## Tuesday, November 18, 2014

### Dividends vs. Capital Gains in Retirement

Trying to maximize your after-tax retirement income is a complicated business. Cheerleaders for dividend investing are convinced that the dividend tax credit makes it a no-brainer that a dividend strategy is best to minimize taxes. However, as we’ll see, there are good strategies that use the 50% capital gains exemption as well.

Let’s look at two investors in the same situation but using different investment strategies. Dean, the dividend investor, and Carla, the capital gain investor, are both 58 years old, single, and living in Ontario. (No, there will be no romance in this story.) They both have a \$500,000 RRSP and \$1,232,000 in a non-registered account. (I’ll explain this cooked-up number.)

Dean owns dividend stocks that we’ll assume earn a 2% capital gain and 4% dividend each year. Dean’s dividend income is 4% of \$1,232,000, or \$49,280. This just happens to be the maximum he can earn and pay no taxes other than the \$600 health premium.

Carla earns the same total return of 6% each year, but she owns the exchange-traded fund HXT so that all her returns are capital gains. Obviously, her return will not be exactly 6% each year, but as we’ll see, her strategy is flexible. Carla’s plan is to withdraw the same annual income as Dean (\$49,280), but she is going to split it between her RRSP and her non-registered account so that her total taxable income is \$14,700. This is the level of income where she pays about \$600 in taxes, the same as Dean.

Carla’s withdrawals will be spread evenly through the year. Let’s assume that the average capital gain on her withdrawn capital during the first year is 3%. If Carla withdraws \$35,091 from her non-registered account, this will be a \$34,068 return of capital and a \$1022 capital gain of which \$511 is taxable. If she withdraws \$14,189 from her RRSP, she will exactly hit her income target of \$49,280 and her taxable income target of \$14,700.

In the second year, Carla will have more deferred capital gains and won’t be able to withdraw quite as much from her RRSP. Her RRSP withdrawals continue to decrease in future years as her non-registered account capital gains build up.

I simulated 7 years of Dean and Carla’s strategies assuming 2% inflation and assuming the income tax thresholds rise with inflation. After these 7 years, both Dean and Carla have paid the same amount of income tax (almost nothing) and have the same total assets. However, their split between accounts is different. Here are their situations at age 65 (rounded to the nearest thousand):

Dean
Non-registered account: \$1,415,000 (unrealized capital gain of \$183,000)
RRSP: \$752,000

Carla
Non-registered account: \$1,515,000 (unrealized capital gain of \$514,000)
RRSP: \$652,000

There isn’t an obvious winner here. Carla has managed to use up some of her RRSP tax-free, but Dean has a smaller unrealized capital gain.

Observe that Dean and Carla are now somewhat locked into their respective strategies because of the unrealized capital gains. Any change of strategy requires some selling that would lead to significant capital gains taxes.

As Dean starts to draw OAS and possibly CPP, his tax situation changes; he will be paying income taxes on his dividends. Carla will pay some taxes as well but she has more flexibility in how much capital gain she chooses to realize.

At the end of the year they turn 71, things change significantly again as they will both have to make mandatory RRIF withdrawals. Dean’s taxable income will rise to the point where he will pay significant taxes including a substantial OAS clawback. Carla’s taxes will rise too, but she will once again have more flexibility in how much capital gains she will realize.

To make a complete comparison of Dean and Carla’s strategies, we’d have to make assumptions about CPP payments, desired income levels later in life, and how much of an inheritance they’d like to leave. I don’t see much point in doing this, mainly because the strategies so far have both of them living on less in their 60s than in their 70s, 80s, and beyond. It’s more realistic to devise strategies that aim for constant inflation-adjusted income with adjustments if portfolio returns disappoint.

I’ve made a few attempts to devise tax-smart strategies based on targeting a constant after-tax real income for life (with adjustments if portfolio returns disappoint). I also included a safer risk level than Dean and Carla’s 100% stock allocation, and took into account TFSAs.

Each time I work out the final result in an analysis that is more complex than the scenarios above, I find that capital gains strategies give slightly higher income than dividend strategies. While I’ve tried to optimize each strategy, I can’t guarantee that I’ve made the best possible choices to minimize taxes.

So, I can’t say with any certainty whether dividend investing is better or worse for taxation, and it may depend on the specifics of account sizes and other factors. What I can say with some confidence is that those who focus solely on the dividend tax credit are missing the big picture.

1. What happens if Dean withdraws from the RRSP as well to increase his taxable income to \$14,700 (He likely will have dividends stream here as well)? You are right that tax optimization is complicated. Through in early retirement, and a few other variables such as couples with mis-matched earnings/CPP etc and I find it too complex.

1. @Gary: Sorry your comment got lost for so long, but I found it, finally. The response is in the next comment.

2. A reader asked a question that seems to have disappeared. Perhaps he figured out his own answer and deleted the comment, but I thought I'd answer it here for others:

"What happens if Dean withdraws from the RRSP as well to increase his taxable income to \$14,700."

Dean's taxable income is actually higher than \$14,700, but he gets the dividend tax credit reducing his taxes owing to near nothing (just the \$600 health premium). If he made a withdrawal from the RRSP as well, he'd definitely have to pay more taxes.

3. Don't know what happened to my original comment (I did not delete it). Thanks for the follow-up - makes sense. As I said I'm my missing comment the optimal tax minimization strategy in retirement is very complicated with so many variables (registered/non-registered, div/cap gains, OAC limits, on top of withdrawal and investment returns.

1. @Gary: Thanks for your comment. You're right that this whole area is complex. I still haven't thought of a good way to explore all possible withdrawal and contribution strategies among RRSPs, TFSAs, and non-registered accounts.

4. I have wondered if dividend investing might help protect against unfavourable sequence of returns risk in retirement. Although most credible blogs suggest that broad market investing (perhaps with value, etc, tilts) is generally superior to dividend investing for long term accumulation portfolios, could dividend stocks help buffer bear markets when drawdowns are required? That is, could the dividend income help prevent the need to sell stocks that have lost capital gains in a bear market that is short enough that major dividend cuts don't occur, and 'bridge' over until markets recover?

Personally, I plan to continue to carry my current broad market 70-30 equity/bond split (including 15% of various tilts) into retirement next year, given that I will also have a moderate pension income to cover basic expenses. But I have been wondering whether to include dividend ETFs for the above reason.

1. @Anonymous: I've never attempted to study how dividend investing fares with sequence of returns risk. This might be a good question for someone like Larry Swedroe.

2. thanks. Maybe I'll have time to work it out myself next year, in retirement, somewhat of course too late!

5. Hi Michael... Interesting post. How did you come up with the \$49,280 figure for "the maximum he can earn and pay no taxes other than the \$600 health premium."??

I thought that taxable income kicked in at somewhere around the \$10,000 mark? Or... is this just Ontario provincial tax?

1. @Mike: The \$49,280 figure is so high because of the dividend tax credit. Normally, so much income leads to having to pay tax in any province. But, our tax system treats dividends differently. First they are "grossed-up" to a larger amount, and then you get a deduction against the amount of tax owing. In Ontario, the tax rates are such that \$49,280 is the most dividends you can make and have all the tax owing except the health premium wiped out by the tax credit (assuming you have no other income). This maximum figure varies by province.

2. Ah... Thanks for the detailed (and quick!) reply.

6. Great post.

There is literature about retirement income but few books do it justice, for the reasons I believe you have discussed, it is "a complicated business" with many variables...

I am a cheerleader for dividend investing but I'm not convinced there is a HUGE tax savings, since capital gains are taxed at a lower rate than dividend income. Until the (government) change the rules, I think in a non-registered account, most investors might actually be better off with a capital gain approach. The only problem I see with that is, what do you do in down markets? Sell the equities? This is where dividend investing helps I think, at least a little bit, investors can rely on the income generated vs. the appreciation and thus avoid selling some assets at depreciated prices.

Yes, you can make about \$50k essentially tax-free in a taxable account with a dividend income approach but very few folks would actually be in position to take advantage of this; i.e, that is good income without any pension or other income supplements in retirement. I doubt most retirees rely exclusively on dividend income.

I will continue to focus on dividends for reasons not in this article but I think what you already know about me, they are tangible and they help me stick to my investing plan. The DTC is a bonus but it's not the be-all end-all, and that could always change with the stroke of a government pen anyhow.

Will definitely highlight this one this week Michael.
Mark

1. @Mark: Thanks. In a more realistic situation, a retiree would have some fixed income investments as well. In a down market for stocks, natural rebalancing would shift withdrawals to the fixed-income side. However, for long enough down markets, an index investor retiree would have to sell some shares of stock. But this isn't much different from the dividend case. Dividend companies choose slower growth to pay dividends. Companies that pay lower dividends, on average, reinvest more of their profits and grow faster. When a retiree sells shares, he is turning some of that growth into cash. At it's core, for dividend stocks to serve investors better, they have to outperform the index. They may well do this, but maybe they'll underperform the index; my crystal ball is cloudy.

You're right that few retirees would have dividend income exclusively, but it might be true for some before age 65. Using dividends as a device to help you stick with a plan certainly makes sense.

7. Great post. I think a combination of dividend and capital gain is the way to go. For non-registered account the capital gain is the way to go. Like Mark I'll continue to focus on dividends.

1. @Tawcan: As Mark said, a big part of the equation is sticking to a sensible plan. I hope it works out well for you.

8. Great blog. Interesting topic. One thing that I can't quite get is the return of capital figure. How can Carla claim most of her withdrawal as ROC? I would have thought that by withdrawing funds she is realizing a capital gain (i.e. half of the amount would be taxable)?

1. @Anonymous: In this scenario, I treated it as though both Dean and Carla had just come into the money in their non-registered accounts. So, their adjusted cost-bases (ACBs) started as equal to their account balances. Carla's initial withdrawals were almost entirely capital in the sense that the difference between sale price and ACB was small. I could have started then both with a smaller ACB (as though they had built the money over time). In this case, Carla would have been able to withdraw less from her RRSP tax-free each year, and by age 65, Dean would have had a smaller advantage over Carla in the area of deferred capital gains. But the uncertainty of who was "winning" would have remained.

9. Hey Michael

Very interesting article. As others have said, the retirement income topic is incredibly complicated especially when considering the taxes now and when RRIF age..

I've been retired for a year and a half and am one of those people living mainly off dividend income. My only other income is CPP. I don't have any company pension.

My wife and I each have an RRSP, TFSA, and non-registered account. My wife was a stay at home Mom so won't be getting any CPP.

Our annual dividend income is now \$96.7k. This means we have extra income so we're still in a bit of an accumulation situation. We just use that extra dividend income to buy more securities which in turn increases our dividend income. Also, 75% of our holdings increased their dividend in 2014 which again increase our dividend income.

I did a major analysis to try and optimize taxes and future potential clawbacks. As a result, we are transferring some of our RRSP holdings in kind into our non-registered accounts and taking a tax hit now to try and save more later.

Anyway, if all goes reasonably well (especially health wise), we will never have to sell anything. I consider this a major bonus of the dividend income approach vs the capital gains approach. It avoids the stress of trying to figure out what to sell and when. Also, anything that is sold for living expenses would be decreasing the future dividend income.

Ciao
Don

1. @Don: It sounds like you have more savings than you need, an enviable position. It's true you have fewer decisions to make because the dividend income keeps flowing whether you need it or not. This is actually a disadvantage from an optimization point of view because you could be deferring taxes rather than paying them. As for avoiding the shrinking of future income, dividend stocks have a built-in expectation of lower growth because the companies are paying a high dividend. However, none of this matters much because it seems you have more savings than you need to succeed with either a dividend or a capital gains approach. No doubt other retirees would be envious.

2. Hey Michael

I have to disagree with your comment on optimizing taxes. We do need a certain amount of income to live off. Let's assume it's \$60k per year. If both my wife and I have \$30k of dividend income, then our taxes are 0. If we have \$30k of capital gains, then our taxes are \$379 each. (Using taxtips.ca for AB residents).

There's also still the problem in the capital gains approach of what to sell to get the \$30k each.

As you commented, it's a moot point for us but it won't be for some other retirees.

Ciao

3. @Anonymous: Once you start collecting OAS and CPP (and later RRIF income), your dividend income won't be tax-free any more. Capital gains start to look better in this case. Also, to live on \$30k each, only a fraction of that will be capital gains because your ACB will be non-zero.

10. Hey Michael

It looks like we are going to have to agree to disagree.

You are talking about Return of Capital and use HXT as your example. HXT doesn't pay any distributions so you'd still have to sell some units to get your income for paying expenses in retirement. It may be fine while you are young and working and don't need the income but that changes with retirement.

I totally agree that HXT is the better option while working but dividend income is better once you are retired and need the income and after-all, your title does say "In Retirement".

If you sell units. that ROC now becomes an actual capital gains and you still end of paying more taxes than with dividend income.

I did a test with CPP & OAS of \$15.5k per year and dividend income of \$15k vs capital gains of \$15k and the dividend income tax is \$650 while the cap gains is \$2301.

I'm still convinced that dividend income is better in retirement. The taxes are lower and you don't need to sell any units so the dividend income remains the same. With HXT, you'd have to sell units so your number of units decreases over time.

Ciao

1. @Anonymous: With HXT, your ACB will never go to zero. Even if the ACB is only one-third of the current value of HXT, when you sell \$15k of HXT, the capital gain is only \$10k. Once the RRIF income kicks in at age 72, dividend income will start to get taxed more heavily because the grossing up pushes you into higher tax brackets (and possibly up to the point of OAS clawback).

When you focus on the number of HXT units decreasing over time, you're making a common mistake. Because HXT grows with the total return, each unit grows faster than each share of a dividend stock, over the long term. The dividends that companies pay effectively bleeds away the potential growth of shares which has the same effect as selling some units of HXT.