets. The average YTM of the RRB market jumps to 2%. Hello ZRR losses! Let's look at the damage (simplified scenario using a zero-coupon 16-year RRB and a financial calculator):
  • N= 16, i= 2, FV= 28125, PMT= 0 => PV = 20488
Our $28,125 has melted to $20,488, a 27% loss. Ouch! But, let's not forget that from now on, we'll be earning 2%* per year on that money!

* [Added:] As I assume that cash will continue to barely match inflation, ZRR would get a bonus in addition to the 2% yield due to the steep yield curve. It would be likely, though, for cash to start returning more than 0% real, reducing the bonus due to the yield curve steepness, but increasing the returns of our model. My point: By not including any bonus and keeping cash yields at 0% real, I am painting a worse than likely scenario.

For now, let's assume that yields remain unchanged for the next 10 years.

We still have one year before our first $10,000 withdrawal. Our $20,488 will grow to $20,898 during the year.

Year 1

Our ZRR holding is worth $20,898 and our cash is worth $71,875. (Don't forget that these are constant dollars and we're assuming that our cash grows at the inflation rate). That's a total of $92,773 before withdrawal.

We withdraw $10,000. We're left with $82,773.

The average duration of our payments decreases by 0.5 year to 5 years. We want our portfolio to have an average 4 year duration. So, we want to put 4/16 = 25% in ZRR, and the rest in cash. That's $20,693 in ZRR and $62,080 in cash. We need to sell ($20,898 - $20,693) = $205 of ZRR and put it in cash.

In summary, this first year we sold $205 of ZRR and withdrew ($10,000 - $205) = $9,795 from Alterna Bank.

Year 2

Our ZRR holding is worth $20,693 + 2% = $21,107 and our cash is worth $62,080. That's a total of $83,187 before withdrawal.

We withdraw $10,000. We're left with $73,187.

The average duration of our payments decreases by 0.5 year to 4.5 years. We want our portfolio to have an average 3.5 year duration. So, we want to put 3.5/16 = 21.875% in ZRR, and the rest in cash. That's $16,010 in ZRR and $57,177 in cash. We need to sell ($21,107 - $16,010) = $5,097 of ZRR and put it in cash.

In summary, this second year we sold $5,097 of ZRR and withdrew ($10,000 - $5,097) = $4,903 from Alterna Bank.

Entire Scenario

It is pretty easy to set up a spreadsheet to make the entire calculation:

Code: Select all

       Year   Duration        ZRR       Cash      Total
          0        5.5    $20,488    $71,875    $92,363
          1          5    $20,693    $62,080    $82,773
          2        4.5    $16,010    $57,177    $73,187
          3          4    $11,908    $51,599    $63,507
          4        3.5     $8,398    $45,347    $53,745
          5          3     $5,489    $38,424    $43,913
          6        2.5     $3,190    $30,833    $34,023
          7          2     $1,505    $22,581    $24,086
          8        1.5       $441    $13,675    $14,117
          9          1         $0     $4,125     $4,125
         10        0.5         $0    -$5,875    -$5,875
We notice that we came up short by $5,875 in the last year. Why did that happen? It happened because the average duration of ZRR was 16 years, when it experienced a significant yield change (+2%), yet, we didn't let the money grow back for a period of 16 years (we had a 10 year horizon). We also artificially reduce our portfolio duration once a year, which can cause losses in our simplified model which underestimates the annual gains of the ETF (due to rolling down the yield curve). So, that small deficit was to be expected.

Note that even with such a pretty much worst-case scenario, the deficit is very small. Had we had a longer time frame and considered the yield curve, probably the deficit would have translated into a small surplus. (With a 16 year duration, the yield curve bonus can be quite interesting).

To dampen this small sequence of return risk, there are two solutions:
  • For fixed plans of 10-years or more (e.g. bridge retirement to OAS+CPP): Just invest an additional 10%, initially.
  • For lifelong plans: Be conservative with the age of death ceiling.
An non-rolling RRB ladder covering 30 years of payments or less doesn't need such buffering and, as a result, could be a bit less expensive. But, if the plan is longer, the RRB ladder will need rolling and will be exposed to real interest-rate risk and require some buffering too.

It's a tradeoff: The RRB ladder would offer the best protection against inflation but would not be manageable by my wife (at least not until it became a non-rolling ladder, when it would be best to buy an annuity, anyway). The ZRR+cash scenario would have some inflation risk but my wife would be able to multiply the sum of ZRR+cash (after withdrawal) with a list of pre-set allocation percentages to rebalance the portfolio. (This could be beyond the abilities of some spouses, though).

To be fair, the ZRR+cash scenario would remain at 100% ZRR (or pretty close to it) until it is time to buy the annuity. So, the ETF choice could be a perfect match, too. :wink:
Last edited by longinvest on 20 Nov 2016 23:19, edited 1 time in total.
Variable Percentage Withdrawal (finiki.org/wiki/VPW) | One-Fund Portfolio (VBAL in all accounts)
User avatar
Peculiar_Investor
Administrator
Administrator
Posts: 13268
Joined: 01 Mar 2005 14:52
Location: Calgary
Contact:

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by Peculiar_Investor »

Fascinating topic and discussion that's giving me lots to consider and ponder. Very educational. :thumbsup: :thumbsup:
Imagefiniki, the Canadian financial wiki New editors wanted and welcomed, please help collaborate and improve the wiki.

Normal people… believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet. – Scott Adams
longinvest
Veteran Contributor
Veteran Contributor
Posts: 3956
Joined: 10 Sep 2012 17:26
Location: QC

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by longinvest »

Quebec wrote:Not having a clearly defined target retirement age makes it very difficult, obviously, to define a suitable savings rate, or asset allocation. With retirement at age 57, there are still many options in terms of savings rates and asset allocations. One of these options is 100% RRBs and maintaining our current savings rate for another 12 years. But that leaves little room for error, I'd like to have a larger pot of money.

I get a feeling I'm going to get sick of my spreadsheets in a few days and just continue to save as much as possible, with a balanced portfolio.
Quebec,

I think that simplicity wins. My wife and I live below our means and invest the difference. We paid off our mortgage some years ago. We have a balanced portfolio spread across RRSPs, TFSAs, and non-registered accounts. We just keep adding to it. It is growing nicely.

Bodie's book provided a refreshing look at the concept of risk and helped reflecting on how to plan a robust retirement. But, I am not adopting his separate needs/wants portfolio approach during accumulation, because we are clearly saving and investing more than strictly enough (according to Bodie's own metrics). We'll only consider setting up a separate RRB portfolio at retirement if absolutely needed in addition to my pension and our expected OAS/QPP pensions (quite unlikely). More likely, we'll want to eliminate longevity risk during our 70s, if still healthy, by buying the minimum amount of inflation-indexed annuities* required to do so.

* Or indexed to a fixed rate equal to the Bank of Canada's inflation target, if inflation-indexed annuities are not available.
Variable Percentage Withdrawal (finiki.org/wiki/VPW) | One-Fund Portfolio (VBAL in all accounts)
User avatar
Quebec
Veteran Contributor
Veteran Contributor
Posts: 1645
Joined: 24 Oct 2009 16:49
Location: Quebec City

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by Quebec »

longinvest wrote:We paid off our mortgage some years ago.
I need to work on that (faster mortgage reduction). That's the main way to reduce risk for us at this stage (cf the book title), and 'pays' more than buying RRBs (or most nominal bonds) in the current low interest rate, low inflation environment. When that's done, I'll look at the ''separate RRB portfolio'' idea again, to cover a fraction of our retirement needs.

I've also decided to try to swim 3 times a week instead of 2. No point in planning retirement to age 95 if I'm not in above-average shape.
Imagefiniki, the Canadian financial wiki: a knowledge base of financial subjects written from a Canadian perspective
OnlyMyOpinion
Veteran Contributor
Veteran Contributor
Posts: 4231
Joined: 24 Jan 2014 23:17

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by OnlyMyOpinion »

Interesting thread. I’d never heard of Bodie and what I know about RRB’s I’ve learned from this thread.

Maybe of some interest that throughout our accumulation years we built a ‘fixed income foundation’ for our retirement savings by maximizing our RRSP contributions every year and doing the unthinkable – investing 100% into fixed income.

Crazy, why? After starting out with our first GIC RRSP accounts in 1979 we were convinced by those friendly but ever-changing advisors at the bank to buy mutual funds. After several years we decided we didn't like seeing a lower account balance each time our quarterly statement arrived by mail, so we jettisoned them and decided to stick with guaranteed investments (typical uninformed consumer).

At the same time though we aggressively paid off our first and only mortgage within 5 years, career income was rising, and I was a bit more informed about investing. By the mid-1980’s we had more annual savings than our RRSP’s could adsorb, and were compelled to begin looking at equities. We started investing in individual 'blue chip' stocks in an unregistered account (this is before ETF’s and TSFA's). Even this was a modest beginning until our comfort in equities grew along with income.

We have continued to this day to use our RRSP's as our 100% fixed income, 'retirement foundation' accounts (we have no company pension). Beginning about 2006 we began to add strip bonds (with corresponding risk). The accounts are now comprised of 22 strip bonds maturing from 2017-2025. Not evenly spaced but 'lumpy' with 0% to 26% held in a given year, depending on the maturing amount and best available yields at the time (>corp bbb). Over 37 years we have contributed $338k to our RRSP's. Today their MV is $1.1MM and their weighting in our overall assets would be considered appropriate. We’ll continue to maintain strips in our RRSP's for another 10 years, and after that as RRIF's. Meanwhile dividends and CPP provide our retirement income.

I don't suggest this as a recipe for accumulating fixed income retirement assets, but merely to point out there are different ways to get there.
I'm sure if we had been smarter about it we could have done much better but we can live with the results.
Flaccidsteele
Veteran Contributor
Veteran Contributor
Posts: 4523
Joined: 06 Mar 2014 12:52
Location: Retired Gen Xer somewhere on the planet earth

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by Flaccidsteele »

OnlyMyOpinion wrote:I don't suggest this as a recipe for accumulating fixed income retirement assets, but merely to point out there are different ways to get there.
I agree with this.

As a person who has never invested in bonds at all, I also like reading about different ways to "get there". It's very interesting.

A million ways to make a million bucks, indeed.
8Toretirement
Contributor
Contributor
Posts: 225
Joined: 26 Nov 2014 14:55

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by 8Toretirement »

ghariton wrote:
8Toretirement wrote:I am assuming withdrawal phase decisions. I am also assuming when withdrawing funds to finance retirement then the retiree would likely prefer limiting volatility in preference of predictability.
I still don't understand this.

Nobody is recommending a 100% RRB portfolio. In my case, I hold a portfolio with RRBs and (mostly) U.S. equities, with the allocation varying over time. Since RRBs are negatively correlated with equities -- estimates of the magnitude of the correlation vary, but their significance is pretty well established -- then a combination of RRBs and equities actually reduces the volatility of the portfolio.

In any case, if there is an emergency need for cash, it is most unlikely to involve selling off more than, say half your portfolio -- if that -- when you are middle-aged or older. With a portfolio containing both RRBs and equities, you can choose which securities to sell in times of need. It doesn't have to be your RRBs. In my case, when my brother-in-law needed money to save his business in April 2009, I sold equities, and didn't touch the RRBs. That was in line with my portfolio allocation (a fixed amount of RRBs) and so volatility of RRBs was not a factor.

To my mind, RRBs are an instrument to build up a retirement nest egg, accumulated over the long term as a part (not all) of your periodic investments. They are supposed to provide the money at age 65 (or whenever) which will permit you to retire even though you don't have a DB pension, or the one you have is inadequate. Emergencies before then should come out of your other investments, the ones intended to fund the nice-to-have rather than the necessities. That is consistent with maintaining the ability to cover your basic needs in retirement. To my mind, everything fits together very nicely.

George
RRB's are very volatile compared to short duration nominal bonds. This is the downside to RRB's. Their volatility. In the withdrawal stage volatility is your enemy. Losses cannot be recouped when the funds are withdrawn for expenditure.

If you have a very large component of RRB's that are underperforming due to deflationary effects on RRB returns, then the investor will be forced to withdraw from other assets. Sounds good right.

Here is the problem. If deflationary pressures continue then the portfolio will become unstable because you would have to withdraw from other assets which increasingly concentrates the portfolio into RRB's.

Longinvest provided the formula to show a drastic drop in the overall portfolio value, you do get an increased interest rate; however, with the long durations of RRB's I suspect the diminishing portfolio will not recoup the losses from higher interest rates due to withdrawals. Using a RRB ETF will increase the effect as they continually roll there duration rates. ZRR has an effective duration of 15.9 years, which is rolling not declining with each passing year.

A measure of risk is volatility, RRB's are much more volatile, therefore risky, compared to a laddered 5 year nominal bond portfolio with a short term duration for the FI component, supplemented by GIC's. Not against RRB's, just not buying the 50% RRB portfolio as decreasing risk. Duration is a measure of risk. And by this measure alone I would be leery of using RRB's for a large component of my portfolio in the withdrawal phase.
longinvest
Veteran Contributor
Veteran Contributor
Posts: 3956
Joined: 10 Sep 2012 17:26
Location: QC

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by longinvest »

8Toretirement,
8Toretirement wrote:Longinvest provided the formula to show a drastic drop in the overall portfolio value, you do get an increased interest rate; however, with the long durations of RRB's I suspect the diminishing portfolio will not recoup the losses from higher interest rates due to withdrawals. Using a RRB ETF will increase the effect as they continually roll there duration rates. ZRR has an effective duration of 15.9 years, which is rolling not declining with each passing year.
I have provided a mathematical illustration of how using an imperfect combination of ZRR and cash instead of a liability-matched RRB non-rolling ladder merely introduces a very small risk of withdrawal deficit in a worst-case scenario where I unfairly underestimated the returns of both cash and ZRR after the rise in real yields.

I have explicitly stated, in prior posts, that RRBs are guaranteed to pay their coupons and par value in inflation-adjusted dollars, regardless of whatever happens to yields. They will do so if there is inflation, deflation, hyper-inflation, or hyper-deflation. These guarantees are part of the written contract which makes up an RRB.

As a consequence, a liability-matched non-rolling RRB ladder will deliver all of its payments as expected, regardless of future yields, inflation, or deflation, guaranteeing that there will not be any withdrawal deficit.

If there is deflation in the future, RRBs will still deliver exactly what they promise: inflation-adjusted coupons and par values.
8Toretirement wrote:A measure of risk is volatility, RRB's are much more volatile, therefore risky, compared to a laddered 5 year nominal bond portfolio with a short term duration for the FI component, supplemented by GIC's. Not against RRB's, just not buying the 50% RRB portfolio as decreasing risk. Duration is a measure of risk. And by this measure alone I would be leery of using RRB's for a large component of my portfolio in the withdrawal phase.
An RRB ETF with an average duration of 16 years will obviously be more volatile than a 5-rung rolling nominal bond ladder with an average duration of 2.5 years. Doh!

As for risk, as defined (not as measured!) by Zvi Bodie (remember the thread's subject!), we can say with full assurance that a liability-matched non-roling RRB ladder is less risky than a 5-rung rolling nominal bond ladder.

Here is Bodie's definition of risk:
In [i]Risk Less and Prosper[/i], Zvi Bodie wrote:
Investment risk is uncertainty that matters.
I further explained this definition in http://www.financialwisdomforum.org/for ... 23#p585616 and http://www.financialwisdomforum.org/for ... 23#p585623.

Why does the nominal bond ladder have more "uncertainty that matters" than the liability-matched RRB ladder? Because while the nominal ladder is less volatile (smaller short-term nominal value uncertainty), the purchase power of future withdrawals from this ladder is highly uncertain. In contrast, while the RRB ladder is more volatile (bigger short-term nominal value uncertainty), the purchase power of future withdrawals from this ladder is relatively certain. It seems quite obvious to me that the purchase power of future withdrawals matters more than the short-term variations in nominal value of the investment.
Variable Percentage Withdrawal (finiki.org/wiki/VPW) | One-Fund Portfolio (VBAL in all accounts)
8Toretirement
Contributor
Contributor
Posts: 225
Joined: 26 Nov 2014 14:55

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by 8Toretirement »

longinvest wrote:8Toretirement,
8Toretirement wrote:Longinvest provided the formula to show a drastic drop in the overall portfolio value, you do get an increased interest rate; however, with the long durations of RRB's I suspect the diminishing portfolio will not recoup the losses from higher interest rates due to withdrawals. Using a RRB ETF will increase the effect as they continually roll there duration rates. ZRR has an effective duration of 15.9 years, which is rolling not declining with each passing year.
I have provided a mathematical illustration of how using an imperfect combination of ZRR and cash instead of a liability-matched RRB non-rolling ladder merely introduces a very small risk of withdrawal deficit in a worst-case scenario where I unfairly underestimated the returns of both cash and ZRR after the rise in real yields.

I have explicitly stated, in prior posts, that RRBs are guaranteed to pay their coupons and par value in inflation-adjusted dollars, regardless of whatever happens to yields. They will do so if there is inflation, deflation, hyper-inflation, or hyper-deflation. These guarantees are part of the written contract which makes up an RRB.
RRB ETF's have a rolling duration as the fund adds RRB's, creating volatility, and I have provided examples where RRB ETF's have lost 13% value compared to around 4% worst loss during recent history on short term Canadian Government nominal bonds. Withdrawing during this type of volatility can reduce portfolio values exponentially.

If, you hold an individual RRB bought on the secondary market to maturity then the investor will receive face value depending on the index value at the time of purchase. If the index value is lower than the time of purchase then the face value will be of lesser value. However, the purchasing power will remain.

If you have to sell prior to maturity, you could suffer losses based on the relative index values from purchase to sale. Do you dispute this?

There is more to RRB's than meet the eye, specifically, since RRB's have to be bought directly from the broker the investor needs to be able to calculate spreads and real interest to maturity on secondary offerings. Make a mistake and you can suffer a capital loss.

There are lots of warnings out there, by the bond experts on RRB's.

See this article for another perspective. http://www.theglobeandmail.com/globe-in ... le4263112/

I am not trashing RRB's but Bodies insistence on TIPS to cover expenses, when converted to RRB's could leave the investor short if they load up on RRB's and need access to additional funds for any reason.
longinvest
Veteran Contributor
Veteran Contributor
Posts: 3956
Joined: 10 Sep 2012 17:26
Location: QC

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by longinvest »

8Toretirement,
8Toretirement wrote: RRB ETF's have a rolling duration as the fund adds RRB's, creating volatility, and I have provided examples where RRB ETF's have lost 13% value compared to around 4% worst loss during recent history on short term Canadian Government nominal bonds. Withdrawing during this type of volatility can reduce portfolio values exponentially.
The withdrawals would only hurt the RRB portfolio if the duration of remaining withdrawals is longer than the ETF's duration.

As an example, if I wanted to withdraw an equal inflation-adjusted amount of money from an RRB portfolio over 40 years (starting in 1 year), the duration of my withdrawals would be 20.5, higher than ZRR's duration. Any loss in the value of ZRR over the period between now and the time when my withdrawals have a duration of 16 years will be fully recovered from, thanks to the higher yields. Actually, such losses would lead to a surplus if yields were to stay up. (I am assuming that the initial size of the portfolio was calculated as to sustain 40 years of withdrawals, of course; the higher risk is if yields go down during this period, as this could lead to a long-term deficit despite an initial increase in value of ZRR).

Someone willing to withdraw an inflation-adjusted amount should not do so from an RRB ETF (or an RRB portfolio) with a higher duration than remaining withdrawals duration. Doing otherwise would introduce sequence of returns risk, similar to the risk borne by balanced portfolio investors taking a fixed inflation-adjusted withdrawal yearly (the famous 4% SWR method).

If you look through my previous FWF and Bogleheads posts, you'll discover that I recommend against the SWR method. The only situation where I agree to using a fixed inflation-adjusted amount is when it is taken out of a liability-matched TIPS ladder (US) or RRB ladder* (Canada). In Canada, the ladder has rungs maturing 5-years apart (or 3 years apart starting in 2041). A Canadian will need bigger rungs covering all annual withdrawals until the next rung, and will need to use cash instruments (like high-interest savings account and GICs) to fight inflation for up to 5 years.

* Or its ETF+cash approximation, of course.

My recommendation, for making withdrawals from a balanced portfolio, is to use the Variable Percentage Withdrawal (VPW) method. This method eliminates the risk of premature portfolio depletion by delivering return-adjusted withdrawals. The risk is that low or negative returns over a long period could bring VPW withdrawals down to an unsustainable level, when the initial portfolio was not big enough. That's why I also recommended to combine VPW withdrawals with a basic income (OAS, CPP/QPP, pension, inflation-indexed annuity, liability-matched RRB ladder) that is not vulnerable to markets and inflation, so that total income (VPW withdrawal + basic income) never gets lower than the retiree's minimal required basic income.

Combining ZRR with cash to approximate a liability-matched RRB ladder is about keeping portfolio management simple in a context where the rest of the portfolio is subject to VPW withdrawals. Any small deficit on the final withdrawal due to the approximation is likely to be over-compensated by the accompanying VPW withdrawal from the rest of the portfolio. If one cannot take such a small risk, one should definitely use a pure RRB ladder as part of basic income, instead of ZRR+cash.
8Toretirement wrote: If, you hold an individual RRB bought on the secondary market to maturity then the investor will receive face value depending on the index value at the time of purchase. If the index value is lower than the time of purchase then the face value will be of lesser value. However, the purchasing power will remain.
We agree.
8Toretirement wrote: If you have to sell prior to maturity, you could suffer losses based on the relative index values from purchase to sale. Do you dispute this?
Yes, selling prior to maturity introduces what is usually called interest-rate risk; a possibility of loss caused by a depressed market value due to an increase in real yields.
8Toretirement wrote: There is more to RRB's than meet the eye, specifically, since RRB's have to be bought directly from the broker the investor needs to be able to calculate spreads and real interest to maturity on secondary offerings. Make a mistake and you can suffer a capital loss.
Yes, bond math can be tricky. RRB math is even trickier. Bond markets are opaque, too. I personally refrain from buying individual securities because it's tricky and my wife would not be able to maintain the portfolio.

I also think that many people confuse bonds and cash. RRBs are bonds; RRBs are not cash. Between issue and maturity, the market value of an RRB fluctuates.
8Toretirement wrote: There are lots of warnings out there, by the bond experts on RRB's.

See this article for another perspective. http://www.theglobeandmail.com/globe-in ... le4263112/
Yes, RRBs can be risky for investors who don't understand how to use them. But, Stocks are even riskier. So, while we see a lot of warnings in the financial press (almost) everytime someone writes about RRBs, it is surprising not to read such warnings about riskier securities.

Let's just analyze the linked article:
The hidden pitfalls of real return bonds, John Heinzl wrote: RRBs typically have long maturities, which makes them volatile when interest rates rise or fall. The longest Government of Canada RRB matures in 2044, the shortest in 2021.

Such long time horizons are fine for pension funds that have to plan decades into the future, but retail investors' circumstances can change a lot in that time. If they have to sell before maturity, they could be hit with a hefty capital loss if interest rates have risen.
So, we're being rightly warned against the volatility risk of RRBs, when sold prior to maturity. Except that we're not being told that this volatility is actually quite small (thanks to mathematics) for the RRB maturing in 2021! Hmm.

Let's continue with the article:
The hidden pitfalls of real return bonds, John Heinzl wrote: If you're worried about inflation, Mr. [Hank] Cunningham recommends setting up a ladder of bonds with maturities ranging from one to 10 years. Every year when a bond matures, you reinvest the proceeds in a new 10-year bond.
[...]
If investors want inflation protection, fixed-income portfolios are the wrong place to achieve it, [James Hymas] said. They should instead look to other asset classes, such as resource stocks, to counter the impact of inflation on their bonds, he said.
We are being told that if we are seeking protection against high inflation, we should shun RRBs because they are too volatile. We should, instead, invest into a 10-year nominal bond ladder or into resource stocks!

Such an article definitely belongs to this thread: Outstanding Financial Pornography

(James Hymas is a member of FWF. He could have been quoted out of context by the journalist. Maybe he will help me understand, if he happens to read this thread.)
8Toretirement wrote: I am not trashing RRB's but Bodies insistence on TIPS to cover expenses, when converted to RRB's could leave the investor short if they load up on RRB's and need access to additional funds for any reason.
It is true that Bodie fails to address change of plan risk. But, let's be fair, here:

For one thing, Bodie recommends to put one's emergency fund into I Bonds which are inflation-indexed government savings bonds (cashable without penalty after 1 year of holding). It's what I would call an inflation-indexed cash investment with no interest-rate risk for a U.S. investor. There is no Canadian equivalent, unfortunately.

For another, the typical portfolio contains a significant amount of stocks. Let's look at what happens when we mix stocks (XIC) with three kinds of bonds, at a 60/40 stocks/bonds ratio: short bonds (XSB), aggregate nominal bonds (XBB), and RRBs (XRB).

PortfolioVisualizer
XIC/XSB (blue), XIC/XBB (red), XIC/XRB (orange):
Jan 2006 - Oct 2016
returns.png
chart.png
We're talking of a difference of less than 2% in volatility between the portfolio using high-duration RRBs and the one using low-duration short-bonds. In terms of drawdown, XIC/XSB dropped -25.3%, XIC/XBB dropped -26.0%, and XIC/XRB dropped -30.7% in 2008-2009*. It's not the real-return bond losses that were worrisome; it's the stocks losses!

* Note that in my personal portfolio, I have a mix of half nominal, half real-return bonds. 60/20/20 XIC/XBB/XRB dropped -28.3% in 2008-2009.

Nobody is recommending to put short-term cash needs into long-duration RRBs! I don't invest my emergency money into RRBs; I keep it liquid in a high-interest savings account that I can access quickly in case of an emergency.

I think that it is important to keep the relative risks of asset classes in perspective. By any reasonable measure, stocks are riskier than real-return bonds.
Variable Percentage Withdrawal (finiki.org/wiki/VPW) | One-Fund Portfolio (VBAL in all accounts)
User avatar
ghariton
Veteran Contributor
Veteran Contributor
Posts: 15954
Joined: 18 Feb 2005 18:59
Location: Ottawa

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by ghariton »

longinvest wrote:I think that it is important to keep the relative risks of asset classes in perspective. By any reasonable measure, stocks are riskier than real-return bonds.
I agree with what you have written. I would make three additional points.

First, the discussion above generally looks at individual securities in isolation. That is misleading, whether one adopts a two-portfolio approach or a one-portfolio approach. Rather, we should be looking at the risk contributed to the overall portfolio by the security in question. In turn, this means looking at correlations as well. In other words, look at the diversification effects of nominal and real return bonds, as well as volatility, etc.

Second, on protecting against inflation with ladders of nominal fixed income instruments. Historically this hasn't worked very well. We've had long periods of time, in the four decades after World War II, when all the tranches of a 5-year bond or GIC ladder earned negative real returns. Essentially, ladders of nominal bonds provide liquidity. The hope is that the liquidity will allow the investor to benefit from higher interest rates that are supposed to accompany higher inflation. Unfortunately, the link between nominal interest rates and inflation is tenuous in the short and medium term. Anyway, liquidity has its own price. Why purchase liquidity when what you really want is inflation protection and it can be purchased directly, more cheaply?

Third, while the mechanics of RRBs are (slightly) more complicated than the mechanics of nominal bonds, both are much simpler than the analysis of common equities, which in turn are much, much simpler than preferred shares. The real problem with individual bonds, as noted, is the reliance on brokers and dealers to quote a fair price, rather than obtaining a price determined by open market forces. The bid-ask spread is usually a pretty good indicator of what you are getting into. If the spread is less than 3% on a 10-year bond you are intending to hold to maturity, that's 15 basis points per year. Compare with the MER on the ETF of RRBs to see which is the better choice.

George
The juice is worth the squeeze
8Toretirement
Contributor
Contributor
Posts: 225
Joined: 26 Nov 2014 14:55

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by 8Toretirement »

ghariton wrote:
longinvest wrote:I think that it is important to keep the relative risks of asset classes in perspective. By any reasonable measure, stocks are riskier than real-return bonds.
I agree with what you have written. I would make three additional points.

First, the discussion above generally looks at individual securities in isolation. That is misleading, whether one adopts a two-portfolio approach or a one-portfolio approach. Rather, we should be looking at the risk contributed to the overall portfolio by the security in question. In turn, this means looking at correlations as well. In other words, look at the diversification effects of nominal and real return bonds, as well as volatility, etc.

Second, on protecting against inflation with ladders of nominal fixed income instruments. Historically this hasn't worked very well. We've had long periods of time, in the four decades after World War II, when all the tranches of a 5-year bond or GIC ladder earned negative real returns.

Third, while the mechanics of RRBs are (slightly) more complicated than the mechanics of nominal bonds, both are much simpler than the analysis of common equities, which in turn are much, much simpler than preferred shares. The real problem with individual bonds, as noted, is the reliance on brokers and dealers to quote a fair price, rather than obtaining a price determined by open market forces. The bid-ask spread is usually a pretty good indicator of what you are getting into. If the spread is less than 3% on a 10-year bond you are intending to hold to maturity, that's 15 basis points per year. Compare with the MER on the ETF of RRBs to see which is the better choice.

George
George, Longinvest, we are tackling the issue from different ends of the spectrum, with some similar ideas.

I think I have more faith the BoC will keep inflation close to their 1-3% control band, this I can handle, any spikes will be covered by the equity component and delay of government pensions, CPP, OAS.

I use Ishares 1-5 year laddered government bond for part of my fixed income. The duration is 2.99 yrs. Also a 1-5 yr GIC ladder, and an ultra short bond ETF with a duration of 0.5 yrs. Fixed income in my portfolio is a safety anchor for the equity component of my portfolio. I believe we became crossed with the discussion on fixed income products by bringing in other asset classes, when the discussion is proxies for FI. I am not big on RRB's but can see their addition in a portfolio under the FI component for advanced investors. The product is mostly used by pension funds and insurance companies to match their long term liabilities.

There is risk in using RRB's if the investor is not fully versed in bond valuation since to get the full value of the product you would have to match maturations of single bond issues with income requirements. Buying these single RRB's can only be done directly through the bond desk, which means the valuation must be determined from their quote.

For the average investor this level of complexity could lead to unexpected capital losses, especially if the portfolio construction doesn't take account of the specific intricacies of RRB's.

RRB ETF's are not the solution as the duration is very long, thus increasing volatility, which in the withdrawal process, will increase risk through quicker depletion of the portfolio. Don't compare 10 year bonds to RRB's, you need to go shorter, especially in this economic environment.

To recap, single RRB's with matching maturities to expenditures leads to increased requirements from the investor to properly value the product or suffer unexpected capital losses. I view FI as the safety component of a portfolio, RRB's are suitable for advanced investors purchasing single RRB's with matching maturities to their income requirements, using them for the specific purpose of guaranteeing purchasing power for their income requirements. There is high complexity with the product if it's to be used properly.

This complexity is a great risk to the average investor.
User avatar
ghariton
Veteran Contributor
Veteran Contributor
Posts: 15954
Joined: 18 Feb 2005 18:59
Location: Ottawa

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by ghariton »

I just don't understand why real return bonds are considered to be more complicated than nominal bonds. The mechanics of each are the same, with the only difference that the RRB uses constant dollars and the nominal bond uses current dollars. All of the concepts -- duration, convexity, yield to maturity, market risk, credit risk, liquidity risk -- are identical, and the measurements use the same methodology. Furthermore the analysis of the one can be translated into terms of the other, merely by multiplying by the appropriate cumulative CPI index.

All of the considerations relevant to choosing between an individual nominal bond and a nominal bond ETF carry over to the choice between individual RRBs and RRB ETFs.

The reasons for choosing between nominal bonds and RRBs, individual or ETF, should be about (1) one's attitude toward unanticipated inflation and the price one is willing to pay for protection (2) the impact of each on diversification of one's total portfolio (3) matching future assets and liabilities. Other factors, such as complexity, should have no role to play IMHO.

George
The juice is worth the squeeze
Flaccidsteele
Veteran Contributor
Veteran Contributor
Posts: 4523
Joined: 06 Mar 2014 12:52
Location: Retired Gen Xer somewhere on the planet earth

Re: Risking Less and Prospering using Real Return Bonds? [RRBs]

Post by Flaccidsteele »

8Toretirement wrote:I view FI as the safety component of a portfolio, RRB's are suitable for advanced investors purchasing single RRB's with matching maturities to their income requirements, using them for the specific purpose of guaranteeing purchasing power for their income requirements. There is high complexity with the product if it's to be used properly.

This complexity is a great risk to the average investor.
I agree with this.

I would go further and say that bonds in general, are more complex than the investors who use them for "safety", realize.
Post Reply