Investing decisions should always be made in the context of your overall financial plan. And although we know short-term forecasts are futile, a retirement plan needs to include some assumptions about returns and risk over the long term. To help with this important task, my colleague Raymond Kerzérho, PWL Capital’s director of research, has just updated our white paper, Great Expectations: How to estimate future stock and bond returns when creating a financial plan.
As we explain in the paper, there are two main approaches to estimating future stock returns. The first is to rely on a historical premium: over the last 50 years, stocks have delivered returns of about 5% above inflation, so one could simply expect that to continue. The second approach raises or lowers that expected premium depending on whether stocks are currently undervalued or overvalued. You can apply similar methods to expected bond returns, using either the long-term premium (about 2.7% over inflation) or the current yield on a benchmark index.
Both methods are flawed, but an average of the two is likely to be a useful estimate. Imagine that you are doing retirement projections going out 30 years. Using an expected return of 4.5% for bonds based on their long-term average seems wildly optimistic. But on the other hand, assuming bonds will yield just 2% for the next 30 years (based on their yield today) seems unnecessarily conservative. An average of these two estimates (3.3%) is a reasonable compromise.
You can dig into the paper for all the details, but here are the numbers we’re using for inflation, bonds and stocks in our plans these days:
Estimated long-term returns (as of December 2015)
| Asset class | Expected return |
| Inflation | 1.80% |
| Canadian bonds | 3.30% |
| Canadian equities | 7.10% |
| U.S. equities | 6.30% |
| International developed equities | 7.20% |
| Emerging markets equities | 9.80% |
Source: PWL Capital
And here’s how those numbers combine in various balanced portfolios. In the table below we’ve also included the standard deviation (a measure of volatility) for each asset mix, and the maximum drawdown (or cumulative decline) experienced in similar portfolios since 1988:
Expected return and risk of various portfolios
| Equities/Bonds | Expected Return | Standard Deviation | Cumulative Decline |
| 0% / 100% | 3.30% | 3.90% | –11% |
| 10% / 90% | 3.60% | 3.80% | –10% |
| 20% / 80% | 4.00% | 4.00% | –10% |
| 30% / 70% | 4.40% | 4.50% | –10% |
| 40% / 60% | 4.80% | 5.30% | –14% |
| 50% / 50% | 5.10% | 6.20% | –18% |
| 60% / 40% | 5.50% | 7.20% | –23% |
| 70% / 30% | 5.90% | 8.20% | –28% |
| 80% / 20% | 6.30% | 9.20% | –33% |
| 90% / 10% | 6.70% | 10.30% | –39% |
| 100% / 0% | 7.00% | 11.40% | –44% |
Sources: PWL Capital, Morningstar Direct
How low can you go?
In this new edition of our paper (which was first published almost two years ago), we’ve added a postscript to help put these numbers in context.
If you’ve looked at the returns of a balanced portfolio over the long-term, you may be surprised (and disappointed) by the expectations we describe in the paper. Even since the late 1980s, traditional index portfolios delivered annualized returns in excess of 7% or 8%, even with a conservative asset mix, compared with our expectation of just 5.1% for a portfolio of half stocks and half bonds. Why so gloomy?
The first important point is that over the last 20 to 30 years, bonds enjoyed a long bull market as interest rates trended steadily downward (10-year Government of Canada bonds yielded close to 10% in 1988). This cannot be expected going forward, so we think it’s reasonable to plan for conservative portfolios to deliver significantly lower returns in the foreseeable future.
It’s also reasonable to expect equity returns to be lower than they have been since 1988. By traditional valuation measures, stocks are relatively more expensive today: for example, the S&P 500 had a price-to-earnings ratio of 14 at the beginning of 1988, compared with 24 at the end of 2015.
Finally, inflation was 4% in 1988, compared with just 1.4% in 2015. The numbers in the tables above are nominal returns, which are not adjusted for inflation. Remember that a 6% return with 2% inflation is very similar to an 8% return with 4% inflation. When viewed in terms of purchasing power, the gap between historical returns and expected future returns is not as wide as it first appears.
Is there likely to be any move in the model portfolios or recommended funds for 2016? I notice you typically do so in the first month or so of the calendar year.
@Dan
Thanks for the update of this paper. This will help me with planning and I appreciated the distinction between a nominal and inflation-adjusted returns.
I also appreciated your article in the recent Moneysense regarding the lack of index ETF content in your CFP and CIM curriculum. I wonder if that would change in the future?
Congratulations on the designations.
@Gavin
I believe that Dan had said in one of his earlier posts (or comments) this year that he won’t be updating his model portfolios for 2016.
The three fund portfolio from Vanguard would remain the same, valuing simplicity, and being adequate to meet to vast majority of investors financial goals.
One could create a very similar iShares portfolio using XQB, XIC, and XAW with the differences between the products being almost negligible.
Best wishes with your investing!
@Gavin: No changes coming in 2016.
@Erik: Glad you liked the piece in MoneySense. I’m sure the educational resources for advisors will improve over time, but we need to remember that the textbooks are created by and for an industry that will always profit from selling skills it does not really have.
Wonderful post as usual Dan! Thanks for the info.
… if only returns were normally distributed. On October 19, 1987 the US Stock market had a drop of over 20 standard deviations. By these estimates of risk, this is impossible within a human lifetime or even over any human-relatable time frame. Events with SDs over 5 are not that uncommon. And within short time periods (<1 hour) you can have movements over 50 SDs.
Having said that, I am still a fan of your website!
I am 28 yo and just started reading a lot about investing. I recently opened TD mutual fund which I will convert it to e-series however I would also like to open the questrade and buy the Vanguard.
After reading and doing some research, I’m thinking to but this following
VIU (new) 15%
VXC 30%
VFV 40%
VAB 15%
I do realize that Canadian market nowadays are shrinking to the bottom so that’s why I am betting my money into US and global market. I feel like am still young and have a long way to save until my retirement, so that’s why I put only 15% allocation on Canadian bond.
What do you think about my portfolio?
Thanks
@Sisi
Are you sure you did some reading and researching? why would you use VIU AND VFV with VXC ? Makes no sense to me and likely most on this blog. why didn’t you just go with 85% VXC and 15% VAB ?
Hello again CP. I love the line in the movie “Simon”. Allan Arkin claims to be quoting Wittgenstein when he says: “You don’t know what you don’t know. Ain’t that great! You don’t know what you don’t know”. I grin when ever I think of the scene.
But what I really wanted to say is that I just clicked on the link above ‘Great Expectations…” It is great article and, I believe, quite accessible. Perhaps you might consider posting the whole thing here on your blog.
You seem tireless–keep up the great work.
jq
@Jake using only 2 ETF is too risky. It’s too much percentage on VXC 85% really?
VFV tracks the S&P 500, it has 5 stars on morningstars and based on my research even when the market crashed during 2008-2009 that still made money.
VIU is the brand new ETF that vanguard just launched and I am interested in betting my money on these. It’s only 15% so not a lot.
@Sisi: With respect, your research is inaccurate. VFV did not exist in 2008-09, and the S&P 500 was more or less cut in half during that downturn.
As for two ETFs being too risky, in what sense do you mean? VXC holds US, international and emerging market stocks. VFV holds US stocks and VIU holds international stocks. So holding all three is redundant and does not reduce your risk in any way.
My model portfolios seem simple but they are carefully thought out. My suggestion is to simply follow one of those for now rather than trying to design something different.
To confirm, the “Expected Return” column is nominal returns, right? Adjusting for inflation, using a 1.8% figure would result in real returns of 1.8% less. For example, a 50/50 portfolio would be expected to return 3.3%/year post-inflation.
@Sisi: VXC includes the US market also — 55% of the fund is the US market, according to the fund’s Portfolio Data page. If you also buy VFV, you may be getting more US exposure than you want. Check out this site’s model portfolios for some good samples. Also, note that VFV is an S&P 500 index fund but it is priced in Canadian dollars, so the fund’s return is combination of the actual performance of the S&P 500 and the performance of the US/Canadian exchange rate. Dan’s posts at Canadian Couch Potato are a great learning resource. I have learned a great deal reading through posts here.
@Tim: Yes, the expected returns are nominal (not adjusted for the 1.8% assumed rate of inflation).
Since a couple of people brought up VFV I have a question regarding VFV vs. VUN. I understand that VFV tracks the S&P 500 and VUN tracks total USA market…
…however according to the Vanguard site VFV has an MER 0.07% while VUN has an MER of 0.15% and VFV has a higher distribution. Am I wrong in assuming that VFV is a better choice given the lower MER and higher distribution?
Thanks again awesome people!!
@Murray: Nothing wrong with VFV, but it holds large-cap stocks only. Over the long term it is reasonable to expect that the inclusion of mid- and small-cap stocks in a fund like VUN should result in returns more than 8 basis points higher. The S&P 500 also has some flaws that can be avoided with a broad-market index:
https://canadiancouchpotato.com/2015/10/30/are-index-funds-fatally-flawed/
I appreciate the article on this topic and used it to play with the numbers in the “die Broke” spreadsheet I use to (help) guide me. However, I guess I’m being a little thick as I don’t understand the “cumulative decline” column. Can you provide a brief explanation? Thx
@John: The cumulative decline is the largest loss in any 12-month period. The idea here is to help investors appreciate that standard deviation doesn’t tell the whole story of risk. Even if annual losses fall within two standard deviations 19 years out of 20, one should be prepared for losses that are much greater than that.
Great – I’ve got it!
Hey Dan, I am setting up my register account and wanted to know what you though of what I have structured for my portfolio. I’m in my late 30’s and plan on retiring in my early 50’s. I already am invested in my work pension fund and maxed my TFSA. Here’s what I’m setting up for the RRSP:
XIG 12.5%
VAB 12.5%
HXT 25%
VXC 50%
Any thoughts?
Dave:
As a derivative-based product, HXT might make sense in a taxable account but it is narrowly based (TSX60). For an RRSP, I would suggest the far broader VCN to get mid/smaller cap exposure. Also, I’m not sure a hedged US Corporate Bond index is really necessary, but others may be able to offer more insight.
Congrats on the workplace pension and maxed TFSA – you’re well on your way. What have you got in your TFSA?
My TFSA is set up as follows
25% HXT
25% HXS
25% VEF
25% XQB
@Dave: I’m only a relative beginner, but the HXS choice stands out as being inappropriate in a TFSA. HXS uses a swap arrangement to transform the US dividend yield into an equivalent rise in the share price, sparing a taxed portfolio from taxes on ongoing dividends, and only being taxed on capital gains on eventual selling. This would be ideal for someone investing in the US stock market who has used up all RRSP and TFSA allocation room. But the MER contains a component that in essence pays for the administration and other costs of this swap. You don’t need this unnecessary complication in a TFSA (all returns are tax sheltered), when plain vanilla Canadian dollar S&P 500 ETFs are available at much more reasonable cost.
(The HXT costs are so low, on the other hand, that you might as well use it to represent your Canadian Equity component, even if the swap arrangement doesn’t benefit you one way or another.)
This article might be of interest to CCP readers – http://www.economist.com/news/finance-and-economics/21695552-consumers-are-finally-revolting-against-outdated-industry-tide-turns
Hi CCP, I hear alot about ppl using mortgage investment corporations as sources of stable income. Since the income is in the form of debt, is this a suitable alternative/supplement to bonds going forwards? I gather a few of them are listed in the TSX so I assume they have stock like volatility. There are not very many of them as well making diversification a challenge. I’m interested to know your thoughts.
Another question is whether you think it’s reasonable to take a HELOC or secured LOC to leverage seeing as interest rates are unlikely to skyrocket in the next few years. I know in general you’ve said you are not a fan of leveraging to invest. Does this apply if you have a stable salary and can afford to pay down the LOC in a short time frame if necessary?
As always, thanks for your invaluable posts!
@Jon: I have not researched mortgage corporations and can’t add anything too useful to the discussion. In general, if they pay higher yields than government or corporate bonds they will also come with additional risks, though these risks may not be obvious.
I still don’t recommend leverage, though as you say, in some specific circumstances it may be a reasonable strategy. For example, borrowing a small amount to top up an RRSP when the you are in the top tax bracket. Overall I think few investors have the stomach to pull it off with discipline.
Hi CCP
Just to add a little more about Jon’s idea of leverage. What do you think about leverage for divided Etf like XEI. The divided is about 5.5% which is 3% higher than the borrowering interest rates. If I can keep the loan long enough,say 30 years, the likelihood of losing the principles is minimal. while at the same time, the dividend will give a decent return even with carrying costs. This is also in favor of tax returns. Dividend credits can be taken advantage of, and carrying cost is tax deductible.
Thanks for your thoughts.
Hi Dan,
Thanks a lot for this paper, it is really helpful. So from what I understand, these expected returns are based on a 30 years horizon, is it right? Is it this horizon that explains the difference between your value of 6.3% for the nominal return of U.S. stocks and the 4% value which is currently anticipated by Bogle for the next decade ( http://www.morningstar.com/cover/videocenter.aspx?id=718639)?
Can you give more details about the method you used to estimate the returns based on market conditions (using Shiller CAPE?) In particular, how do you get the Shiller CAPE for equities outside the U.S. market and where is the horizon entering in your calculations?
Thanks!
@Herbert
Regarding those dividend and bonds ETF that seem to provide yields above 3.5%…
Check the fund’s breakdown of the distributions and you’ll find that the column labeled Return of Capital is not empty. Meaning that the fund gives back a portion of what you put in. Essentially, the yield gets inflated due to that.
IIRC, XTR yields around 5.7% but that’s only due to the ROC. Remove the ROC and it just yields like 3.6% from the underlying stock’s dividend payouts.
Check to make sure that XEI isn’t similar as XTR.
Thanks CarlieFo
It looks like the XEI has the eligible dividend about 5% then it distributes ROC once a year on top of that.
I am new to this, please enlighten me if you think otherwise. Here’s the link:
https://www.blackrock.com/ca/individual/en/products/239846/ishares-sptsx-equity-income-index-etf
Great stuff as usually!
@Pierre-Luc: The expected returns are actually based on a 50-year horizon. The data for the white paper were complied by my colleague, Raymond Kerzhero, using Morningstar Direct, so it’s not freely available. Please email me directly if you would like more details.
@Herbert: In theory, the argument for leverage makes sense for all of the reasons you mention. The main argument against it is that most investors lack the discipline to execute the strategy for long periods. In 2008-09, for example, dividend funds fell my as much as 50% in about six months. Imagine how that feels when you’re investing borrowed money. It can work, certainly. I just feel that most investors are better off avoiding it.
Hi Dan,
Just wanted to get your opinion my RRSP portfolio below:
10% VAB
30% VUN
30% VCN
20% XEF
10% XEC
Do you think VXC would be a good substitute for XEF/XEC/VUN?
Thanks
@nhr594: This is very similar to my models. Substituting VXC for XEF/XEC/VUN would give you very similar exposure with fewer moving parts, but either option is fine.
okay thanks. I guess I should just keep what I have in XEF/XEC/VUN instead of selling them and replacing it wit VXC right? For future investments, I am thinking of just buying VXC for the reason you mentioned below (fewer moving parts).
Thanks Dan for all your insight. I have a small account of $65k so cost is an important factor and what burns me is that RBC Direct tells me that Vanguard ETF’s are ineligible for DRIP Plans in my RRSP. Are there any alternatives ?
@Ming: Yes, that is a frustration with RBC Direct and Vanguard ETFs. Personally I don’t see DRIPs as that important, especially on a relatively small account where you may only receive one or two shares at a time anyway. But if RBC will DRIP iShares products you could use the ones that most closely match the Vanguard ETFs in my model portfolios, such as XIC, XAW and XBB or XQB.
Good report, I think most people still have an unrealistic expectation of returns. I had a client just recently that said he wouldn’t be happy with anything less then double digit returns! The above is a reasonable expectation to hopefully bring investors back down to reality!
It doesn’t say in the paper, but it appears that a rate of return of 7.6% was used for the international component. I think this works out to 85% international developed and 15% emerging markets.
For those like me without emerging markets (I have XEF) the 60/40 portfolio rate of return is 5.44% That’s just a little lower but presumably a little less volatile.
Thanks Dan and Raymond for this interesting article and paper!
Hi Dan,
Regarding asset location, does it not make more sense to hold US and International equities in a TFSA and bonds in an RRSP given the higher expected returns of the equities? The RRSP will eventually be taxed as income whereas the TFSA will never be taxed. Your asset location article from 2013 seemed to advocate holding bonds in the TFSA and US/International equities in an RRSP, wouldn’t this lead to higher overall taxes down the road? Maybe I’m missing something…
@Landon T: I don’t think you’re missing anything, and I would agree with you in most situations. It’s difficult to make hard and fast rules about asset location, but these days we are indeed using most of our clients’ TFSAs to hold equities while keeping low-growth assets such as bonds in RRSPs for the reasons you’ve outlined. An exception might be if there was an expectation the investor might need to pull some money out of the portfolio for down payment or other major expense. In that cases, bonds in a TFSA is an option.
That said, we also frequently put US and international equities in RRSPs, since there is less drag from foreign withholding taxes when US-listed ETFs are used. Remember that TFSAs are still relatively small, so investors with significant portfolios won’t be able to keep all of their equities in tax-free accounts.
hey there couch potato.
would you ever consider adding an emerging market index to your e-series model portfolio? just curious and if so which one?
@Amus: Unfortunately there is no emerging markets e-Series fund.
Hey CCP, great post and thanks for the read.
I personally allocate 0% to bonds with market interest rates being to low right now. Instead, I allocate my “safe haven” money to REITs because the yields are way better and they demonstrate similarly low beta compared to the stock market.
Again, good writing!
FC
Hi Dan.
Last quarter I opened my Locked in RSP account when I left my previous government job. I have the portfolio setup as 70% stock and 30% bonds (XEF, XIC, VUN, VAB & VEE). I have only so far used about 50% of the total account holdings for this ETF portfolio and I’m keep the other 50% as cash and trying to spread out the buying power throughout the next year or so. I guess I’m worried about purchasing everything at once and trying to purchases the ETFs as they fall throughout the year. Not sure if this is a good strategy or not? Your thoughts?
Also, I figured instead of keep 50% in cash what did you think of XTR ETF. It pays great dividends, at least that way my cash as some return if I end up not adding a lot to my above portfolio.
Thanks, and great website.
@Paul: Investing gradually is a reasonable strategy if you are very nervous about entering the market all at once. But keep in min that your pension assets were fully invested before you took the lump sum, so you are not really starting from scratch. And of course there is always the risk that you will simply be buying in at higher and higher prices, or that you’ll continue to be nervous and not buy in at all:
https://canadiancouchpotato.com/2013/05/31/does-dollar-cost-averaging-work/
Whatever you decide, using XTR instead of cash is certainly the wrong approach. The fund is about 60% bonds and 40% stocks, so that will be your exposure. The yield is irrelevant. If you want to be 60% bonds and 40% equities, then just build your portfolio with the allocation and gradually move toward your 70% equity target.
I am curious whether you will be updating this article in light of the economic events of 2020. Might be a timely piece.