Asset allocation is the single most important factor in determining the returns of an investment portfolio

CFA Society UK

We would do well to remember that asset allocation is the most important thing for investors. A lot of the noise in the financial media is about selecting funds. Even I fall foul of that, talking about the funds I invest in, and not about my asset allocation. We live in a bizarro-world where the advice on which fund to buy is heavily regulated. But advice on asset allocation is almost ignored. The reality is how you allocate your investment funds across assets will be the biggest determinant of your investment return.

## What I do

Rather than talk about the theory, let’s go straight into what I do. Overall, I’m 90% equities, 10% bonds in my portfolio. Including my various cash balances, I’m probably about 80% equities, 10% bonds, 10% cash.

I’m much more invested in equities than the traditional 60/40 portfolio. That’s because I’m after long-term investment growth and, historically speaking, the greater the allocation to equities the greater the long-term growth return on your investment. Another reason is that I can stomach a lot of the volatility that comes with having an equities-dominated portfolio (more on that in a bit).

The last reason is that I’m more worried about **Inflation risk **and **Shortfall risk** than **Capital risk**.

## Hold on, there are three types of risk?

There are actually lots more types of risk. Inflation risk and Capital risk are the ‘big two’ for investors, but there is a third, often overlooked risk, Shortfall risk.

**Capital risk** is the risk that your investment may be worth less in the future than it is today.

**Inflation risk** is the risk that the purchasing power of your investment will be eroded by inflation.

**Shortfall risk** is the risk that your investment will fall short of the amount you require to make your financial goals.

In this respect, no one asset is ‘low risk’ in all regards. Cash has a low Capital risk, but high Inflation and Shortfall risk. Equities have a relatively high Capital risk, but a low Inflation risk. Over the long-term, it has a low Shortfall risk.

Circling back, for me, I’m more concerned that my investments will be eroded by inflation and that my portfolio will fall short of what I require to continue to be Financially Independent.

## Notice what I didn’t say

The words ‘volatility’ and ‘uncertainty’ are missing from those definitions. Yet, that is what lots of Financial Economics refers to when we are talking about risk (and I’ll be using that later too).

But I’ve never heard anybody say: “*Ooh I’m not going to invest in equities because the standard deviation is 18% compared to 12% for bonds*.” Rather, the conversation is usually framed around what risk investors are **comfortable** taking and what risk they can **afford** to take. You can see that ‘comfort’ marries up with Capital risk and affordability with Inflation risk. The final consideration is what level of risk is associated with the return I require to meet my financial goals. This risk **requirement** matches up with Shortfall risk.

Investing is not a ‘have your cake and eat it’ type of thing. You will have to trade-off between those three risk types. And what’s suitable will vary for each investor. Unfortunately, financial economics tends to ‘fall down’ when we start getting pesky people with all their emotions involved!

That’s why I don’t like the term ‘risk’. It’s a red herring. I prefer the term **risk tolerance** (considered with my other three factors: Time-horizon, access and affordability. See link).

## The efficient frontier

The fundamentals of investing are based on *Modern Portfolio* Theory (MPT)*. *The theory states that risk-averse investors can build an investment portfolio to optimise their expected return based on a given level of market risk. It’s often boiled down to: the trade-off between risk and reward. The risk we are talking about here is *volatility*; not the three risks (capital, inflation, shortfall) we’re talking about above. That’s because volatility is easily observed and measured – it’s much more difficult to calculate the three risks we care about (although some smart people come up with some ingenious attempts).

From MPT it’s possible to build an **efficient frontier**. The set of optimal portfolios offering the maximum possible for a given level of risk. Plotting these points creates a curve. The idea is that investing in any portfolio that is not on this curve is undesirable.

## A UK passive investor frontier

Using the amazing Portfolio Charts calculators I’ve created a set of efficient frontiers for a UK passive investor. I’ve looked at three simple portfolios:

- World Equities / Long-term duration Gilts
- World Equities / Intermediate duration Gilts
- World Equities / Short-term duration Gilts

I’ve created 11 points for each portfolio ranging from 100% in World Equities to 100% in Gilts. At each point, I plot the average annual return and the standard deviation ( how much the returns differ from the average).

Here are the charts:

(real average annual return with reinvestment on y-axis (note: not CAGR), standard deviation on x-axis, ticks at 10% intervals, data since 1970).

### World Equities / Long-term Gilts

### World Equities / Intermediate Gilts

### World Equities / Short term Gilts

At 100% in world equities, the average annual return is just shy of 7% with a standard deviation of just under 18%. Each portfolio curves downwards as we move from Equities to Gilts. With the return falling sharpest as we move out of Equities into short-term Gilts (T-bill equivalents). This is more obvious when we plot the curves on top of each other.

## Using the Efficient Frontier

There are two ways of using the Efficient Frontier.

The first is to look at whether your portfolio has a lower return than the frontier curve at its level of standard deviation. For example, if our portfolio is returning 5% at a standard deviation of 15%, Modern Portfolio Theory tells us that our portfolio is sub-optimal. We would benefit by moving into the portfolio at the frontier.

The second use is to ride along the frontier to the most efficient point on the curve. This is where use diversification and derisking to get a ‘better’ risk-adjusted return. If we draw a line from the 100% equity point, to the 100% bond point, this the frontier we’d get if there was no diversification benefit from holding equities and bonds together. So when we ride along the efficient frontier, part of our change is from derisking – accepting a lower return for less volatility and part is from diversification – getting a higher return because we are putting our eggs in multiple baskets:

## Finding the sweet spot

Using this technique we can find the sweet spot – the mix of risky assets that delivers the best risk-adjusted return. To do this, we need to find the return on a riskless asset. The closest we have is cash. According to the Barclays Gilt Survey 2016, the 50yr real return on cash is 1.4% per annum. If we plot this on our graph and find where it meets a tangent with our efficient frontier we can find the sweet spot:

This chart shows that something around a 50/50 World Equity/Intermediate Bond portfolio would give the best risk-adjusted returns (i.e. 50% VWRL, 50% VGOV). Thus my somewhat flippant phrase in my How I invest my money post: “*There is good evidence that on a risk-adjusted basis a more even split between bonds and equities is superior (in risk-adjusted returns).*”

But the return on cash varies significantly over time, were one to look at the 2010 Barclays Survey, the real return was 1.9%. If we plot that, we get a much different answer:

Our graph now tells us that something around 60/40 to 70/30 is the sweet spot.

## But…

Cash isn’t riskless. In terms of *nominal* volatility it might be, but once we account for inflation cash is anything but risk-free. We must also bear in mind that we are, generally speaking, less fussed about volatility than we are about capital loss and inflation erosion. So even if we’ve found the ‘sweet spot’ we still need to tailor it around our needs for our risk comfort, what risk we can afford and what risk we are required to take.

At this point, you’re probably cursing me for making you go through all that math for nothing. But I hope it underlies an important point about investing: you can’t invest based on numbers alone, you must also consider the qualitative side too.

## A final chart

I have one final chart for you. This is a look at the spread of annual returns based on my portfolio allocation. Again, courtesy of the amazing Portfolio Charts:

(again, real average annual returns, since 1970)

I’ve input my asset allocations based on my How I invest my money post. First off, you can see that the return and standard deviation are relatively high, well towards the top right of any of the frontier curves above. You’ll also notice that in 27 years out 100 this portfolio would lose money in real terms.

Another interesting thing is that 77% of the time did the portfolio returned less than 2.2% or more than 12.2%; which highlights why you have to take average returns with a pinch of salt – investors rarely get the ‘average return’.

Lastly, you’ll notice two huge outlier returns. The loss is -41%. That’s right, you would have lost almost half the value of your portfolio in one year. That year was 1974. However, 1975 saw the other massive outlier, a return of 45%. (note that the UK data is patchy pre-1975, thus the shading).

Would you have been able to stomach such a massive drop? Well if you’d have been invested in this portfolio 10 years ago, you’d have seen a drop of 23% in 2008, followed by a gain of 21% in 2009.

I hope I wouldn’t flip my lid and keep cool in such market turmoil. Though that is easy to say before it happens. I know that one day, there will be such an occurrence. Hopefully, by putting this out here, it’ll keep me honest for that day.

All the best,

Young FI Guy

### References:

Please do visit the amazing Portfolio Charts which is an excellent data visualisation resource for investors: portfoliocharts.com

Part of the inspiration for this post was from Karsten at Early Retirement Now who created efficient frontiers for US investors to ask the question of how much bonds diversify from equities: https://earlyretirementnow.com/2017/04/26/have-bonds-lost-their-diversification-potential/

The other inspiration is Monevator’s piece on UK historical asset returns and asset allocation: http://monevator.com/uk-historical-asset-class-returns/

If you want to, you can look at the underlying excel spreadsheet: Link

Thanks for this – efficient frontiers are are intellectually appealing but sensitive to assumptions.

A few months ago I started thinking about my portfolio (trackers, Deferred FS pension, BTL, P2P, pref shares etc) and whether there was any way to construct an EF. Most EF seem to have only 2 categories and building a more complex version would involve estimating a correlation matrix between various classes (pretty subjective stuff). Do you have any views or ideas on this?

B

Ps adding leverage to the sweet spot is tempting!

Hey Boltt.

You’re exactly right, as with lots of financial economics, the results are very sensitive to assumptions. There’s only so much I can cram into one post, but I tried to leave two examples in there with the different curves for ostensibly ‘similar’ equity/bond portfolios and with the impact of different cash returns.

It would be difficult to create a more than 2 asset efficient frontier. I mean you can do it, but when you start going into 3d/4d(!) space a lot of the benefit (easy visualisation) is lost! One potential option would be to plot the curve of the Vanguard LifeStrategy funds (or a similar set of risk-spectrum funds, bear in mind though, very small sample size). You could also use the https://portfoliocharts.com/portfolio/annual-returns/ calculator that I used, making subjective calls about what the different points on your chart would be. But I think in terms of the Efficient Frontier and MPT we’ve got to remember it’s not about mathematical precision (and I say that as an Accountant and Math grad!) – the main thing is to understand the two important lessons on the benefit of diversification and trade-off between risk and reward.

And yes, leverage is very tempting (in a way, this is what Mrs YFG and I do now as we have a mortgage). Early Retirement Now has an excellent post (and much better than anything I could ever write) on this exact subject: https://earlyretirementnow.com/2016/07/20/lower-risk-through-leverage/

Hi,

Great blog – have been an avid follower for a while.

A question on your own rationale for your c 80:10:10 equity:bonds:cash split. You state that equities will generate the highest long term return (as a tech analyst, I agree!). You also state that you can stomach the higher volatility of equities.

Q: Why do you hold bonds at all?

Whilst I appreciate that the bonds should give you a smoother ride, you also accept that you will (probably) have a lower return over the long run. Yet you repeatedly comment that you are more concerned with inflation and shortfall risk. Is this logically consistent?

If you had a shorter investment horizon, then I completely buy holding fixed income. You, however, have an investing horizon of 50+ years from now, over which time, as we all know, a minor change in annual returns will generate a huge difference in overall returns.

Could you not hold 100% equities (at least in the mid-term), with your additional 10% cash buffer to cover daily expenses etc if required?

Do you hope to rebalance the fixed income exposure back into equities when they’re cheap (ie being 100% equities for a short while)?

Alternatively, is it to minimize psychological impacts if (/when) equities take a tumble?

I don’t mean this as a criticism, I’m genuinely curious about your thought process – in part to guide my own decision making (ie whether I should add gilts etc to my own modest equity holdings).

Thanks again for the blog!

Hi Young Analyst, thank you for commenting and for the kind words!

Very good question, and probably worthy of a post of its own. Apologises in advance for my rambling answer. Here are three reasons why I hold some bonds and cash:

1. Sequence of returns risk – it’s the #1 enemy for early retirees, and the data suggests you are much more vulnerable to it if you have a 100% equities portfolio. From the analysis I’ve seen, somewhere between 60/40 and 80/20 equities/bonds seems like the best asset allocation for safe withdrawal rates. A really good visual explanation can be found at portfolio charts (again!): https://portfoliocharts.com/2015/11/17/how-safe-withdrawal-rates-work/

2. Volatility – Even a small step down from 100% equities to 90%/80% significantly reduces volatility, making it (arguably) worthwhile. For example, just using the charts above, moving from 100% to 90% with intermediate bonds reduces the standard deviation by 10%, and almost 20% if we got to 80%. The effect is even stronger for cash. (note: however, there is a valid criticism that bonds don’t diversify you that much away from equities, especially at the 80/90% equity allocation, as the volatility from the equities ‘swamps’ the bond volatility).

3. Psychology – somewhere between 60/40 and 80/20 appears to be around the best risk-adjusted return, so in a way, I’m deviating from what’s ‘best’. I could deviate some more, but my brain kinda says: “that’s enough, stop being a smart-ass”.

There are plenty more reasons, and I’m sure lots of them are contradictory and full of thinking errors! I’ve kind of reached a point where my asset allocation “is good enough”. It’s pretty safe from the big risks, so I don’t need to worry about it too much. Does that make sense?

p.s. I don’t really buy into the ‘valuation’ style asset allocation stuff. I know a lot of people do, but I find CAPE unconvincing (perhaps from my bias of being a professional valuer and finding it to be too noisy). Most valuation strategies also require a heavily-subjective judgment process to know when to buy/sell out. I’m sure some people get it to work marvellously for them, but I’m incredibly lazy. So I’m happy with bi-annual rebalancing.

Great article YFG, I’m still trying to get my head round MPT, and the Efficient Frontier particularly took some head-scratching. I didn’t understand this at first glance, but on a second focussed read-through it all fell into place.

I do find it interesting you’re using such an allocation whilst also opting for a drawdown plan. Do you think this could ever change?