## Tax Impact on FixedResetPremium Yields

The discussion ten days ago on this excellent blog about the impact of tax on premium bonds led me to consider if there is there a tax downside to owning premium resets.

How is the capital loss between the purchase price of say \$28 in todays market and the call price of \$25 dealt with?

Not many of us have capital gains to use as an offset!

He was referring to a comment by prefhound on my essay The Bond Portfolio Jigsaw Puzzle. And, naturally enough, his use of the phrase “excellent blog” virtually guaranteed a response!

In order to investigate the problem of tax effects, we need:

The last two requirements are permanently linked on the right-hand panel of this blog under the heading “On-Line Resources”. Note that we don’t really need the “FixedResets” version of the calculator; since we’re only going to be calculating yield to the first call, the regular version will do the same job; but we’ll use the souped-up version anyway. Why not?

Calculations will be performed for an Ontario resident with taxable income of \$150,000. Ernst & Young claims the marginal rate on capital gains is 23.21% and the marginal rate on dividends is 23.06% (compare to the marginal rate on income of 46.41%, which is not used in this calculation).

pugwash specified a price of \$28 for discussion, so for discussion purposes we’ll examine HSB.PR.E, which closed last night at 28.01-15. It pays 1.65 p.a. unitl the first Exchange Date 2014-6-30, when it resets to GOC5+485, or is called at 25.00.

The dividend rate of 1.65 implies quarterly payments of 0.4125. For taxable accounts, we will assume that this is reduced by 23.06% to 0.3173775. Dividends are paid at the end of June/Sept/Dec/Mar and the next ex-date is March 11 (estimated) so we’ll get the next dividend.

It should be noted that a horrifyingly precise calculation will not pay the tax on day of receipt, as assumed in the above paragraph, but pay annual taxes in the following calendar year. HIMIPref™ does this calculation, but the current calculation using the spreadsheet software doesn’t.

The maturity price is 25.00, which is all we need for the non-taxable calculation, but taxable accounts with capital gains will be able to claim the 3.01 capital loss until maturity. The tax rate of 23.21% on capital gains implies that this deduction will be worth 0.698621, so a taxable account with capital gains to offset the loss may use a maturity price of 25.698621.

Having accumulated the data, we can fill in the calculation spreadsheet:

 HSB.PR.E Yield-to-Call Calculations Data Non-Taxable Taxable with Capital Gains to offset loss Taxable without Capital Gains to offset loss Current Price 28.01 Call Price 25.00 25.698621 25.00 Settlement Date 2010-1-28 Call Date 2014-6-30 Quarterly Dividend 0.4125 0.3173775 0.3173775 Cycle 3 Pay Date 31 Include First Dividend 1 First Dividend Value (if different) [blank] Reset Date 2014-6-30 (irrelevant) Quarterly Dividend After Reset 0.4125 0.3173775 0.3173775 (irrelevant) Annualized Quarterly Yield to Call 3.43% 2.61% 2.04% Effective Tax Rate 0% 23.91% 40.52%

Note that the calculated yield on the taxable account with no capital gains (2.04%) is a little harsh, because it does not reflect the fact that the investor will have a capital loss of 3.01 that may be used at some time to offset future capital gains. However, if he never makes any capital gains, this asset will be worth zero.

Note also that the effective tax rate for a taxable account with capital gains (yield of 2.61%, effective tax rate of 23.91%) is in excess of both the capital gains rate and the dividend rate. This is because the investor is paying tax up-front (when the dividends are received) and receiving the benefit of the capital loss later.

### 8 Responses to “Tax Impact on FixedResetPremium Yields”

1. Chris says:

James,

Last night there was a very interesting post up about ultimateMaturity (in reply to my previous comments). I skimmed through it and was looking forward to reading the details today. However it seems to have disappeared off the site. Any chance of getting it back?

Thanks!

2. jiHymas says:

Any chance of getting it back?

Eventually! I found some problems and have to (i) decide what to do, and (ii) do it.

3. prefhound says:

Naturally, James never lies, but there are three points I would add:

1. The investor with no capital gains today, and who never gets capital gains to offset his/her loss, will, I believe, be able to take the capital loss at death. The eventual tax benefit needs to be discounted at the after-tax rate for however many years it takes to realize it. If, for example, one assumes waiting five years past call (i.e. 9 years to 2019) the effective tax credit is at a rate of 23.21% x (1 / (1.025)^9) = 18.6% — not quite a complete writeoff, although long delayed — and certainly not spendable by the investor! Some people have so many historical capital losses, they must wait until they die to get a credit (and the executor has to remember) — 28 years gives a credit worth half of 23.21%. Note that these estimates are quite sensitive to the assumed after-tax return AND the time to realization.

2. Depending on how much dividend income one gets and on one’s tax situation in general, CRA will demand quarterly tax payments approximating last year’s tax not paid by source deduction. In practical terms, an investor counting on significant dividend income will be paying tax on it quarterly, so the delay between receiving the dividend and paying the tax will be low (typically 45 days, on average). Therefore, I usually make no adjustment for paying taxes later than income is received.

3. Mr. Hymas is an astute investor with reasonably well-off clients, so it is sensible for him to use Ernst & Young tax brackets for \$150,000 income. However, this is a critical assumption because the marginal tax rate on dividends is very sensitive to income (and becomes negative, I recall, below \$60,000 taxable income). Thus it is essential to repeat James’ calculation with a correct estimate of the marginal tax rate on dividends applicable to the investor. Of course, capital gains rates will vary somewhat with income too, but not as much as dividend rates.

When you take these points into consideration, investors with lots of capital losses and higher incomes should probably avoid both premium bonds and prefs in taxable accounts. If the investor has taxable income less than \$60,000, he/she should do a proper after-tax comparison to see if the lower taxes on dividends work in their favor.

4. jiHymas says:

able to take the capital loss at death

Really? I didn’t know that!

marginal tax rate on dividends is very sensitive to income

Additionally, because of the “gross-up” part of the Dividend Tax Credit and Gross Up, seniors subject to OAS clawback can find that their effective marginal rate of dividends is greater than for those with no OAS

5. prefhound says:

http://www.hrblock.ca/your_life/losing_loved.asp

Check out Net Capital Losses on this HR Block Page. It is probably in CRA documentation somewhere, but I found it here first.

6. jiHymas says:

H&R Block says:

Normally, you can only claim capital losses to the extent that they can be used to offset taxable capital gains. The balance becomes a net capital loss that can be carried back three years or forward indefinitely to offset capital gains realized in those years. However, in the year of death, both capital losses incurred in the year and net capital losses carried forward may be claimed against other income.

So it’s only good to the extent that there is income in the year of death – but that’s still better than nothing!

7. tobyone says:

CRA has a handy 32 page document T4011: “Preparing Returns for Deceased Persons.” Relevant section on page 25 wrt to net capital losses…..”If there is still an amount left, you may be able to use it to reduce other income on the final return, the return for the year before the year of death, or both returns…….”

8. […] March edition will contain an appendix discussing the Interest-Equivalency Factor, Tax Effects on FixedReset Premium issues, with a discussion of the odd case of BAM.PR.R thrown […]