Are RRSP's the best way to save for retirement? Part 1 of 3

You may be surprised to learn that there are many people who absolutely detest the idea of RRSP’s. I remember a meeting with a senior citizen, well into retirement, who became visibly red in the face with anger when explaining to me how much less he had to spend in retirement thanks to the tax laws surrounding RRSP’s and RRIF’s. And he is not alone.

ChasingTheMarketsOrRisk.jpgI don’t want to scare you: RRSP’s can be great for some/most people. BUT, if not planned properly, you can end up shooting yourself in the foot come retirement time. If anything, this section is designed more to get you thinking about how well you plan your finances and the future (as best as one can given the long time frames and uncertainty of capital markets and tax laws, etc.).

So let’s begin by taking a sample Canadian. Let’s assume our test subject, Anna, is 30 years of age today (2007) and earned $50,000 last year. She has decided that this year, and from now on, she will save 10% of her gross income for retirement which she plans on taking at 65. I’m going to assume inflation is 3%, but her salary increases at 4% per year (i.e. she is getting a raise over and above the increase in the cost of living in other words). Consequently, her savings will also grow at 4% per year in line with her wages.

I’m going to be comparing the same investment portfolio in all cases, but while the "tax identity" of the returns in the RRSP is insignificant (since the RRSP is a tax shelter), it IS significant from a taxation point of view in non-registered investment accounts. The long term rate of return is 8%, which is broken down as follows: 2% Interest Income, 1% Dividend Income, 2% Annual Realized Capital Gains, 3% Deferred Capital Growth. The portfolio has a Standard Deviation of 12%.

Standard Deviation is important to know if you are using Monte Carlo Sensitivity Analyses since you can use them to model the probabilities of outcomes with portfolios that actually behave like real life portfolios – in other words, I’m not going to assume that the portfolio returns 8% per year, every year. The standard deviation tells you a bit about the return pattern you can expect from a portfolio if you also have a good idea as to it’s long term rate of return.

For example, for Anna’s portfolio that has an 8% expected rate of return and a 12% standard deviation. That means the annual return in any given year will lie within ONE Standard Deviation (which is 8% plus or minus 12%) approximately 68% of the time. Another way of saying that is "68% of the time, the portfolio’s annual return will be between -4% and 20% (which is 8% plus or minus 12%). Further to that, roughly 95% of the time the portfolio’s annual return will be within TWO Standard Deviations (which is 8% plus or minus 24%). 99.73% of the time, the annual return will be within THREE Standard Deviations (8% plus or minus 36%).

Okay, so we’ve set the stage – let’s get on with the topic at hand! :)

Case 1: Anna Saves 10% to Her RRSP – Doesn’t Use Tax Refund Productively

As a base plan, I’m going to assume that Anna saves her 10% to her RRSP account, but does NOT use the tax refund productively. And I mean that purely from a "retirement planning" point of view, since if you spent your tax refund on an annual vacation down south, it can do wonders for the soul! :)

Based on this set of parameters and taking into account current tax legislation with respect to CPP and OAS, etc. Anna can expect to have $43,700/year after-tax, and in today’s purchasing power in retirement from age 65 to age 90. Her probability of success when incorporating her portfolio’s standard deviation is 76%. This is graphed below:


Let’s now compare this to saving to a non-registered portfolio instead of her RRSP…

Case 2: Anna saves 10% to a Non-Registered Portfolio, There is no refund to use

In this case, while Anna is no longer taxed her full marginal tax rate on all her withdrawals at retirement (she will only be taxed on the growth), she WILL have to pay some taxes along the way. If she is able to pay the annual tax bill as she is saving (which will get quite high closer to retirement) then she is in better shape AT retirement since her annual after-tax income in today’s dollars is a whopping $51,300/year – that’s an increase of $7,600/year! Her probability of success is 74%. The results are graphed below:


There is one noticeable difference right away just from eye balling the graphs. You’ll notice the overall distribution of the various "trials" are more squished together. You’ll also notice that there are some more outliers in this graph (by outliers, I mean data points that are extremely unlikely, however possible). This serves to increase the scale of the Y-Axis – so if you compare the graphs with the same scale – they are actually more similar looking. This just happens to be an anomaly of this particular set of trials. I am only using 100 different iterations. To be more academic, I would have to use something like 10,000 iterations, but if I do that every time, my CPU would have melted by now. :)

In any case, we have just seen that by saving money to her non-registered portfolio instead of her RRSP, Anna would have more money to spend in retirement – and by a healthy margin. So how important are tax shelters if the withdrawals are taxed as income? Certainly something to consider.

Remember though, there are a few points to also consider that may affect the feasibility of this approach: the annual tax on the non-registered portfolio will grow and grow and we are assuming that Anna is paying for this out of her employment income up to retirement (after she retires I am fully factoring in all taxes on the non-registered portfolio).

We also haven’t looked at what happens if we do indeed use the RRSP refunds productively. In Part 2, I will examine two other scenarios. One in which Anna is using her RRSP refund to contribute to a non-registered portfolio each year, and a second where she is taking the RRSP refund and putting it right back into her RRSP.

Part 3 will look at comparing the probabilities of each scenario against a fixed level of income (since each on their own will produce different levels of annual income, a comparison of probabilities of success for each will not be an apples-to-apples comparison). Part 3 will also include some commentary on the psychological aspect of each method as well as some other points to consider. I hope I have your interest piqued so far! :)


If you found this article of interest, please <

a href="" target="_blank">consider subscribing to my RSS Feed. If you want to learn more about what an RSS Feed is, click here.

For special deals for readers of (that’s you!), please visit the "Deals For Readers" section

Preet Banerjee
Preet Banerjee an independent consultant to the financial services industry and a personal finance commentator. You can learn more about Preet at his personal website and you can click here to follow him on Twitter.
Related Posts
Showing 0 comments
pingbacks / trackbacks