EquityMar 20 2019

For & Against: Does diversification still work?

twitter-iconfacebook-iconlinkedin-iconmail-iconprint-icon
Search supported by
For & Against: Does diversification still work?

Convention says diversification is a basic aspect of risk aversion, but the truth may not be that simple.

For: Andy Gadd is head of research at Lighthouse Group 

Everyone knows and surely accepts that at its most basic level diversification reduces risk. If you own one stock then that will be risky, while owning two different stocks is generally less risky.

That is intuitively correct and makes even more sense if the two stocks are what is known as ‘negatively correlated’.

By this I mean that the risk of owning one stock can be significantly reduced by adding another stock, especially if that additional stock behaves differently to the original stock in various situations.

When considering investment options, in my experience asset allocation and diversification are often, wrongly, used interchangeably.

Correlation basically describes the connection or lack of connection between different investment options.  

A perfect negative correlation would mean that if the price of one stock goes up then the price of another stock is likely to go down.

So rather than owning just one stock that makes, let’s say ice cream, you might own two stocks, one an ice cream maker and one an umbrella maker. The result is that when the sun shines the ice cream business makes money, while when it rains the umbrella manufacturer makes money and vice versa.  

In investment terms this is known as ‘specific’ or ‘unsystematic’ risk – and in the context of an investment portfolio, unsystematic risk can definitely be reduced through diversification.

The companion of unsystematic risk is ‘systematic risk’, which is also known as ‘market risk’ or by some as ‘undiversifiable risk’.  

The argument is that this type of risk cannot be eliminated by diversifying within an asset class, but rather only if you are able to find an asset class that is uncorrelated to other asset classes.

When considering investment options, in my experience asset allocation and diversification are often, wrongly, used interchangeably.

The fact is that even if you are 100 per cent invested in cash, that is still a form of asset allocation. It is not, however, diversification.

When considering diversification and asset allocation one might aim to construct a portfolio that is not only diverse in terms of the underlying assets held within asset classes but ideally holds different asset classes, some of which are potentially uncorrelated so that in various circumstances the positive performance of certain asset classes balance out the negative performance of others, with the aim of smoothing out overall portfolio returns within a particular risk budget.

Ultimately, a correlation matrix is based on past performance – and as we now all fully accept, especially I think since 2008, past performance is no guarantee of the future, especially as we work through what I would describe as the greatest economic ‘experiment’ in history of quantitative easing and tightening.  

This means that although it may currently be more difficult to find asset classes with negative correlations, I believe it is still possible to find asset classes with low correlations, and that is likely to prove to be advantageous.

All investing is fundamentally about the trade-off between risk and reward; you need to accept enough risk to achieve your desired returns but not so much that you stay awake at night worrying.  

Whatever anybody says, for me diversification remains key to managing risk.

Against: Mark Polson is principal of the Lang Cat

I will start this off by noting that just as Brandon Flowers once sang that he had soul but he was not a soldier, I have got higher economics and have not got a clue about macro-economics in 2019.

Nonetheless, there are some things going on that are worrying me a bit, and none of the people I have asked who understand more than me have been able to allay my worries.

A fundamental tenet of portfolio construction for decades has been the equity-bond correlation – as equity prices rise, bond yields also rise, reflecting a decline in their price.

The change in correlation – in the United States at least – appears to be continuing, for reasons that clever people argue over.

To put it another way, equities and bonds tend not to tank at the same time, and so bonds act as a great diversifier to equities in portfolios.

This accepted wisdom is reflected in virtually every centralised investment proposition I see, where a predetermined split of (usually) large-cap equity allocations is gradually reduced as you go down the seven or 10 risk grades in favour of (usually) investment-grade bonds.

Again, this reflects decades of experience, and so it must be right, right?

Well, maybe not. In the first quarter of 2018, equities and bonds fell in unison – the traditional correlation was reversed. This freaked the life out of the investment industry, and continues to do so.

The change in correlation – in the United States at least – appears to be continuing, for reasons that clever people argue over.

Now, I do not know if that change in correlation will affect advisers’ model portfolios or not. But what I do know is that most advisory models I see work in the same way, with roughly the same increases in bond exposure as you go down the risk spectrum. The bond funds people buy tend to be commonly held, and all in all it all looks a bit samey.

That is fine if the old correlations hold true. But what if they do not?

If everything heads the same way at the same time – as it did a decade ago – we are looking at a change that will fracture the way portfolios are built in retail intermediated investment.

Where will advisers go? What tactical changes will they make? 

You all tell your clients they should not worry about short-term changes, because you do not either.

But every long-term change starts as a short-term one. And when, or if, there is a shift, advisory models in particular are slow and cumbersome to adjust – just think about getting every single client’s written permission to make such a major change in their portfolio. 

It is the everyday things; the things that become a fundamental part of the way we do business, that we seldom question on an ongoing basis because the pain of doing so is just too great.

But that just means if something profound does change, we are spectacularly ill-equipped to deal with the fallout.

It is incumbent on anyone who assumes the mantle of running money for clients – which of course is something that is quite a different discipline from the practice of financial planning – to not only question asset allocations and the skill of fund managers, but also the underlying assumptions that are the very building blocks of model portfolios.

Look, it might all be fine.

But I wonder if it is not a good idea to plan out a scenario for what happens if we do hit harder times, and if the foundations upon which we have built so many centralised investment propositions turn out to be based on assumptions that no longer hold true.