“Three steps to planning your spending in retirement” was published by Firstlinks on 7 July 2021, having appeared earlier as the leading article in the (London) FT Money edition in April. Judging by the 14,000 reads on Firstlinks, it was helpful to many readers but they also had comments and follow-up questions.
Here I’ll respond to three of them:
- flexible withdrawals
- using equity dividends for safety
- tax implications.
First, as background, let me summarise the approach I described that I use for my wife and myself.
I identified two independent financial risks: that we might live longer than most people of our age, and that our equity investments might drop in value early and stay low for a while, making us withdraw spending money from a depleted pot.
To reduce the chance of each to no more than 25%, we selected a planning horizon of 31 years (based on our ages), and invested five years of withdrawals in short-term securities (our insurance bucket), because historical statistics suggested that a falling equity market recovered in real terms within 5 years 75% of the time.
We calculated the sustainable annual after-inflation and before-tax withdrawals that went with this combination and invested everything that was not in the insurance bucket in a global equity index fund (our growth bucket).
There were lots more points, but that’s the essence of it.
Flexible withdrawals
I appreciate the Australian comments that I’ve over-thought the issue, and that my analysis and introspection must make retirement impossibly difficult for me. “Lighten up,” I was told. Invest in well-run companies and the rest will take care of itself.
Thanks for your concern. I’m actually thoroughly enjoying retirement – particularly because of my analysis. The worry is over. I re-examine our position every year, and adjust our withdrawals a little bit, as I’ll explain in a moment. The other 364 days, I’m happy.
With no insurance bucket, and 100% in well-run companies (tough to identify them in advance, isn’t it, and it’s too late after the badly-run ones reveal themselves), that wouldn’t have worked if, for example, we had retired at the start of 1973 (the worst-case historical test, and worse than I’m planning for, obviously, but still very instructive).
I used US stats and found that the first two years of equity real returns were -8.3% and -34.2%. After 5 years the cumulative equity real return was still worse than -30%. Subtracting 5 years of withdrawals, the remaining assets were down to 33% of their original value.
What would the insurance bucket have achieved? The bucket itself would have been exhausted (slightly earlier than expected, because cash cumulative returns were -7% over the period), and the remaining assets would be down to 43% of their original value. That may not sound like much of an achievement, but future withdrawals from that point forward would have been at 70% of the original withdrawal rate, compared with 53% with the growth-only approach. Sad, but noticeably better.
A number of readers said that they use flexible withdrawals in response to changes in the market value of their growth pot. An excellent idea! I do too. But I don’t vary the withdrawal in proportion to the market swing (like reducing it by 30% if the future indicated supportable withdrawal rate is 70% of the initial rate). Instead, I spread the amount of the swing (30%, in this example) over our remaining planning horizon, to smooth it out.
Spreading relies on a belief in (or at any rate a hope for) ‘mean reversion’, the notion that, over the long term, returns will come back to the average. But if future market declines are steeper and longer than before, the gradual declines in withdrawals won’t be enough. It’s a risk to be conscious of.
I wonder how the reader with a 100% equity portfolio would have reacted by the time 1975 arrived. Regardless, it’s easier to cope in hindsight!
I’m reminded of the saying that “no battle plan survives its first contact with the enemy.” That’s right: no plan will ever work out perfectly. But the work that’s gone into the plan will help you adapt, as circumstances change – that’s why we plan. And that gives you resilience.
Using equity dividends for safety
An excellent comment said, in essence, that the safety bucket should take into account the cash flow from equity dividends, and that this would increase the size of the growth-seeking bucket. Quite right.
As an example, with a 30-year planning horizon and a 5-year safety bucket, and using my 4% real annual return assumption for equities, a 1% dividend yield used towards the withdrawal would reduce the initial size of the safety bucket by about 1/3, a 2% dividend yield would reduce it by about 2/3, and a 3% dividend yield would reduce it to virtually zero.
That ought also to increase the implied sustainable real withdrawal. And it does. But by very little, unfortunately. Even the 3% dividend yield only increases the annual withdrawal by about 2%.
Tax implications
In most countries (although not for most retirees withdrawing their superannuation in Australia), each year’s withdrawal is subject to tax, so you don’t get to spend it all. For many, tax will itself be a significant expenditure. So if there’s a way to minimise it, that becomes important.
I have no general principles for you. It’s as if you’re asked to approach the taxing authorities with all your money stuffed in various pockets in the clothes you’re wearing, and they say, “From Pocket A we’ll take 40%, from Pocket B 20%, you can keep whatever is in Pocket C, from Pocket D …” And so on.
Naturally, before you approach them the following year, it makes sense to rearrange your money across the pockets. But finding out how to do this is difficult, and whole tribes of people make their living by getting to know the complexities and advising non-technical citizens about how to minimize the total.
Don Ezra, now happily retired, is the former Co-Chairman of global consulting for Russell Investments worldwide, and the author of “Life Two: how to get to and enjoy what used to be called retirement”. This article is general information and does not consider the circumstances of any investor.