Pooled Retirement Pension Plan Draft Legislation Revealed - Will PRPP Help?

Last Thursday November 17th, the Federal government announced and tabled draft legislation (Bill C-25) for creating Pooled Registered Pension Plans.

Will this help the target group - people who have no company pension and are not voluntarily saving through RRSPs or TFSAs? The answer is a mild "yes" in absolute terms and a resounding "no" in relative terms.

Yes, it does help somewhat.

• Anything is better than nothing - The most critical problem is that those people right now are not saving for retirement, so any program that gets them to save will help. Despite the provisions of the PRPP legislation that do not require companies to opt in and that allow individuals to opt out, the default auto-enrollment rule (par.39(1)), the default investment option rule (par. 23(3)) and a default contribution rate (par.45(1)) will be quite effective in getting more people to save. "Nudge" works.
• Cost and fees have a fair chance to be less than for mutual fund RRSPs - Why might there be a grain of truth to the confident assertion by Minister Menzies who told the Globe and Mail, subsequent to receiving financial industry assurances, that management fees will be "substantially less" than they are for RRSPs? 1) Money will be locked into the PRPP unlike the RRSP. You may be able to switch between funds but the fact that the investment management company knows the money will not be withdrawn means less need for cash balances and less urgency for short-term return-chasing by funds that undermine returns. 2) Marketing costs, which are embedded into management fees could well be less since the target market for the financial industry is not the individual consumer but small and medium-size companies. That should mean no glitzy TV ads. There will be no trailer fees to salespeople (aka financial advisors); in fact, the draft bill specifically bans kickbacks by fund companies to employer sponsors (par.33). 3) The regulatory structure should help to control the most egregious overcharging. Good 'ole politics might even have a bearing since the Governor in Council, i.e. the government, gives itself the power to decide what "low cost" fees means (par.76 (1) (j)).

No, the PRPP will NOT help, it may even do harm.

• NoHype Investing author Gail Bebee emailed me her comments, which I think are spot on, so I'll simply reproduce them with my highlighting):
  

  "1. Employers are not required to offer this, or any other, employer-sponsored pension plan.

  2. Employees will be able to opt out at will.

  3. The financial industry, the same folks who charge Canadians some of the highest mutual fund fees in the world, will be managing the pension funds and will have a major say on the fees charged to do so.

  4. More government bureaucracy will be set up to regulate this new program.

  5. RRSPs already offer a similar retirement savings option. The issue is that not enough Canadians participate."

• Wealthy Boomer Jonathan Chevreau's comments in the Financial Post gave some nuance to Gail's points.
• The PRPP will complicate and confuse - The retirement landscape is already tough enough to understand, what with RRSPs, Defined Contribution and Defined Benefit Pension plans and TFSAs. Another layer of complexity is added. How will people make intelligent informed decisions about which to choose under what circumstances? There is no advice-giving component included in the PRPP structure. Blogger Preet Banerjee was right on the mark pointing this out in a CBC article on the announcement. With each province being required to implement its own complimentary legislation, it's a sure thing Canada will add another patchwork of permutations and combinations in rules, all of it totally unnecessary. What will happen when people move to different jobs in different provinces or with different companies? The minister says the plans are portable and transferable but sure as to betsy a lot of folks will end up with several plans. I already have two different LIRAs that cannot be combined (one is federal, the other provincial) on top of an RRSP and a TFSA. Another possibility to add?! Gimme a break!
• Low pay workers will likely get scr***d - Tax-wise, the logic of the PRPP will work the same as an RRSP. In a couple of posts here and here on the HowToInvestOnline blog comparing the TFSA and the RRSP for retirement savings, it is quite clear that anyone earning less than $37,000 is better off using a TFSA. If such workers get auto-enrolled into the PRPP and they don't opt out (who is to tell them to opt out except for lowly bloggers that they never read anyway?), they will be appreciably worse off in retirement. Thanks for your help, government!
  
• PRPPs pale in comparison to the CPP - The much debated alternative solution, that of expanding the CPP, as we have previously argued here and here, meets the criteria of what is needed much better. Over at Moneyville, author and pension expert Moshe Milevsky points out that the Pooled Retirement Pension Plan doesn't even live up to its name. It doesn't provide a pension - a lifetime of secure guaranteed income - at all, it is only a savings and investment plan that will go up and down with stock and bond markets.
  

Later addition: Commenter Leo's new blog PRPP Canada devoted to PRPP (the blog's mere existence is a sign of the additional complexity Canadians are soon to face) contains a link to lawyers McCarthy Tétrault's review of some of the ins and outs of C-25. From what McCarthy says, the law won't require that there be a default option if the PRPP offers a menu of investment (aka fund) choices that a "reasonable and prudent" person could use to assemble a retirement savings portfolio. That's one less small but critical nudge.

Education Savings Options: RESP, TFSA or RRSP?

Back in 2007 I compared the RESP and the RRSP as investment savings options for funding higher education, excluding the TFSA since it did not yet exist. The conclusion at the time was that the first $2500 of savings should go into an RESP to take advantage of the free money (courtesy of other taxpayers) available from the federal government in the form of the Canada Education Savings Grant. The $500 annual CESG (20% of contributions up to $500 per year and $7200 maximum lifetime - more details here from TaxTips.ca) made all the difference.

It's time to revisit the question. First, the TFSA now exists. Second, an anonymous comment this past July on the original post suggested the RRSP might be better if one takes into account the possibility that the student can transfer up to $5000 in annual tuition deduction during the time of eventual study, which gives the parent a 15% tax credit (i.e. $750) on the tuition transferred. Excellent question! With interesting results too.

I've built a downloadable spreadsheet (look for the download link on the right hand side of the web page once the spreadsheet opens up as a Google doc in your browser) for readers to play with beyond what I have already done.

Conditions Applied to My Analysis:

• Parent Must Have Enough RRSP Contribution Room - The whole analysis presumes you can put in $2500 per year new money plus up to $2100 reinvesting the tax refund each contribution generates, plus the reinvested refund on the reinvested refund, plus the reinvested refund on the reinvested refund on the reinvested refund etc ... (remember that child's song, there's a hole in the bottom of the sea? this is the tax refund version of it); that's why the RRSP part of the spreadsheet extends way out to the right. At the top Ontario marginal tax rate of 46.41% (see TaxTips.ca's tables for personal tax rates in each Province, which readers can use to check what happens in their own bailiwick) that means needing another $2100 or so of extra annual contribution room. Doing this gives the RRSP option its most favourable conditions.
• Only $2500 in Annual Contributions - This condition is to give the RESP its most favourable conditions, namely that it gets the most free CESG money, so that each contribution buck is getting the most bang.
• Child Must Not Have Enough Income to be Liable for Taxes during Higher Education Years - As I noted in the original post, the RESP's advantage disappears if the student has to pay taxes, even at the lowest tax bracket (see the Student_Taxable tab in the spreadsheet).
  

Results: (the summary numbers for the discussion below are in the Results tab and the calculation table with inputs you can use to plug in your own numbers is in the RESP_RRSP_TFSA tab; other tabs contain the calculation tables from the original post)

1) RESP is (Almost) Always Best ... if the Child Takes Post-Secondary Higher Education - No matter what the parent's tax rate, the RESP comes out ahead after tax, as shown by the green numbers. The only circumstance when it does not is when, as shown by the red numbers in the Results tab, the parent's tax rate at time of withdrawal is at least three tax brackets lower than at time of contribution - e.g. taxable income goes down from $100k to $70k as in retirement - and the investments earn a low (2%) to medium (5%) annual return. In this latter case, the RRSP wins, but not by much.

2) TFSA Wins if the Child Does Not Attend Higher Education and Parent's Tax Rate Stays the Same - The green numbers under TFSA show that no matter what tax rate the parent is in and regardless of investment returns, the TFSA does better, but not by a lot, than both the RESP and the RRSP.

3) RRSP Wins if Child Does Not Attend Higher Education and Parent's Tax Rate Drops at Withdrawal - The green numbers in the RRSP column show that the RRSP does better and better the more the parent's tax bracket drops between the date of contribution and withdrawal. The higher the investment return, the bigger the effect. When it is three brackets lower and there are high (8%) returns, the net difference is $25,000 more than the RESP and $15,000 more than the TFSA.

Bottom Line:

• if you are confident that your child will go to college/university before you retire, put that first $2500 into the RESP; if you have several kids, the chances should be higher that at least one of them will go on.
  
• if you really are not sure your child will go on to higher education , then the RRSP is the better hedging option. The TFSA's advantage when the child does go on isn't big enough to offset the TFSA's lower results compared the RRSP when the child does not go on. Also in the RRPS's favour is that the RRSP's disadvantage compared even to the RESP when the child goes on is much less when investment returns are low and you are in the lower tax brackets.
  
• if you believe that your income will drop two or more tax brackets by the time the higher education decision will need to be taken, the RRSP looks better even than the RESP. When there is no higher education the RRSP is always superior to the RESP and even when there is higher education, at two brackets lower you are ahead except at high investment returns. If your investments within the education account are cautious and low risk, low to medium returns are what you will get.
  

CD Howe and Pension Study - The Proper Choice of Reno Tools

Head of the CD Howe Institute William Robson told us in the Globe and Mail last Friday to Fix pensions with screwdrivers, not sledgehammers based on the results of the Institute's newly-published study Sizing Up the Retirement Challenge: How Well are Canadians Preparing for Retirement?. With all due respect, since the study, of which Robson is an author along with Kevin Moore and Alexandra Laurin, is an interesting and worthwhile piece of work, after reading through it, I think we should be saying "Don't patch pensions with the wrong colour paint, cover the whole wall over ".

Through its reasonable, for the most part at least, assumptions the study projects that the likely pension income shortfalls cover such a broad spectrum and such a large and steadily growing number of future retirees that spot fixes like the proposed PRPPs won't do the job.

The study summarizes the future problems thus (my highlighting):

"The principal finding of this study, however – that is, a projected gradual increase in the proportion of future retirees likely to experience a significant decline in their standard of living upon retirement – persists even with differing assumptions for future real wage growth, inflation, rates of return, RPP coverage, and future saving rates." (page 2)

"... the proportion of newly retired individuals unable to replace at least three-quarters of their average pre-retirement consumption from the sources we model is projected to nearly triple over the next 40 years (see Figure 9). If current trends persist, by the 2046-50 period, about 45 percent of workers currently aged between 25 and 30 years would not meet our 75-percent threshold ... " (page 20)

"This decline in potential consumption replacement would be felt across the entire earnings distribution ... " (page 20)

We are just at the start of a long-term uptrend in inadequate pensions that is little affected by such things as increased saving in RPPs, which is essentially what the new PRPPs would offer to people working for small companies and the self-employed. Look at this chart from the study. Other charts in the study have similar inexorably upwards trends based on "business-as-usual" in pensions savings methods/plans.


All this looks like a pretty substantial problem to me.

Ironically, the following quote suggests an alternative:

"... the public pension system, which is mandatory and has nearly universal coverage, provides high levels of consumption replacement to individuals with low pre-retirement earnings. The higher a person’s earnings, the more voluntary saving by the individual (and/or his or her employer) through RPPs, RRSPs, home equity, or other instruments is needed to replace consumption in retirement." (page 13)

In other words, the combination of OAS/GIS and CPP have been doing a great job. When higher income income earners have to rely on the various other means, they fall short.

There is one problem with the public pension option, however. The beneficial influence of OAS will progressively wane. By being indexed to CPI inflation, it only maintains the standard of living at the moment of retirement. The standard of living slowly rises i.e. real wages increase and real consumption does too, faster than CPI does. Since the pension objective assumed in the study is to maintain a level of consumption, anything that goes up only by CPI will produce less and less consumption relative to new retirees. Here is the relevant study chart.

A couple of assumptions probably understate the future pension challenge.
1) Home equity drawdown - the study assumes as base case that 50% of equity in a home would contribute to retirement income. The methods for accessing home equity, like high-fee reverse mortgages, or downsizing, seem unlikely to happen for most people except under forced circumstances. To the authors' credit they run the model excluding that assumption and the difference is another 5% or so today, rising to about 8% in 2050, of the population with less than the target 75% consumption replacement rate.
2) Consumption replacement at age 70 - this number derives from actual government data based on what is used from various sources, including investment return sensitive limited capital RRSPs and RRPs from defined contribution plans. The study does not, that it says anyhow, model consumption to track adequacy all through the retirement years. What happens to the income at age 75, 80 and 85? Chances are it won't go up. Age 70 is likely the maximum income / consumption. One of my relatives has been using a RRIF to supplement her income but it will run out in a few years so she is facing a significant drop in income/ consumption. The model thus does not seem to attempt to evaluate the effect of longevity risk sharing, or not, (running out of money before you die).

One telling point founded on the actual past statistics is the fact that " ... the net real rate of return received by individuals in the future is roughly 1 percent for RRSPs and 2.5 percent for defined-contribution RPPs." That's a shockingly low return compared to the actual historical asset class returns of 4% averaged for a portfolio. The combination of fees and poor investment decisions by individuals trying to manage on their own really cuts deep. If the real return for investors could magically be 1.5% higher, it looks from figure 13 as though another 5% of Canadian retirees in 2050 would escape the inadequate income threshold of the study.

People need to assess the problem properly before deciding on how to fix it. Like a bad paint job, if you don't do it right the first time, you will soon be doing it again.

Senate Weighs in With Some Useful Retirement Savings Suggestions but ...

Canada's Senate committee on Banking, Trade and Commerce announced a half-dozen recommendations on how the government could enhance retirement savings in its Oct.19 report Canadians Saving for Their Future: A Secure Retirement.

The recommendation that would likely have the most beneficial effect is the suggestion to establish a Canada-wide plan for retirement saving and investing. The new plan would entail setting up five or so professionally-managed, competitively-sourced investment funds into which savings deductions/contributions of Canadians 18 and over would go. It's a pretty good but incomplete plan. Why?

1. Auto Enrollment - the report calls the plan "voluntary" but that means an optional opt-out, which few people will do. As the famous book Nudge explains (and as the use of the word in the report slyly suggest that the Senate committee is aware of the idea), the difference between voluntary opt-in and opt-out is huge and participation rates will be as good as universal, up in the 90+% range. Goodbye to the costly sales and marketing overhead cost of retail funds because it's a captive market.
2. Fiduciary Duty Governance and Management and Competitive Sourcing - they call it a commitment to avoid "real and perceived conflicts of interest". Professional managers can add diversification, discipline and net value when the fees they charge are restrained - i.e. the gross investment return isn't sucked dry by the fees. Hello to much lower fees from the powerful negotiating position that such a massive plan will have and hello to a resulting much higher net return to investors with much higher end value retirement savings.
   
3. Optional RRSP or TFSA - it is valuable to have the flexibility of being able to contribute to the right account for one's tax situation / income level (the familiar question about whether your tax rate will be lower in retirement - RRSP better, or whether your absolute income is low - TFSA better) and retirement goal (if legacy desired, TFSA better).

The report does not address a few key issues related to this idea:

• Savings Deduction Rate? - how much should it be? Maybe 9% would do, the same as for CPP, which aims to replace about 25% of pre-retirement income, so such a contribution rate in this plan would provide another 25%. A less desirable method would be to allow the contribution rate to be chosen by the contributor but then the new plan should have a default rate with option to change it (another nudge).
  
• Sequence of Returns Risk - the danger of a market plunge, such as happened in 2008, at the intended time of retirement is that the total available to purchase an annuity is vastly reduced and permanently low retirement income would result. The alternative of withdrawals from a RRIF would see much lower sustainable withdrawals. Of course, nobody would retire after a market crash if they possibly could and they would deal with the market returns risk by continuing to work however long it took for market and retirement savings recovery. That's not the only way to deal with this risk though. The method of the CPP is to have a defined benefit payment coming no matter what the state of the market - did the CPP announce a reduction of payments in 2008 even though its investment portfolio dropped about 20%? The reason the CPP can maintain payments is that it can, as a fund with a very long term investment horizon, smooth out market humps and bumps, knowing that savers continue to provide cash inflow. There is time risk sharing going on within CPP that the Senate's proposal lacks, which to my mind is a very important feature of making retirement saving feel secure and actually be so.
• Conversion to Retirement Income, Inflation Risk, Longevity Risk and Annuities - a retirement savings plan, such as the one proposed, must be converted into an income stream and the report does not consider how this will be done, except for brief off-hand references to buying an annuity. Yet the income conversion vehicle, its cost and its effectiveness in countering inflation and longevity risks determine the success of the whole retirement income exercise. This cannot be considered apart from the savings phase method with the assumption that all will be well. Choose an annuity and even low, normal 2% inflation eats away a huge portion of the value of a fixed payment annuity over the longer and longer retirement periods of today. Real constant-value CPI-adjusted annuities are almost absent from the Canadian marketplace. Most annuities on the market in effect provide income for life at a fast (high inflation) or slow declining standard of living. I bet that's not what people want or need. There is also the problem that the market is lop-sided - the people who want to buy annuities are those who figure they will live longer and not those who will die off sooner and whose cash helps maintain a higher standard of living for the survivors. (Those who believe this is unfair could be reminded that sharing the risk means everyone gets higher payments than if no one shares) The annuity-selling insurance companies know about likely-to-live longer annuity buyers of course, and so annuity payouts are even lower. Contrast that with CPP where everyone, early and late deceased, automatically and without choice to opt out, gets into the annuity payment stream. Choose the other option to generate income, a RRIF from which withdrawals are taken, people have the very hard job to figure out how much to withdraw given their uncertainty how long they will live and need income. Live too long and you run out of money. There is no longevity risk sharing. People can either be very cautious, withdrawing slowly, and perhaps live a much more restrained lifestyle than they might have liked, or they can live high, perhaps to discover that they must drastically reduce their spending later on. Pooled assets with no opt out (i.e. with longevity risk sharing) during withdrawal means higher payments for everyone and much less worry along the way. The prime example of a successful end-to-end solution is the CPP - you pay in a certain amount per year and you are guaranteed (by the most stable provider around, the Federal government) a certain inflation-adjusted amount for however long you live.

The report includes several other worthwhile but less significant suggestions:

• Set a TFSA lifetime contribution limit of $100,000, which could be used immediately in full any time e.g. for an inheritance; helps present-day retirees with taxable accounts
  
• Remove the effect of RRSP withdrawals on means-tested benefits; makes things less complicated and less punitive
  
• Defer RRSP conversion age to 75; helps those who work longer
  
• Have the Financial and Consumer Agency of Canada do financial education and monitor investment advisors (I think they mean financial advisors, which is much broader than investment advisors) - pretty wimpy, they could and should have recommended that fiduciary duty for financial advisors be put into law with some body given policing powers
  

Pension Reform: a Comparison of CPP(IB) vs RRSP / RRIF / LIRA / LRIF / LIF

There's been a lot of talk lately in Canada about pension reform, a very necessary and worthwhile subject, but unfortunately most of the analysis is from the viewpoint or from the self-interested position of government, regulators and the financial industry. Herewith I present a modest contribution to the debate by comparing two of the existing major options from the viewpoint of the retiree or pensioner, in whose interest all this reform supposedly is ultimately most important.

The two options:

• Canada Pension Plan (CPP) and its investment arm, the CPP Investment Board - thus my new acronym CPP(IB)
• RRSP / RRIF / LIRA / LRIF / LIF - the family of registered retirement plans available to individuals, first to save for retirement, and then to draw income from during retirement

The objective is to determine which best meets the first and highest priority of retirement income, the essential spending needs to maintain a lifestyle. I defined my criteria for this objective in my post of June 15 Pension Reform and What Retirees Need, so now I turn to the comparison evaluation as promised then.

First order analysis: What the alternatives deliver today, as promised and as possible. The CPP is very straightforward - once you are eligible and fill in the form to start payments, you receive a monthly cheque, indexed (increased but never decreased e.g. during deflation) for inflation, for the rest of your life. The registered plans are more complicated and the income must somehow be created from investments. I've assumed that the pensioner will follow what I consider to be the best available method to invest within the plans - a portfolio of passive index ETFs, possibly with annuities purchased at some point.

Below is a table with my comparisons and ratings. I haven't bothered with an overall score because the CPP is clearly and massively superior based on doing what it does now.


Second order analysis: drilling down beneath the surface, how sustainable and sure, and what are the risks of each alternative. Below is the table for those results. Again, the CPP is Victoria to St. John's distance ahead of registered plans.


Third order analysis: what investing challenges must be met, what effort and skills does each require to be successful. Big surprise huh? A pattern seems to have emerged as the CPP is again far superior to registered plans.


One could say that the CPP is the best thing since, and for, sliced bread.

Time to Put Some RSP in the RRSP?

Once you deposit cash into the RRSP the question then becomes what to invest the money in. One of the key pieces of most diversified portfolios is US equities and a popular choice is an index fund based on the S&P 500, perhaps the SPDR S&P 500 (symbol SPY).

Here's an intriguing complement or perhaps alternative - the Ryder S&P Equal Weight Fund which has the memorable ticker symbol of RSP (the connection to Canada's RRSP cannot be a coincidence, surely this is a prophetic sign ;-). For those more pragmatic and try-to-be-rational people, like me, a closer look at RSP reveals some tantalizing data.

The RSP is a passive index fund that differs from the traditional cap-weighted SPY by weighting the portion of each stock holding equally - i.e. each of the same 500 stocks in the S&P 500 comprises 0.2% (100% divided by 500) of the total. Keith Hawkins' excellent Investopedia article S&P 500 ETFs: Market Weight vs Equal Weight explains the similarities and the differences between the two approaches. The article compares results based on the underlying indices but what about the actual ETFs?

The simple Google Finance chart below of SPY vs RSP since the 2003 launch of RSP looks mighty good as RSP is up 42% compared to SPY's 12.6%.

But that's not the whole story. First, there is the question of total returns, which takes account of differences in taxes/ turnover/ capital gains, MER, bid-ask spreads, dividends etc. Turning to Morningstar, the Performance tab of the data on RSP and SPY shows us that RSP looks just as good if not better on a Total Return basis - despite an MER of 0.40% vs only 0.09% for SPY and turnover of 22% vs only 7%, RSP outperformed SPY by a massive 2.21% per year (2.08% 5-yr annualized trailing total return for RSP vs -0.13%) from 2003 to date. What is more, RSP's tax efficiency (see Tax tab) as expressed in the lower tax ratio of 0.48 vs 0.57 is better and RSP has capital losses stored up (against which future capital gains will be offset so that no capital gains will be distributed to fundholders causing tax liability) of minus 23% vs a gain of 3% on the books of SPY. Given the much higher turnover of RSP, I'd guess what is going on is that RSP is accumulating capital losses by having to sell losers leaving the S&P500 at the bottom (they sure cannot be obliged to sell by some company moving up, the S&P 500 is the top category!).

That's pretty good, but the second question is critical. Is this outperformance is only a manifestation of the value and smaller-cap tilt inherent in RSP's indexing method or of a fundamentally better method of tracking the market and thus sustainable in the long term through different market cycles and conditions? That the latter might be the case finds support from studies done by the EDHEC who found that cap-weighted indices in the USA, Europe and Japan were inefficient compared to equal-weighted indices they built, which were similar though not identical to RSP - e.g. Assessing the Quality of Stock Market Indices: Requirements for Asset Allocation and Performance Measurement by Noël Amenc, Felix Goltz and Véronique Le Sour. Standard and Poors' Equal Weighted Indexing Five Years Later on SSRN by Srikant Dash and Keith Loggie also reach the conclusion that equal weight indexing really works both for US stocks and internationally.

If the net effect of equal weight indexed funds is only to under-perform during strong bull markets (and bubbles) and outperform during bear markets, as the commentary by the Rydex's Carl Resnick says in this interview, then that is a very valuable quality, especially for the portfolios of people in retirement, when downside risk is a prime concern.

The big question I have not seen addressed directly, though the Dash-Loggie paper does show the recent varying correlation between the S&P 500 Equal vs Cap-Weight Indices, (it lessened considerably during the tech bubble which is a very good thing), is the correlation of equal-weighted indices with other asset classes. It is the combined effect in the overall portfolio that counts above all. If equal weight is significantly un-correlated with them, the added volatility of RSP on its own is not a concern since the overall portfolio volatility will decline. That characteristic is the reason I think having commodities in my portfolio, among other holdings, is worthwhile.

The idea of putting some RSP into an RRSP merits serious consideration.

New Data on TFSA vs RRSP and Canada's Rube Goldberg Tax System

Cartoonist Rube Goldberg was famous for his drawings of incredibly complex, convoluted machines. That's Canada's income tax system. A new paper from the CD Howe Institute Saver’s Choice: Comparing the Marginal Effective Tax Burdens on RRSPs and TFSAs (kudos to Don Cayo of the Vancouver Sun on whose blog I found the link) reveals the gory detail of the complexity created by the interaction of all the start and stop levels of tax credits, tax brackets, tax surcharges, rebates and clawbacks. The table they show for an Ontario taxpayer has no less than 33 income levels at which tax rates either go up or down. Contrary to popular belief, the tax you pay on your next dollar of income, the marginal rate, does NOT go up constantly and smoothly. It bounces up and down by more than 100% for very small rises in income. These tax items are not special rules for individuals in unique circumstances, it is what everyone faces.

What's more, and what is important for the average person trying to decide whether to put savings into a TFSA or an RRSP, as a result the better choice flip-flops back and forth between TFSA and RRSP. The answer varies by: Province, by income in retirement compared to during working life (the replacement rate) and by working life income level.

CD Howe's Findings

• $20-30,000 or so working income, TFSA always is better and by a massive amount, the GIS clawback being the primary cause as TFSA withdrawals are not included as income for the calculation while RRSP withdrawals are included.
  
• $35-45,000 or so working income, RRSP is better but not by nearly as much as the TFSA advantage in the bullet above
• the boundary between TFSA and RRSP shifts higher as retirement income replacement is lower e.g. in Ontario, at 80% retirement replacement, the TFSA is better up to around $30,000 working income but at 60% replacement, the TFSA is better up to about $39,000
  
• TFSA is better across most of the working life income spectrum for Alberta and Quebec and most income replacement levels
• most surprising, TFSA is everywhere best for the highest income earners of $110,000+ even when their income replacement is only 60% - one would have thought they would end up in a much lower tax bracket and thus conform to the general principle that RRSP is best when your tax rate is less in retirement.
  
• above low income levels, the advantage for TFSA or RRSP is not enormous (less than 10%+/- in marginal tax rate), except for huge spikes up or down in Ontario
  
• a change of only a few thousand dollars in working income can shift the balance, sometimes drastically, from TFSA to RRSP or vice versa, especially in the band $35,000 up to about $80,000 in Ontario (which leads me to conclude that Ontario residents face the most uncertain, difficult and chaotic tax system as far as TFSA vs RRSP planning goes)

Ontario residents in the income range from $35,000 to about $90,000 probably need most to hedge their bets about where their retirement tax rate will end up and to contribute to both their TFSA and RRSP. At least both benefit from the powerful advantage of tax-protected compounded growth while funds are in the plan.

Unfortunately, CD Howe only looked at the numbers for Ontario, Alberta and Quebec, so taxpayers in other Provinces must be wondering where they stand. Don Cayo got preliminary data from CD Howe about BC, which he says is similar to Ontario.

The Federal government could do us a favour by expanding TFSA contribution room to make it equal to the RRSP, or make it a combined total that people can divide between the two as they choose.

Government Ban on RRSP - TFSA Swaps Revisited: One Red Herring and the Real Problem

Sometimes I'm a bit thick and it takes a while for the real reality to distinguish itself from the illusory reality.

The Illusory Reality: Yesterday I noted the scenario mentioned by two other bloggers - see here and here - for supposedly moving funds tax-free from an RRSP to a TFSA. The illusion is that the investor moved funds but what has actually happened is that the investor made a profit on an investment in the TFSA account and made a loss in the RRSP account. To see this, it is only necessary to remember that the exact equivalent result could be achieved by simply buying and selling on the market instead of doing a swap. In fact, a swap is just that - instead of the investor buying or selling on the open market, his accounts buy and sell to each other.

Another tack is to think of it in the investor's shoes - after the stock price rises in the TFSA, you are $XXX better off in total wealth. Would you really want the stock to decline after your RRSP buys it so that the TFSA can buy it back? After the round trip of swaps and the stock decline, the investor has less money in total than after the TFSA made a profit. The TFSA is the same but the RRSP is worse off. In any case, there is no guarantee that the stock will happily fluctuate up and down within the range needed to come out ahead on a net basis. That's why day trading is a highly risky proposition.

This non-problem is a manifestation of the sunk cost fallacy. At each step of the process, the investor is faced with a clean slate and a new decision about how to invest. The past, however recent, is irrelevant. I certainly hope the Department of Finance policy is not meant to stop this kind of investor operation because the government would then be taxing the profits of normal risky stock purchases and sales.

The Real Reality: The real problem is revealed in the discussion on the Financial Webring TFSA thread. It is the fact that the tax rules allow an investor to choose which price within a security's trading range on the day of the swap to have applied to calculate the value of the swap. The difference between the high and low price is what generates the tax-naughty riskless profit for the investor who has eliminated the market risk through judicious use of options. A more volatile stock, or a volatile day to perform the swap, is better because it produces a higher high-low spread and that increases the risk-free profit. Options also use less capital than straight stock, which boosts the returns.

That being the case, the Government would seem to be engaging in throwing out the swap "baby" with the dirty hi-lo price "bathwater". Instead of banning swaps (which doesn't make sense anyway since direct stock trades can effect the same outcome) or changing the rule that any price during the trading day when the swap takes place may be used to value the transfer amount, either declare that two-way swaps of the same security will automatically be valued at the same price within the same day, or perhaps within 30 days in a manner akin to the superficial loss rule. There's even a catchy name to give it - the superficial swap rule.

TFSA Ban on Asset Swaps with RRSPs

Updated later in the day - see red text.
The proposed new rules to eliminate potential abuses of TFSA accounts announced the other day by Finance Minister Flaherty includes one strange rule (see the Department of Finance's Backgrounder section on Asset Transfer Transactions) that bans swaps between TFSAs and registered accounts like RRSPs, LIRAs, RRIFs.

I must admit I was puzzled since I had not previously seen anyone proposing a way to avoid taxes by doing a swap.

There seem to be several explanations of what the rule prevents:

1. A visit to the Financial Webring where all the usual suspects gather and gleefully point out such tax "work-arounds" uncovered in a TFSA thread a post by Marty123 on Oct.20th detailing a highly sophisticated strategy using massive over-contributions and options.
   
2. Blogger Michael James on Money's post TFSA Abuse shows another scheme that seems to fit the bill.
   
3. The Canadian Tax Resource blog gives a similar example to Michael's.
   

Perhaps it is the space limitations of traditional media but the clear online examples put to shame the ambiguous description in the Globe and Mail article on the subject.

My direct question to the Department of Finance for an example has yet to be answered (stay tuned for what they eventually tell me). Here is what they said: "The idea would appear to be that overall, advantage is rarely gained but the scheme is such that a large number of small swaps, particularly involving volatile stock, could enable capital gain to exceed tax liability, using financial software and hedging strategies."

In this first year of the TFSA when the contribution limit is only $5k, it is likely Michael's strategy would hardly be worth it considering each swap is charged a fee by the broker ($45 flat fee per security swapped at my discount broker) and it would eat up much of the tax savings. However, down the road when accumulated TFSA room gathers bulk, the benefit becomes more attractive.

Yet to be confirmed also is the import of the tax penalty. The Department of Finance phrase is: "TFSA amounts reasonably attributable to asset transfer transactions will be taxable at 100%." I would think that what it means is that you would be taxed on whatever "excess" you had managed to transfer - the $1000 in Michael's example - at your normal marginal tax rate i.e. just as if you had withdrawn the $1000 directly from the RRSP, and not at a 100% tax rate, which would amount to confiscation by the government of the excess shifted, an action that even for the government is a tad harsh. " Update: I was wrong, ouch! quote from an official spokesman of the Department of Finance - "No it’s not the marginal rate, it’s a levy of the full amount of the *gains*. It won't matter much anyways since the brokers will all block any sort of swaps with TFSAs and registered accounts.

I wonder if the government will now ban swaps between locked-in registered accounts and non-locked-in registered accounts since the same technique Michael describes could be used to move value and unlock locked retirement money.

RRSPs and the Government's Secret Tax Weapon

Jonathan Chevreau posted today on a BMO Retirement Institute survey finding that boomers expect to receive huge amounts of inheritance money to pay for their retirement.

A quick skim of the Passing it on study reminded me that in the case of inheritances "what you see is not what you will get". The example box on page four of the report shows how an estate worth $350,000 at death can be chopped in no time down to $223,500. That's even before lawyer and executor fees, or probate fees (taxes) get taken out. The reason is that instead of inheritance taxes Canada has a nifty little tax rule called deemed disposition, whereby upon death a person is considered to have sold all property at fair market value and to have cashed in all registered plans like RRSPs, RRIFs, LIRAs, LRIFs etc. There is an exception that a spouse can transfer the registered plans into his or her name without tax being triggered, but eventually the spouse too will die. Sooner or later there will tax to pay and since everything enters the person's income all in the same year, for almost everyone that will mean falling into a much higher, if not the highest, tax bracket.

Think about it, how many people die penniless, their registered plans depleted to nothing? You and I may withdraw RRSP money during retirement at lower marginal tax rates than while we were working but it is pretty sure that come death, our estate will be paying top rates. The government will end not only recouping its tax deductions but may well end up with more tax! That's the secret weapon. If you are dead, you may not care about paying unfairly high taxes on your estate and it is certainly difficult to complain at that point, or to vote against the government.

I've tried getting stats out of Stats Can and the Canada Revenue Agency about the extent of the issue - e.g. what is the average size of registered plans at death - but none seem to be readily available. (An indicator that this issue is real is suggested by Stats Can's table Private Pension Assets of Family Units, which shows that elderly families in 2005 had a median of $186,000 in private pension assets. Another table shows that the net worth of families 65 and older was a median $303,000, of which maybe half might be a tax-exempt principal residence, but the rest would mostly be in some form of taxable savings.) Better it not be too well known, I guess. People might get upset. The problem may get worse over the years as the decline of defined benefit plans means RRSP-type savings become more and more important.

This issue is one reason I am using my TFSA first since that account is "tax pre-paid" for contributions and any gains are tax free, even at death. It is also a reason to take money out of the RRSP when my tax rates are low.

Finally it is a good reason to donate money to charity since the tax credit effectively would mean no tax to pay even at the top marginal rate on the donated amount. Maybe that's why significant numbers of seniors are planning to donate parts of their estate to charity, not to bequeath it all to family.

Is the RRSP Refund as Contribution to TFSA Dipsy-doodle Worth It?

Some commentators have suggested the best way to solve the conundrum of deciding whether to contribute to an RRSP first or a TFSA is to make the RRSP contribution then use the tax refund to put into the TFSA.

Does this make sense? On first glance, one gets the impression that more is being protected from tax. For example, if you put $1000 into an RRSP and are in a 40% marginal tax bracket (despite the fact that no such tax bracket exists, it is convenient to use a round number for illustrative calculations), you get 40% x $1000 = $400 back, which can be put into a TFSA, which seems to result in a total of $1400 being "saved" while a straight $1000 in a TFSA produces no tax refund and so only $1000 is set aside.

The answer is that you are no better off doing the RRSP refund into TFSA than the straight TFSA contribution if you stay in the same tax bracket when you withdraw the RRSP money. Save yourself the trouble. Want to see the numbers? Look at the table below where I've worked through a simple example.



There are other factors to consider in deciding between the TFSA and the RRSP, most of which come out in favour of the TFSA as various people have said like Ed Rempel on MoneyvsDebt.com and on Million Dollar Journey. The one thing that still goes in favour of the RRSP is when your tax rate upon withdrawal will be lower than at contribution. It may still not come out in favour of the RRSP if you lose income-tested benefits like OAS and GIS. Taxtips.ca has a handy calculator in which you can plug in numbers to test RRSP vs TFSA with differing tax rates and considering the OAS and GIS clawbacks.

Hooray for TFSA! Canadian Federal Budget Delivers on My Wish List

Well, you read it here first on my Christmas Wish List 2007. Minister Flaherty has delivered on my request to institute a tax-free savings account just like the ISA that exists over here in the UK. Wonderful. These accounts are so simple and straightforward, they achieve much better than RRSPs the goal of getting people to save. There are no books written about ISAs, unlike the tome Preet Banerjee recently put out about RRSPs, because there's so little to write about. The worst thing about the new accounts is their awful acronym - TFSA. How the heck do you pronounce that - TaFSA? Didn't they learn with RRSP, where is the marketing savvy?

The $5,000 annual contribution limit is too low; better would have been double that and even better would have been four times higher, which would start killing off RRSPs. The "if you don't use it, you don't lose it" feature of the annual contribution limit is a great measure for added flexibility since many families with young children might not get the chance to save for a number of years.

I'd expect the new TFSA, as the info sheet suggests, to be heavily used by seniors, for instance for funding inheritances. The money can be put aside, grow tax-free and be non-taxable at death. While it is in the TFSA, it can serve as a safety cushion for unexpected health care costs and if not needed, passed along to the next generation. Increasing life expectancy could allow amassing a tidy sum.

In short, I believe the info sheet blurb is not too far off (except for the niggardly $5k) when it says "It’s the single most important personal savings vehicle since the introduction of the Registered Retirement Savings Plan (RRSP)."

It was nice to see our government paying attention to this humble blogger. I suppose they are waiting for the next budget to put in an annual tax-free capital gains exemption.

Book Review: RRSPs by Preet Banerjee

This book is written by Preet Banerjee, a financial advisor by day and a dedicated blogger at WhereDoesAllMyMoneyGo by night. The subject - Registered Retirement Savings Plans - is near and dear to all Canadians' hearts, especially at this time of year, when all the banks and mutual fund companies are exhorting everyone to invest for their retirement before the end of February, the deadline for contributions to apply to the 2007 tax year.

The book is thus a timely contribution and a very useful one at that. It is divided into three sections. The first section explains the basic rules about RRSPs and how they fit into the Canadian tax system. The second section is the guts of the content with a series of 41 chapters on a wide variety of topics involving RRSPs in one way or another, ranging from simple tips, rules subtleties to complicated investment strategies. Due to the nature of this content, it is not necessary to read the book from one end to the other. The strategies are stand-alone. The third section is a brief appendix with other sources of information, notably including blogs that Preet has found useful. It is flattering to find my own blog listed among those he finds to be useful - how could I disagree! His description of me as a "very detail oriented writer" is accurate and for me, at least, a compliment - I've found that paying close attention to details is necessary to avoid unpleasant 'gotchas'. This book is not an exhaustive, definitive guide and reference source. Its aim is to provide inventive ways of effectively using RRSPs.

Preet's writing style is informal and free-flowing, very readable and natural. He does a good job explaining terms and technical details, a real bonus for those just getting into RRSPs.

Preet's day job as an advisor shows through in the realism of the examples he chooses and the comments he makes. Key choices that most people must make are analyzed - e.g. pay down the mortgage or contribute to the RRSP, since most people don't earn enough to do both simultaneously. And he does the analysis with numbers, which is far more believable. He also comments on the psychological challenges that often prevent people from gaining the benefit of a strategy, such as the temptation to spend the tax refund from an RRSP contribution instead of investing it somehow or the panic that causes a sell-off at precisely the wrong time when using leverage.

The introduction of Monte Carlo simulation of investment returns in the analysis of various scenarios such as RRSP vs mortgage and RRSP vs non-registered account is a major improvement over traditional modeling, which usually assumes some constant rate of increase. Monte Carlo is vastly more realistic. It gives a much better sense of the risk involved in investing when returns swing amongst a range of positive and negative results from year to year. The fact that an 8% average growth rate with different patterns of positive and negative years can result in vastly different total wealth / retirement incomes will probably be a revelation to many. The book is well worth it for that analysis alone.

There could be a simplified presentation of the long analytical chapters 26, "RRSP vs Mortgage" and 28, "RRSP vs non-Registered Account". So far as I can tell, some factors and information are irrelevant to the conclusions, like the inflation rate, the value of the house and its rate of value gain and how long the investor lives. Though this extra data was likely added to present a believable scenario, it makes no difference to the choice of which is better to do. Another quibble is the conversion of the net savings/wealth to retirement income. We are never told specifically how that happens, though I assume again, if it is done consistently, it makes no difference. The possible confusion is that all the charts show wealth on the Y-axis not income. Some of the charts appear to show unsuccessful outcomes with higher net wealth than successful ones, a puzzle for sure. Could this just be a problem with graphics?

One part of the RRSP vs mortgage analysis that I question concerns the assumed contribution of the $2009 monthly mortgage payment to the RRSP in various scenarios after the mortgage is paid off. At $60k salary, the contribution limit is 18% times $60k or $10,800 yet the analysis posits contributions of 12x$2,009 or $24,108 per year plus an amount of $3,000 already being contributed that grows with salary. Since the investor's salary and contributions are assumed to grow at 4%, in 25 years here is what I calculate: salary $153,798, new RRSP limit $27,684, existing $3k contribution up to $7,690 plus $24,108 equals $31,798, or $4,114 too much. This especially affects scenarios B and D where the investor is assumed to pay off the mortgage early then contribute the mortgage payment amount to the RRSP. I wonder how the conclusion would differ taking that lower RRSP contribution room into account? Would scenario D still beat C?

At the end of the many pages on RRSP vs mortgage, the summary table on page 125 is the key. The basic answer is that when the rate of return of the RRSP exceeds the mortgage rate, the RRSP is better and vice versa. Moshe Milevsky had already shown and explained that in his book Money Logic so it's good that they arrive at the same conclusion. Beyond that, the relative certainty of the mortgage rate against the uncertain, variable investment return is a major consideration that Preet examines at length.

It is surprising to see that the book does not present a case or scenario in chapter 28 on the RRSP vs non-Registered account comparison that shows one of the core supposed benefits of the RRSP - namely, that when one gets deductions at a higher tax rate when working and a lower tax rate on withdrawal when retired, one is much better off saving in an RRSP. The scenario of Anna in the book appears to show that she ends up in the same tax bracket, though that is not explicitly stated. However, it is significant that the non-Registered account strategy only beats the RRSP when the on-going investment taxes of the non-Reg account are paid with other funds. Such an approach does not compare apples with apples, as Preet himself says on page 148.

There are several spots where the author says that a number of other factors could affect the conclusion (e.g. page 155 regarding the RRSP meltdown strategy) but he neglects to explain even the nature of the effects. That's frustrating for the reader.

I notice that some of the book's content seems also to be on the author's WhereDoesAllMyMoneyGo blogsite so interested readers can get a sense of the book for themselves there.

All in all, the book could be better but it still quite useful; it only takes one good idea to be worth the money.

My rating: 3.5 out of 5.

The book can be purchased at the RRSP Book website,

The Impossible: Buying US Mutual Funds in Canada

A reader asks this question:
"I am currently setting up an RRSP for my daughter who plans on providing small monthly contributions. ETF's seem out of the question because of the monthly transaction costs associated. Vanguard index mutual fund for american and international exposure seemed like a good idea since their MER's are so much lower than comparable Cdn products. However my broker (TD ) tells me that they don't offer this. Are you aware of anyone in Canada that does? Any suggestions?"

Maybe your broker should have said instead that it is illegal for US funds to be sold in Canada. All funds must be licensed / registered with the provincial securities commission, e.g. the Ontario Securities Commission, and file a prospectus with it. So I believe Vanguard could theoretically be sold in Canada if it went through all the legal rigmarole of registration and licensing but it hasn't so far. I tried quickly without success to find a link to the actual regulation on the OSC website, though I came across this snippet from another document that confirms the situation:
"Policies that prevent US mutual fund companies from soliciting business in Canada, forbid advisors to recommend US funds to Canadian clients, and prevent the distribution of US mutual funds without filing a prospectus in Canada, all serve to protect the Canadian mutual fund industry. Severe tax consequences also deter Canadian investors from investing in US mutual funds." Source: letter to OSC, Re: An Alternative Trading System, Oct.13, 1999 (I believe the last sentence may refer to the fact that income received from US funds would lose their tax-advantaged character as dividends, or capital gains)

The other alternative of opening an account with a US broker used to be possible years ago but has been shut down by the US authorities - see this (justified) rant by ByloSehi. Sadly, you must forget those great Vanguard mutual funds.

You are aware of the Vanguard ETF option and its limitations for your situation. Perhaps you could accumulate the cash in the TD RRSP in a money market mutual fund to gain a little interest and get the RRSP deduction as well as the regular savings in operation. Once or twice a year you could buy the ETF to minimize the trading fees. I'd also check out brokers like Questrade that offer lower trading fees even for small accounts.

Another option would be to buy TD e-Series (Internet only) mutual funds for a few years. Their fees aren't great but at around 0.48% are not nearly as high as most Canadian mutual funds. After accumulating $20-25k, you could then switch to ETFs. Who knows, by then Vanguard may have expanded to Canada or the Canadian funds may have substantially lowered their fees (both of which are likely to occur at the same time ... call it the Walmart effect)?

I leave you with this statement from the same letter (see p.3) linked to above:
"The result of Canadians being denied access to the US mutual fund industry is the existence of a healthy and productive mutual fund industry in Canada that benefits the Canadian economy." ... but not the Canadian consumer / investor!

RRSP Loans: and Another Thing - Timing is Crucial

My last post concluded that it is worthwhile to borrow to make a lump sum RRSP contribution just before the deadline at the end of February if the tax refund is used either to reduce the loan principal or to contribute back into the RRSP.

What happens if you have barely missed the deadline and are sitting there on March 1st wondering if it still makes sense? The answer is most likely NO. You are probably better off to start making monthly contributions (I'd suggest you use automatic deductions from your bank account so you actually do it), for instance by putting in the same monthly amount as you would have done to pay off the loan.

Why is this so? The reason is that the tax refund won't be coming back to you till the following year instead of within three months or so if you manage to meet the deadline. The absence of that immediate extra boost to your savings destroys the value of the loan in most reasonable scenarios. The only scenario in which the loan still comes out ahead of the regular monthly contributions is when the investment return rate exceeds the loan borrowing rate by 2% or more. Think of the probabilities. On the one hand the loan rate is a sure thing, it won't be lower by chance; on the other hand, the investment return is very uncertain if it is in equities, which is the type of investment that could most likely provide returns in excess of the loan rate. Or, if you invest in something that gives a known return, like a GIC, you will only get a much lower rate that will almost surely be less than the loan rate? Is the loan worth it, then? In my view, no.

As the months pass from March to the following February, the balance gradually shifts in favour of the loan option. At some point, the loan will be the better option. I have not tried to calculate the exact month when it looks better - perhaps half way, after six months, is a guess. But don't put off the contributions just for that reason. The February contribution scenarios all come out ahead of the March contribution scenarios. Putting the power of compounding to work as soon as possible by contributing sooner is the most important principle of all.

RRSP Loans: What to Do and Not to Do

Suppose you have unused RRSP contribution room but it's the end of February and you don't have the cash on hand to make a contribution before the deadline. Should you take out a loan or should you just start contributing monthly to the RRSP? There doesn't seem to be much substantiated / well analyzed advice out there on whether taking out a loan to contribute a lump sum to an RRSP makes sense so here goes. (This one is for you Nicole since you brought up the question.)

Being a non-believer in loans generally (the only loan I've ever had in my life is a mortgage), I have been greatly surprised to find that taking a loan to contribute to an RRSP is a smart move if it is done right. And it is useless if done wrong. The do's and the don'ts are very simple, easy to follow principles.

Do This (in order of priority):

1. by all means take out a loan BUT use all of your tax refund for your RRSP as soon as you receive it by either reducing the loan balance or by reinvesting it in the RRSP (it doesn't matter much which you choose as I explain below).
   
2. repay the loan as soon as possible - the shorter the repayment period the better as you will increase your net savings faster; make sure you can handle the monthly payments, including some leeway in your budget for the inevitable major car repair/vet bills/dental work at the worst possible moment;
   
3. a multi- year loan to catch up a backlog of past accumulated contribution room is beneficial as long as you follow rule #1 above every year
   
4. reduce the loan payment amount, using the reduction to continue to contribute to your RRSP if you still have contribution room and if you have no contribution room left, start investing the payment difference in a non-registered account, or
   
5. keep paying the same amount so that your loan is paid off as soon as possible
   

Do NOT Do This:

1. spend the tax refund; if you do, you would much better off taking the equivalent amount of the monthly loan payment, contributing it to the RRSP each month, then spending the refund you get a year later; spending the refund blows the whole loan scenario out of the water - you are wasting your money.
   

The Scenarios I Looked At:

1. Take Out a Loan, Make a Lump Sum Contribution to the RRSP, Reinvest the Refund in the RRSP and Do Not Reduce the Loan Principal
2. Take Out a Loan, Make a Lump Sum Contribution to the RRSP, Reduce the Loan Principal and the Payments with the Refund, Add the Monthly Difference in the Payment to the RRSP
   
3. Instead of a Loan and Lump Sum, Invest in the RRSP the Same Monthly Contribution as the Loan Payment Would be and Reinvest Tax Refunds in the RRSP as Received

With all the scenarios, I used a spreadsheet to calculate. I plugged in different loan rates and investment return rates, loan repayment durations from one to five years and yes/no refund reinvestment decisions. In all cases, the loan was considered to be taken out just before the February deadline so that the contribution could be counted for the previous tax year and so that the tax refund would be received 3 months later. My criteria for success was net savings wealth, as measured by cumulative RRSP value at the end of each loan repayment.

Caveats, Assumptions and Things That Don't Matter

• if you do happen to have the cash on hand but wonder whether it is better to take out a loan anyway, that will only pay off if the rate of return on the RRSP exceeds the loan interest rate; Moshe Milevsky explained this in Chapter 7 "Borrowing to Invest" of his book Money Logic
• the tax bracket you are in (as long as you pay some taxes ... but then you wouldn't have contribution room if you didn't, would you?) does not affect whether you should take a loan, it only increases the benefit if you are in a higher bracket because you get a bigger tax refund
• surprisingly, the "which is better" answer is quite insensitive to the rate of return on the RRSP investments and the loan interest rate - for differences of 1 or 2% between loan rate and investment return rate, the end result of the two loan scenarios was quite close; in cases where the loan rate exceeded the investment return, scenario 2 was better, and where the return was higher than the loan rate, it was the other way round. It takes a difference of 4-5% to make scenario 1 or 2 significantly better than the other. The reason for this is a key concept for all situations - the tax refund is the critical benefit and main differentiator - the sooner you receive it and invest it, the better off you are. In addition, the loan principal amount on which you pay interest begins to decline from month 1 onwards while the RRSP goes up constantly with any positive return and is boosted significantly by the refund, so that the net interest gained is higher from the RRSP than that paid on the loan - e.g. for a $1000 twelve month loan at 8% the balance in month 6 is $509.97, the interest is $3.95 while the RRSP with annual return of only 4% has gone up to a balance $1333.28 (assuming reinvestment of the $310 refund for an Ontario taxpayer with taxable income in the $37-62k band and a marginal tax rate of 31%) gains $4.43 in return
  
• that also the main reason that scenario 3, where you simply invest on a monthly basis and take no loan, comes out behind the loan options except in extreme cases where the loan rate is very high and the investment return is more or less zero; when you invest as you go, you are always playing catch up, effectively one year behind in receiving those tax refunds and always getting a lower return on a lower RRSP balance
  
• in my calculations, I assumed the refunds from the reinvested refunds would also be reinvested; they get smaller and smaller but they do make a difference; the more you have invested and the sooner it is invested, the greater the effect of compounding
• if you have taken out a multi-year loan and still haven't paid it off but have generated more RRSP room in the past year but still don't have the cash, does it make sense to take out a new loan for a new lump sum contribution? my answer is yes, as long as you can handle, with some safety margin, the combined payments; the same logic applies and it still makes sense for the same reasons.

Othe Factors to Consider:

• a loan can be much more forceful in ensuring that you actually make the savings because it's not just your decision to reimburse or not, you are compelled to do it; a big problem in saving is actually getting round to do it and sticking to it
• a loan is riskier in case of unforeseen urgent expenses; it is easy to interrupt RRSP contributions; you need to have confidence in being able to pay the loan
• with scenario 2 and a small loan, the monthly reinvestment amounts may be hard to reinvest at a high rate within the RRSP; having the contribution sitting in cash will ensure that your rate of return is below that of the loan rate; maybe using a mutual fund is the way to invest that monthly amount in the RRSP
• many/most RRSP loan lenders will allow you to defer the first payment for a few months to allow your refund to come in; that may allow you to reduce the loan principal and payment to an affordable level and thus allow you to make a larger catch-up contribution
  

A Canadian Investor's Christmas Wish List

It's the time of year that Santa comes around and this year I'd like him to bring me these small gifts:

• the capability to hold foreign cash in registered accounts like RRSPs and LIRAs, starting with US dollars but why not other currencies like GBP, EUR and JPY, to avoid having to settle trades back into Canadian dollars and then to repurchase USD again to re-invest, which incurs foreign exchange commissions on each end and 2% extra trading costs. I had suggestions from one discount brokerage (not one of the big five banks) that it was finally about to launch such accounts this month - fingers still crossed.
  
• discount brokerage accounts, both registered and non-registered, that enable low cost trading ($10 per trade sounds reasonable) directly on other major world exchanges like London, Tokyo, Paris, Frankfurt; TD Waterhouse, this should be especially easy for you since it already exists in your UK service offering
• passive index tracking mutual funds with low MERs (0.3% or less is a good target) like those of Vanguard in the US as a competitive alternative to ETFs; this will enable small purchases, re-balancing and will simplify reinvesting and tax returns. Note to TD Canada Trust - start offering your e-Series index funds through other brokerages and not force people to open an account with you ... oh, and lower those fees a wee 0.10% please; at Christmas you will find that if you give something, you will receive too.
  
• combined account portfolio reports from my discount brokerage for all types of accounts, into one integrated portfolio, to save me the trouble of copying all the data from my regular trading account, my RRSP and my LIRAs into a spreadsheet; a very useful extra capability would be to enable me to add labels of my choosing for asset classes and to summarize that as well across all accounts; plus, capital gains tracking on the regular accounts, to make it easier to do my income tax return plus plan year-end tax-loss selling or gains lock-in.
  
• real tax-exempt savings accounts (wonder if Jim Flaherty reads this blog) from our federal government like the ISAs in the UK, in addition to the tax-deferred RRSPs; it's so much simpler and more flexible - no tax deduction since the funds come from after-tax income but growth is completely tax-sheltered and no tax is due on withdrawal, no matter what the type of investment, one's income or age.
• again from the federal government, an annual tax-exempt capital gains amount, say $10,000, like that of the UK

That's not a lot to ask is it, only six things? And I've been a good boy all year.

Capital Losses and Superficial Loss Rule Using an RRSP

Investoid posted a comment that is worthy of a new separate post. In it he says he was not aware of the tactic to sell a holding in an open taxable investment account to lock in a capital loss then to immediately re-acquire the same holding inside an RRSP.

Oops, should have noticed that in the document before posting. There's a rule of investing - if it looks too good to be true, it probably isn't. It applies here. Or does it? Let us say we have conflicting opinions, including from the CRA itself!

This extract from Chapter 5 of the CRA's T-4037 Capital Gains Guide seems to say no, you cannot do that, the superficial loss rule applies. It states:
''you, or a person affiliated with you, buys, or has a right to buy, the same or identical property (called "substituted property") during the period starting 30 calendar days before the sale and ending 30 calendar days after the sale ....
Some examples of affiliated persons are:

• you and your spouse or common-law partner;
• you and a corporation that is controlled by you or your spouse or common-law partner;
• a partnership and a majority-interest partner of the partnership; and
• after March 22, 2004, a trust and its majority interest beneficiary (generally, a beneficiary who enjoys a majority of the trust income or capital) or one who is affiliated with such a beneficiary.''

In other words, the RRSP is a trust and you are the beneficiary so the CRA says no it isn't allowed. Probably, people started taking advantage too much and the CRA changed the rules in 2004 to stop it.

That's not the end of the CRA story, however. It doesn't actually say trust = RRSP and you = beneficiary. In my zeal to get a definitive confirmation I called the CRA helpline. After a good long wait to get to a rep and then again while he went off to consult with someone, the answer was, the CRA says yes it is allowed, and quoted me from an internal document #2001-008077, written in 2001. When I expressed my doubts and asked him to re-confirm, since I was about to post this on the Internet and I am not out to embarrass the CRA (really!), he said he would get back to me by next Tuesday at the latest. (For those who think badly of the CRA for this, ask yourself whether a service rep at a typical corporate helpline, say Bell Canada's, would even consider looking further into such a matter.) Wouldn't it be nice if the CRA added a specific mention of RRSPs (and RRIFs since they would presumably be similarly affected) to their Capital Gains guide regarding this matter?

Web sources don't seem to agree either, perhaps no surprise. Here's a brief sample of results from a bit of Googling:

No, It is Not
AIM Trimark's Capital Loss Planning
Posting in Canadian Business forum in April 2006
Another posting in Canadian Business from Jan. 2006 that mentions an Altimira Funds publication saying No as well.
CIBC Wood Gundy article on Tax Loss Selling with no date and the added footnote, hilarious in light of the absent date, ''The information contained herein is considered accurate at the time of posting.''

Yes, It is OK
Milestone per the previous post (April 2002)
TaxTips.ca. And it's one of my favorites!
Advisor.ca column by Jamie Golombek (a VP at AIM at the time of writing in Nov. 2002)
Canadian Shareowner article from NovDec 2000
Bylo posting of Jonathan Chevreau article in the National Post from Nov.21, 2000
Sterling Mutuals article Avoid Superficial Losses in Nov.2001
Institute of Chartered Accountants of BC Tax Traps and Tips article Nov. 2003

My bet is on the ''not allowed'' side. Any opinions? I suspect all this is to re-discover the sad truth of stale content on the web and the fact that one cannot necessarily believe everything that is written, even from reputable organizations.

Thanks, Investoid for the topic!

Update Oct.10 - Finally got a call back from Revenue Canada and the answer is now NO, you are not allowed to do it, or more precisely, your capital loss will be declared superficial and denied on your tax return. The relevant subsection is 251.1 (g) of the Income Tax as modified in 2005.