Martin Gale form Efficient Market Canada just wrote an amazing piece, laying out the case for RRSPs over non-registered investment in a growth stock, a dividend stock, and a mortgage payment. This article is a MUST READ for anyone who has every wondered if investing outside an RRSP has some advantages over investing inside an RRSP. Or if you have ever wondered if you should pay down your mortgage rather than investing in an RRSP. Here’s the set-up:
Are people really better off saving money outside of an RRSP? Some say buying and holding growth stocks that pay no dividends is another way to defer taxes, and that investments in an RRSP lose access to the dividend tax credit. Others say that it’s better to repay your mortgage than invest in an RRSP There are a lot of people out there who say things like this and they are almost always wrong. The arguments for non-registered investments generally involve a lot of handwaving and grand claims so let’s break it down and look at the numbers. First we’ll look at buying stocks in and out of an RRSP, then we’ll compare an RRSP to a mortgage payment.
This article is going to be a straight-forward proof that investments in an RRSP beat investments outside of an RRSP in terms of expected returns.
A couple notes about his calculations… As he says, “Note that if anything these assumptions bias against the RRSP because for many people the tax on withdrawals from an RRSP will be lower in retirement than the tax saved at contribution time.” He goes on to give a good reason for why this is usually the case. Another reason these calculations are biased against the RRSP is because he neglects the fact that the investments outside the RRSP might be taxed every year if you are selling your investments every year and buying new ones, for example, whereas inside the RRSP your investments will experience tax-free growth. I just did a quick back of the envelope calculation and this does make a difference.
In the end the RRSP wins out in each case, even though some simplifications in the calculations bias the calculations against RRSPs. The key reason the RRSP comes out ahead, as he explains so well, is because of double taxation outside the RRSP. Taxation on your income (off your paycheque), then taxation on the gains made off those investmentes. In the RRSP you are only taxed once, when you withdraw the entire thing, your gross income plus gains from the RRSP as income.
Almost everyone who argues that the non-RRSP is better ignores the fact that the non-RRSP payment was made with after tax dollars, in other words, they ignore the double taxation.
Around this same time last year I looked at a Phillips, Hagar, & North report that compared investing in an RRSP with investing outside an RRSP. It was called “The Retirement Savings Debate: Inside or outside the RRSP structure” and their conclusion was that
saving for retirement using a registered plan (RRSP) is more beneficial than saving in a non-registered, taxable account. There are a few exceptions to this, but for the most part, this conclusion will hold true for the majority of middle- and upper-income earners.
If anyone finds that PH&N report, let me know. That blog post also had a quote from Derek Foster, which after looking at it again, now seems like total bunk.
I also wrote about paying down student loans vs. contributing to an RRSP although I neglected a few things (see comments).
Another rrsp advantage that his example doesn’t show is the fact that the average tax rate paid on the withdrawal in retirement will almost certainly be significantly lower than the average tax saved on the contribution, even if the marginal tax rate is the same.
ie (ignore inflation) someone making $100k for 35 years and contributing $10k per year – saves $4338 in tax each year for an average tax saving of 43.38%. During retirement this same person withdraws $100k per year (it’s just an example!) and pays $29,374 in tax for an average tax rate of 29.374% which is a lot lower than the original 43.38% saving.
I think this is an advantage?
Holy crap you are right Mike. I never realized that before. You are always contributing at the marginal rate but withdrawing at the average rate. Thanks.
One thing I would correct in the RRSP over anything debate would be to use the 1000 for the non registered and use 1000 tax savings. This better reflects the real transaction of getting your refund and reinvesting it which is the only way RRSP work. Does anyone know a site that does different calculations based on different rates of return, tax rates and time? Maybe some math wiz could solve these.
“This better reflects the real transaction of getting your refund and reinvesting it which is the only way RRSP work.”
Not necessarily the only way RRSPs work. If you make your RRSP contributions every year in February, deduct the contribution when you file, then get your refund in April and shove the refund into your RRSP right away, that closely approximates the situation described in the Martin Gale article. You are correct though, that this assumption is something that would bias the calculations slightly in favour of the RRSP, although I think the difference would be ever so slight.
I’m pretty sure I’ve done an involved example on a spreadsheet before. But I think Martin Gale’s calculations are better. He makes proper, sound assumptions (that mostly bias the calculations against an RRSP, the winner in the end) to make the calculations easier to understand and follow. If you make the calculations too complicated there is more of a danger in making an error.
There are opinions all over the place on these calculations.
A study published by AIM Trimark states that RRSP only come out ahead if the refund is also invested. IF the refund is used for other purposes then the un-registered investment is a better choice.
RRSP vs. Non-Registered.pdf
This is what Martin Gale also states — you have to invest the whole $1000 (the $640 you have in hand, plus the refund you get).
David Ingram states that a mortgage reduction is a better choice as the interest that you never pay due to your pre-payment becomes savings at your pre-tax rate, i.e. you never have to earn the money to pay the interest.
http://www.centa.com/CEN-TAPEDE/2003/expert/expert119-01.html
As always, YMMV!
David
David,
Regarding the AIM article, I must be missing something. How can a taxable account come out on top of a non-taxable account if you’re putting the same amount in each year and have the same rate of return? Even if you don’t assume that the taxable account is fully taxed each year, at best it can match the non-taxable account’s ending value (no realized tax).
I guess I’d like to see the full set of assumptions they use in that article.
Investoid,
The AIM article makes three comparisons: an un-registered portfolio, a RSP with tax return spent, and a RSP with tax return invested. The article is further discussing the amounts you would have in your wallet at the end of the period, not the amount in the investment account. Since the various instruments are taxed at different rates (i.e. dividends or capital gains vs. interest or income) between the account and your wallet, that explains the value in the un-registered account.
Please be aware that I am not a financial advisor, and thus offer an opinion, not advice!
David
Here’s some more fodder. Sterling Mutuals published this study in November 2000:
http://www.sterlingmutuals.com/Research/Monthly/account_nov00.htm
It excplains the differences between investment choices in greater detail.
Please be aware that I am not a financial advisor, and thus offer an opinion, not advice!
David
question: why do say that the derek foster quote is bunk?
David,
That was a great post!
I have to agree with the number logic that for most people in middle to upper class should go with an RRSP.
The only problem of course is knowing when you’re an exception to the general rule such as low income or those using a family trust.
CD
David, thanks for the AIM link. Totally makes sense and thanks for bringing up the fact that the refund must be invested, not squandered. This is implicit in Martin Gale’s calculation, as you pointed out “This is what Martin Gale also states — you have to invest the whole $1000 (the $640 you have in hand, plus the refund you get).”
ace: He’s basically got taxable dividends from a $100,000 non-registered investments (assume they are invested and compounded) competing with $6000 in an RRSP (compounded) that is taxed on withdrawal. Basically if you do the math, it can work out to be the same or different but it depends a lot on the assumptions you use. I might play with it some more. There is risk involved with buying up $100,000 in blue chips with borrowed money. Who knows maybe I’ll try it someday, but for now it’s on my list of things not to do.
I said it was bunk because it isn’t exactly good advice suitable for general audience on whether or not to use RRSPs.
Ace: a couple of other considerations when evaluating the “Derek Foster” method is that as I recall he never made much income in any particular year when he was working so living a frugal life would probably be a natural extension of that fact. Also – if you don’t make much money then you don’t pay much tax and the incentive to contribute to an rrsp is much less (possibly zero). I don’t think there is anything really wrong with his strategy in his situation but as Dave said, it obviously doesn’t apply all that well to the general public.
What about the effects of annual fees for the RRSP. In the example there was about a three hundred dollar advantage for the RRSP and that could easily get wipe out by the annual admin fees
Kevin, here’s what my strategy would be. Start with an RRSP that has no fees then go to a self-directed account once you meet the minimum amount for the fee to be waived.
For example:
TD has mutual fund RRSP accounts that have no fee. You can buy TD Index eFunds which have very low fees. Once you have $25,000 in an RRSP you can switch to E*Trade as E*Trade waives fees for those with over $25,000 if I remember correctly. Or you can move to TD Waterhouse or any of the other places.
The PHN report can be found here.
I am continuing this debate in a post in the near future, so I’ll put my comments there.
My post will focus on the RRSP versus mortgage debate, so I’ll comment on RRSP versus taxable portfolios here. Martin is right about RRSPs being better than taxable portfolios for almost everyone not in the lowest tax bracket. People generally underestimate the corrosive nature of taxes and don’t appreciate the importance of RRSPs in deferring taxes on investment gains, dividends and interest. Warren Buffett calls unpaid capital gains as a free loan from the IRS. Imagine what he would say about a vehicle that lets you defer ANY tax, even the initial capital!
Hello,
You can find the PH&N Report at the following address:
https://www.phn.com/Portals/0/PDFs/ArticlesandAdvice/Publications/retirement_savings_debate.pdf
It’s a good read.
Regards Everyone.
Mark, thanks for the updated link. I’ve updated it on my original post.