The most important thing: asset allocation

  • Written by 
  • 17/12/2015

In investing, as with all things in life, it is wise to devote your attention and time to getting the big things right, rather than squandering your energy on the trivial. Getting your portfolio to the right starting point in broad-brush terms is much more important than fiddling with the details. At Barclays we focus on the factors that have the biggest effect on our clients’ portfolios. And they don’t come any bigger than asset allocation.

What makes investors happy? 6 of 6 Earnings dispersion 4 of 6

Residents of the United States, please read this important information before proceeding

Asset allocation is defined by the proportion allotted to each category of investment. There are many ways of categorising investments, but any good method ensures that assets that share sufficiently similar characteristics are grouped together. For example, our moderately risky asset allocation is shown in Figure 1. This grouping allows us to analyse, understand and control a great deal of what really matters using a relatively small set of levers: the portfolio’s overall risk and return characteristics.

Sound investment management starts from an asset allocation to get the big picture right, and then fills in the details from there. However, even if your portfolio has not been constructed this way, it is worth noting that every investor has an asset allocation, whether they’ve thought about it or not – even a motley collection of investments adds up to a strategic big picture. And crucially, the wealth languishing uninvested in savings or cash accounts is part of your asset allocation and has a significant effect on risk and return expectations.

Asset allocation tells you the bulk of what you need to know to predict the performance1 of your portfolio, and understand its overall risk and return. A well-constructed allocation gives you the best chance of coping with the inevitable risks of investing and growing your wealth over time, and has a “major impact on whether you will meet your financial goal[s]”2.

Building the right levels of risk

Risk is the chance of an undesirable outcome. For investors, risk reflects the danger that you might not meet your financial goals, or that your portfolio drops in value at the point you need to draw on the money. Most investment managers measure a portfolio’s risk using the standard deviation of the statistical distribution of possible returns: a measure of how much returns might deviate from the average, or most likely, return. The more returns might differ from the average, the higher the risk. However, this measure makes an implicit judgement about what matters to investors that assumes a bizarre kind of symmetry: do steep gains really add to risk just as much as sharp losses do? Trying to reduce risk, measured this way, is effectively trying to “protect” investors from unusually high returns as well as low.

At Barclays, we look at things differently. Firstly, we appreciate that risk is one-sided, so we measure it accordingly to mirror what investors should wish to protect their portfolio from: the chance of bad outcomes, not good. Using our downside-focussed behavioural risk measure in building our asset allocations means we are more wary of assets which have a tendency to drop rapidly in value.

Secondly, we acknowledge that something is only ever risky insofar as it is risky for someone for whom some possible futures are better, and some worse. Without a preference for how the world turned out there would be no motivation to prevent or avoid certain outcomes. Risk is inherently subjective. In investing, higher returns are preferred to lower, but different individuals have different levels of Risk Tolerance: the degree to which they are motivated to avoid bad outcomes. Our Financial Personality Assessment is a scientifically robust method of uncovering an investor’s risk tolerance (amongst other aspects of financial personality), and we directly link this psychometric way of establishing preferences for risk to the optimal balance between risk and return in our asset allocations.

Thirdly, we do not target an arbitrary preset level of risk in the way many of our competitors do. We think it common sense to take on risk only to the extent that we think it’s going to be rewarded. For example, many approaches to asset allocation fix a risk level for each profile, for example 9% annual standard deviation for a moderate risk investor. They then aim to build the portfolio that delivers the highest returns for this 9% regardless of how the world changes. To us this makes no sense. In times when returns are lower and risks higher why would you keep targeting the same level of risk? Since at such times we get rewarded less for each unit of risk we take, it is logical to take less of it. Similarly when the environment rewards us more for each unit of risk, we should be prepared to take more. Our process allows the optimal risk level to fluctuate somewhat as circumstances shift.

The only free lunch: diversification

A robust asset allocation will use diversification to get the best trade-off between risk and return, effectively diluting risks and stabilising the portfolio. We can’t know the future, so we should avoid the hubris of thinking we can identify the perfect portfolio in advance. As the old saying goes, you shouldn’t put all your eggs in one basket.

A selection of varied investments will not all behave in the same manner at the same time, so getting a good mix diminishes risk when viewed in aggregate. Since as humans we care substantially more about avoiding worse-than-average outcomes than we do about achieving better-than-average outcomes, reducing the risk and thus limiting the range of outcomes with the same likely return is beneficial. We should only accept an increase in the range of potential outcomes if we’re paid for it: if the most likely return increases as a result.

Most portfolio methodologies assume that future risks, returns, and relationships between assets are precisely known – they perform fantastically in a world where their assumptions remain true, but often poorly in reality. We use a number of sophisticated techniques to reduce this reliance on our fallible knowledge of the future. One part of our process involves repeatedly “re-sampling” from all available data on past returns, putting them back together in new random combinations, effectively creating a huge number of possible histories – worlds that might have occurred, but didn’t. This reduces our reliance on what actually happened in the past - the one thing we know for certain is that the future won’t be the same as the past, so whilst we need to use the data we have, we should do everything we can to avoid optimising for a single history that is now behind us.

This creates an asset allocation that has greater resilience to different environments. The resulting portfolio is designed such that it will still do extremely well if our best guess of how investments behave in the future turns out to be right, but sacrifices a little of the best fit for this world to protect us, just in case we’re wrong. These portfolios are not just diversified across asset classes, but also across alternative views of history, making them more robust to a changing world.

In summary, an asset allocation is the high-level proportion of broad asset classes in a portfolio: every investor has one and should be aware of it. Your asset allocation should adapt to make most efficient use of the investment environment and stay tuned to your changing ambitions and circumstances.

Imagine a choice between 10 different investments, each most likely yielding a return for the next year of 5%, but each having a possible range of outcomes between -5% and 15%. If you chose only one of these investments as your entire portfolio you would have a substantial risk of getting an outcome of less than 5%. However, if you purchased equal amounts of all 10 investments for your portfolio, most likely your overall return would still be 5%, but since the investments all perform differently, the potential range of returns would be significantly dampened. This can be seen by understanding that for the portfolio with ten investments to return -5% overall, all ten of the investments would have to return -5% simultaneously, something vastly less likely than just one of them doing so.

1 Ibbotson, Roger G. “The importance of asset allocation.” Financial Analysts Journal 66.2 (2010): 18-20.
2 http://www.sec.gov/investor/pubs/assetallocation.htm