Monday, October 1, 2012

Is a Lump Sum or Annual Contributions Better for an RESP?

Nancy Woods at Globe Investor answered a reader question about an RESP for his newborn grandchild: “does it make more sense to contribute a lump sum and forgo the government grant (CESG) or do I make annual contributions and take the government’s free $500? Signed Bill”. Unfortunately, her analysis failed to find the best option.

This question only matters if Bill actually has $50,000 (the maximum total contribution RESP amount) available right now. So, he either throws it all into the RESP now or he puts a certain amount in each year and invests the amount held back in a non-registered account.

The advantage of the lump sum right now is longer tax-free compounding. The advantage of spreading out the contribution is that each year the government will match 20% of the contribution up to a maximum of $500 per year and a lifetime maximum of $7200. (There are also catch-up provisions, but they are not relevant in this case.) Woods concludes that Bill has two options:

1. Make a $50,000 contribution right away and get only one $500 government grant.

2. Maximize the government grants by making a $16,500 contribution in the first year, then $2500 more each year for 13 years, and a final year’s contribution of $1000.

Assuming a 5% annual investment return (tax-free in the RESP and taxed in the non-registered account), option 1 leads to an RESP balance of over $121,000 when the grandchild turns 18, but option 2 only gives $119,000 in total between the RESP and non-registered account.

Woods concludes that tax-free growth trumps the government grants by a small margin. However, she fails to consider the best option, which is somewhere between options 1 and 2. To see what I mean, consider the strategy of making a $47,500 contribution right away and $2500 the next year. This would get $1000 in government grants. Compared to option 1, this strategy loses out on one year of 5% tax-free growth on $2500 but makes 20% on this $2500 with a government grant. This is clearly better than option 1.

Even better would be if Bill makes an initial $45,000 contribution right away and $2500 for 2 years. No doubt you can see where this is going now. The best answer is for Bill to make a sizeable initial contribution and then $2500 for several years until he reaches the maximum $50,000 lifetime maximum. The optimum number of years of smaller contributions depends on the rate of return assumptions and Bill’s marginal tax rate.

If we assume a 5% investment return and a 40% marginal tax rate, Bill’s best option is a $35,000 initial contribution and $2500 more each year for 6 years. This gives Bill’s grandchild a final total of over $124,000 at age 18. The best strategy will vary with the return and tax assumptions, but one thing is certain: contributing an initial lump sum of $50,000 is not optimal under any sensible assumptions.

19 comments:

  1. This analysis is mathematically correct, but it
    - assumes an unspoken risk that the grandchild will be able to actually use up all the investment,
    - neglects income tax deductions for the contributor,
    - ignores a risk of grandparents not being able to contribute through the 18 years

    Whichever way you slice it, if we assume that funds destined for RESP are already invested in the same market, then all analysis becomes much simpler. Consider this strategy: invest $47.5K in market privately and $2.5K through RESP and receive $500. Every year move 2.5K from private investment to RESP. Repeat until 18, until grandparent die or realise that child will not choose to become a doctor. This way investor collects all "lump sum" growth and all grants. It's the same as if 50K stayed invested through 18 years and government paid 500 annually. Obvious without math that this strategy beats any other, but math says: 50 grow to 120, 500 annually grow to 14K. And all possible taxes are saved, and flexibility is left.

    ReplyDelete
  2. @AnatoliN: Regarding your 3 points:

    1. You're right that there is a risk of a child not using up all of an RESP for school. However, this risk is the same across all of the scenarios considered because the grandparent, Bill, intends to make the full $50,000 contribution while the child is still young.

    2. With RESPs, there are no tax deductions for the contributor. The benefit of RESPs is tax-free growth.

    3. The assumption in this case is that Bill will be committing %50,000 immediately. The only question is how to split the money between an RESP and a taxable investment account.

    The scenario you describe is the way that most people would approach RESPs because they don't have $50,000 available immediately. However, in Bill's case, the disadvantage of the approach you describe is the lost tax-free growth from not contributing a large amount to the RESP right away.

    ReplyDelete
  3. I had never considered this issue before. Seems to me $50,000 contributed at age zero would indeed appreciate enough to fund a great (well, expensive anyway) education.

    As you say not many people are faced with the opportunity to invest a $50,000 lump sum when a child is zero, so this is moot for them. However, I'm going to be looking at the issue a little differently since I do have some unregistered investments and a recently vanquished mortgage.

    Thought provoking as usual, Michael.

    ReplyDelete
  4. @Gene: It definitely feels good to kill off a mortgage. Are you considering being a very generous grandparent by filling up an RESP?

    ReplyDelete
  5. Actually, my daughter just turned six, I've contributed just enough since 2006 to get the full grant amount. Anything extra seemed like a waste, since I've always looked at the contributions from a perspective of maximizing government grant money, more or less ignoring/discounting the tax sheltering.

    I might review Mike Holman's RESP book regarding RESP withdrawals and consider a large lump sum contribution to shelter part of my portfolio from taxes. Er... I mean provide ample education funding for my child. ;-) A second child would help shelter even more investments.

    This reminds me of one of your prior posts where you speculate that an individual could spitefully ensure that a child's family could not contribute to an RESP by opening an RESP for that child and contributing the maximum allowed, then collapsing the plan twenty or thirty years later.

    ReplyDelete
  6. @Gene: I guess I shouldn't make age assumptions based on the fact that you've paid off your mortgage; I paid off my first mortgage at age 28.

    Be careful with the RESP if you don't think all the money will go toward your daughter's schooling. The government has structured the RESP rules to discourage those who are just looking for a tax shelter. Reviewing Mike Holman's book is a good idea.

    ReplyDelete
  7. Some of my school classmates are grandparents, but at 41, that's pretty quick. 28 is very young to own a home outright, we didn't buy a home until I was 34.

    Since I've been focused on paying down the mortgage, I still have room in my TFSA, so I still have a few sheltering options. Of course, buy and hold in a non-registered account can be a good tax-deferral strategy as well.

    ReplyDelete
  8. While I don't have neither kids nor $50,000 right now (the former may change soon, hopefully the latter could change as well), this is a quite thought provoking, and the answer is not that obvious when just looking at the extreme scenarios as you proved. If I were Bill, I would probably go somewhere in the middle as suggested.

    ReplyDelete
  9. @UI: For a lot of problems the best solution is at one extreme or the other, but this is one case where the optimum point is somewhere in the middle.

    ReplyDelete
  10. This is a great spin on the average article about RESP's.

    As a former financial advisor, I found that many people still think that they get a tax break on contributions. Unfortunately, it's only on the growth once in the RESP. As the basic grant offers a 20% annual

    ReplyDelete
  11. Doesn't such an analysis also fail to consider the good old concept of dollar cost averaging? ie: by spreading out investment timing over a longer period, one reduces the risk of buying at a high and allowing one to pick up more when low?

    ReplyDelete
  12. @Marko: You're right that many people don't understand well how RESPs differ from RRSPs.

    @Sleepydoc: The method of investing the money could be separate from the timing of adding to the RESP. For example, the RESP could hold some assets in cash that gets invested over time.

    ReplyDelete
  13. But if you are arguing that the lumpsum advantage is tax free compounding - there wouldn't be much advantage in just putting it in cash - there wouldn't be any growth to compound.

    ReplyDelete
  14. @Sleepydoc: Keep in mind that the $50,000 exists right away and we're deciding whether to put it all in the RESP right away or split it between the RESP and a non-registered investment account. So, we have to make the decision to invest it all right away or spread it out over time regardless of whether we put it all in the RESP right away or not. All studies have shown that if you have the money to invest, you're better off investing it all right away. Dollar-cost averaging applies when you are saving a fraction of each pay cheque.

    ReplyDelete
  15. Outstanding article, Michael. This choice being relevant to me, I analyzed both extremes but failed to consider the possibility that the optimum solution lay in the middle. Thanks so much for writing this.

    ReplyDelete
  16. Awesome article, thanks. I have a question. In our situation we have a higher expected return (7%) and a higher tax rate (50% next year) than Bill. I'm trying to work out what our optimal RESP strategy would be, given that our other tax-sheltered accounts are maxed-out. Would both of these changes mean we are better off contributing more upfront than Bill?

    ReplyDelete
    Replies
    1. @James: By my calculations, the optimum for these new inputs is contributing $42,500 up front and contributing $2500 each year for 3 years. But make sure your situation lines up with the assumptions. Good luck.

      Delete
  17. Thanks Michael. I wasn't expecting you to run the calculation! Much appreciated.

    ReplyDelete