This article is addressed to individuals in the decumulation or drawdown phase, not to pension fund managers. But most investment principles for individuals were developed by applying pension fund principles to the limited case of individuals: for example, what would a fund do if it were down to its final member?
So I will trace the history of asset allocation as it relates to pension funds, and then extrapolate to individuals.
Where the history of asset allocation takes us
My logic here is as complete as possible but some points are in the footnotes to keep the flow of the article as smooth as possible.
1. We need growth assets
At the two extremes, if you have so much money that you don’t need growth, or so little that even a lot of growth won’t get you to your target, then none of this applies to you. So I’m only addressing those who absolutely need future asset growth (1). They either don’t have quite enough to lock in their desired spending for their longevity horizon, or they’re only slightly above the necessary amount.
So we start with growth-seeking assets as the base of the portfolio. The simplest approach is to buy a stock index fund, in the form of a managed fund or an ETF. This was traditionally a local-country index fund, but these days it’s more likely to be a global one (2). In the old days this needed to be a simple index fund, but these days it’s also possible to get an index fund devoted to ESG characteristics or other factors.
2. Growth assets bring problems
We know, however, that while growth is highly likely over the long term, it is not guaranteed – that’s what William Bernstein calls 'deep risk'. And this uncertainty leads to short-term price volatility, which is bad when we’re forced to interact with the market at an adverse time – what Bernstein calls 'shallow risk (3). Deep risk is unavoidable. Shallow risk is either avoidable or capable of being acceptably mitigated, and that’s where fixed income comes in.
3. Pension funds and volatility
Once upon a time, there used to be defined benefit (DB) pension plans flourishing on this earth. (Yes, I know this rare creature is not yet extinct). The contributions required to finance the promised benefits are necessarily unknown, because they depend heavily on future investment returns. The practice used to be to update the estimated required contributions every three years. This meant that three-year market volatility was reflected in contribution volatility.
To reduce the contribution volatility to what was considered acceptable by the plan’s guarantor, the practice arose of diluting the growth exposure with 40% of fixed income assets, like bonds and mortgages. These had the effect of providing some investment return, though less than that expected from the stocks, while reducing the portfolio’s overall volatility and therefore the contribution volatility.
That 60/40 allocation become the default approach for many, many years (4). In fact it got to be a joke: “No matter what the investment question is, the answer is always 60/40.”
4. Pension funds discovered a new need
This 60/40 policy worked fine when new contributions exceeded benefit payments, because there was no forced need to sell assets: the benefits could be paid from the new contributions. But if a DB plan closed to new entrants, regular contributions in the current year became smaller and then vanished, and asset sales became necessary to pay benefits. And any forced asset sale right after a market decline locked in a permanent loss.
So funds now had a new need: make sure there’s enough cash to pay the benefits, at least for a few years. And that led in turn to a realignment of the fixed income assets, to structure them in such a way that their combined interest and maturity payments matched the benefit cash flows required. This became known as 'liability-driven investing' (LDI for short), and fixed income now took on a maturity structure matched to the liability cash flow, rather than just accommodating the idea that a fixed income index fund would do the volatility-reducing job.
5. What is different today?
In my early days, fixed income at least promised a relatively high stream of interest payments. Today those interest payments are very low, in fact edging closer and closer to zero, sometimes even negative when considered in real (after-inflation) terms. What difference does that make?
My answer: None. That’s just tough.
You still need to reduce stock volatility, you still need to match benefit cash flows. The two needs haven’t changed. All that has happened is that the reward expected in the old days from interest payments has essentially gone (5). Too bad, but life has simply become more expensive.
And there isn’t an adequate substitute for fixed income. You can consider high-dividend stocks, but (a) they don’t reduce the volatility problem and (b) the dividends make only a small dent in the cash flow needed to pay benefits. So they don’t solve either of the two essential problems.
This is why so-called alternative assets, such as real estate and private equity, have become so popular. Not being traded daily, they aren’t subject to the volatility of traded stocks, so they help with the volatility problem (although low volatility is an artificial characteristic here but I won’t pursue that angle). But they definitely don’t solve the LDI problem.
So the two problems that fixed income used to happily solve still exist. That’s why I say that the role of fixed income is the same as before. What’s changed is the degree of comfort with the solution. Too bad. It’s like living in the same world as before, but with higher taxes. Your problems are the same, the solutions are largely the same, you’re just worse off than before.
(However, see 7. below for another way to deal with the new conditions).
6. What about individuals?
I’ll apply all of this to individuals in the drawdown or decumulation stage of their financial lives. They have accumulated assets. Individuals need growth, and they also need predictable cash flow for a few years, because they want to avoid the need to sell stocks right after a market decline.
The ratio of the two parts (growth and income) may or may not be 60/40, but they will certainly reflect the lifestyle risk tolerance of the couple or person involved.
That risk, of having to sell into a market decline, is now well known. It means that, when you think of the sequence of the volatile stock returns over time, you’d rather have high returns in the early years, when your assets are at their peak, than in the later years, when your assets have been mostly spent.
When Bob Collie, Matt Smith and I wrote our book (6) in 2008, we couldn’t find a recognised name for this risk. I think it was Matt who came up with 'sequential risk'. Today there’s a standard term: 'sequence-of-returns risk'.
The notion is that a drawdown portfolio should consist of a base of growth-seeking assets, on top of which you place fixed income (typically bank securities) that generates the cash you need for a few years. If the stock market falls, you have mitigated the 'shallow risk', giving time until the market recovers.
How many years? As I said, that depends on your risk tolerance. My wife and I use five years. Why? Because historically, stock markets have recovered enough to provide at least a 0% real (after inflation) annualised return over five-year periods – three-quarters of the time. The remaining 25% of the time is our acceptable shallow risk. Others I know use three years or even one year. In the opposite direction, one particularly risk-averse friend prefers 10 years.
Whatever the time period chosen, that shallow risk is mitigated but not eliminated. The dividends on the stock portion of our portfolio don’t come close to meeting our annual cash needs, though they help.
And now that our bank securities provide hardly more than zero interest (actually negative, in real terms), well, that’s just too bad for us. It doesn’t change the nature of our problem, it doesn’t change our solution, it just means we’re worse off than when interest rates were higher.
7. But wait – there’s more
Clearly, the world has changed, or the parameters that define the problem have changed. My own reaction has been to accept the reduction in return. But there’s another valid way to react, and that’s to decide that you’re willing to accept more risk than before, in order to get your expected return back to where it used to be.
In other words, you’re thinking: it’s a new world, and a new me, willing to take more risk than before. Then you’ll reduce the amount of fixed income relative to what you used to be comfortable with. At the extreme, you’d reduce it to zero; in which case, something like high-dividend stocks would be your preferred route. Expected return goes up again, risk (both deep risk and shallow risk) go up too, but you’re willing to bear it. That’s a valid reaction too.
No magic solutions even for a new you
I think of low-to-zero interest rates as a world of higher taxation. There are no magic new solutions to the problems. We’re just worse off than before. Or, in search of a higher return, the new you can also accept higher risk, with the inevitable consequences.
Footnotes
1. Strictly we need a return rather than growth. But I’ve used the word 'growth' extensively in the past that I’ll stick with it. Fixed income gives you a return too, of course. Typically stocks give you a higher return over the long term, with higher volatility and uncertainty.
2. I’ll skip the stuff on currency risk, as it’s not relevant here. The same goes for active management – a separate issue.
3. https://www.amazon.ca/Deep-Risk-History-Portfolio-Investing-ebook/dp/B00EV25GAM
4. I noticed, practising in three countries, that while the US default was 60/40, it tended to be 70/30 in the UK and 50/50 in Canada. To some extent this may have reflected different fixed income yields. But the same principle applied everywhere: how much contribution volatility was acceptable?
5. I’m assuming that, with the decline in fixed income yields, the expected return from stocks has declined by the same amount: in other words, that the equity risk premium has not changed. This is consistent with the rise in the stock index that was triggered by the decline in yields and becomes the new base for future projections. With the same equity risk premium, if your risk tolerance doesn’t change, your asset allocation should also not change. Only your expected return changes – downwards.
6. https://www.amazon.ca/Retirement-Plan-Solution-Reinvention-Contribution/dp/047039885X
Don Ezra has an extensive background in investing and consulting and is also a widely-published author. His current writing project, blog posts at www.donezra.com, is focused on helping people prepare for a happy, financially secure life after they finish full-time work. This article is general information and does not consider the circumstances of any investor.