Rob Carrick had a column in the Globe today that examines the effect of the recently announced changes in dividend taxation:
Still, there are going to be cases where investors pay more tax without the offsetting benefit of a higher dividend. Preferred shares are one example, while another is the shares of companies that maintain a steady dividend.
Since the dividend tax credit was enhanced a couple of years ago, dividends have in many cases been the most tax-efficient form of investment income (we’re talking here about so-called eligible dividends, or those typically paid by large corporations). Mr. Mida said dividend income may lose this distinction to capital gains, but not by a big margin.
The status quo will hold in dividend taxation until 2010, when a three-year phased adjustment begins.
Mr. Carrick’s source for the figures used in his report appear to be those of Price Waterhouse:
These are different from the Ernst & Young figures that I normally use. Presumably, the two accounting houses have used different assumptions regarding what constitutes a ‘base-case average taxpayer’. Not a big deal … it would be nice to know just precisely what the differences are, but we can’t have everything for free.
Enough! Let’s do some work here! Using Mr. Carrick’s published figures and assuming no change in the marginal rate charged on income:
| Projected Taxation Factors | |||
| Year | Income | Dividend | Equivalency Factor |
| 2008 | 46.41% | 23.96% | 1.419 |
| 2009 | 46.41% | 23.06% | 1.436 |
| 2010 | 46.41% | 24.56% | 1.408 |
| 2011 | 46.41% | 27.59% | 1.351 |
| 2012 | 46.41% | 30.19% | 1.303 |
So … the estimate is that the equivalency factor is going to revert to approximately what it was in the nineties.
Before we take the next step, let’s emphasize to ourselves that these are estimates, approximations and forecasts! In the first place, accountancy firms can’t even agree with each other on what the top marginal rates are, such is the idiotic and increasing complexity of the Income Tax Act. In the second place, a five year forecast of something political like tax rates is going to be even more subject to error than a five-year forecast of investment returns … at least when you perform the latter operation, you can assume that at least a tiny minority of the players have functioning brain cells!
So. This is an estimate. Do with it what you will.
| Estimated Effect of Tax Changes On Perpetual Discounts |
|||
| Year | Equivalency Factor | Change in Spread if Prices Constant |
Change in Price if Spread Constant |
| 2008 | 1.419 | 0 | 0 |
| 2009 | 1.436 | +9bp | +1.33% |
| 2010 | 1.408 | -6bp | -0.89% |
| 2011 | 1.351 | -37bp | -5.48% |
| 2012 | 1.303 | -63bp | -9.34% |
Note on calculation: I use base case figures for 2008 of a PerpetualDiscount yield of 5.39% and a long corporate yield of 5.90%. At the 2008 equivalency factor of 1.419, the current interest-equivalent on PerpetualDiscounts (IE Spread) is 7.65%; the current spread to long corporates is therefore 175bp.
Figuring out the change in IE Spreads is easy – multiply today’s yield by tomorrow’s equivalency factor to get tomorrow’s estimated IE Yield; subtract the (constant) corporate yield to get tomorrow’s IE Spread; subtract today’s IE Spread to get the change.
To estimate the effect on price if the IE Spread is constant, I multiply the change in IE spreads by 14.82, which is the modified duration of the PerpetualDsicount index. Thus, for year 2012, the change in price (from now) is 0.63 x 14.82 = 9.34. To check this … let us assume we have a $100 pref yielding 5.39% at the moment … therefore, it pays $5.39 p.a. If the price drops by 9.34%, the new price will be $90.66; and the yield will change to (5.39 / 90.66) = 5.95%. The Interest Equivalency Factor is 1.303, so this yield will be equivalent to 7.75% interest. We were hoping to get 7.65%, but 10bp difference is due to convexity effects (the modified duration will decrease as the price decreases; modified duration is, strictly speaking, applicable only to infinitesimally small changes in price … and a 9.34% drop is not “infinitesimal”). Additionally, the extremely precise modified duration of 14.82 is calculated using HIMIPref™’s limitMaturity, which assumes a maturity at the current price in thirty years. This is not strictly accurate in itself and is not consistent with the use of Current Yield as an approximation of YieldToWorst. So a 10bp error isn’t bad!
Update, 2008-2-29: This post updates Federal Budget – Effect on Prefs
[…] Update, 2008-2-29: Follow up article, with projected rates is Dividend Taxation Changes. […]
James, I would be wary of making a projection like this, even with all your warnings in place. Investors should not forget that the current dividend tax credit regime, which is very generous, is only a couple of years old. Prior to that, there was considerable double taxation of corporate profits because of poor integration between corporate and individual income taxes. It’s certainly not clear to me from the historical record – and this is not ancient history – that prefs got marked up a lot when the effective tax rate on them was reduced for individuals, so I’m not so sure that they will be marked down a lot when the reverse occurs.
I suppose we should be quiet about it though. Perhaps preferred investors could ignore the above and lobby issuers for compensation for the coming tax change. 😉
It’s certainly not clear to me from the historical record – and this is not ancient history – that prefs got marked up a lot when the effective tax rate on them was reduced for individuals, so I’m not so sure that they will be marked down a lot when the reverse occurs.
Me too, frankly! When the equivalency factor went up, I spent a long time waiting for a really good pop in prices … it never came.
For what it’s worth, my base case wild guess for long term returns is that PerpetualDiscounts should have an interest equivalent yield of about 100bp over long corporates. At the moment, I consider long corporates to be cheap to Canadas and PerpetualDiscounts to be cheap to long corporates. A decline of 70bp in PerpetualDiscount Interest-Equivalent Yield (if it happens!) will merely eliminate the second cheapness and make them merely fairly priced relative to long corporates, which would still be cheap to Canadas …. assuming any opinions in this paragraph can be relied on, which they ain’t.
Long term readers of this blog, though, will remember that I am extremely hesitant about long term predictions. I certainly would not take investment action based on this analysis – I would simply add it to my list of “things that stand a reasonable chance of sort-of happening, maybe”.
For those of us not at the highest marginal tax rate,will the reduction in dividend tax credit simply reduce the amount of dividend income that can be earned tax free? Even taking AMT into account, an Alberta resident today can earn over $50,000 in dividend income tax free.
lystgl – I really don’t know. I’m not a tax expert and I haven’t seen much commentary one way or the other.
Carrick noted in his article:
… so who knows? If you’re a subscriber to Canadian Moneysaver, you may wish to try the “Ask the Experts” facility there.
[…] The quoted top marginal rate is very different from the 30.19% in 2012 estimated earlier! […]
[…] – there are two tax changes scheduled to have an effect on preferreds in the next few years, Federal Bad and Ontario Good, netting out to a modest […]