Adjusting the sails

23rd February 2022

The ambition of our investment management process is to invest in a way that achieves specific objectives. At the start of every calendar year, we take the opportunity to reassess and re-examine the data that informs our decisions. We start with a clean slate, building fresh sets of analyses which together help us answer the question: where and how should I invest?

 

“The pessimist complains about the wind, the optimist expects it to change, the realist adjusts the sails” – William Arthur Ward

 

Data diving

The main building blocks of our annual strategy-setting exercise are known as capital market assumptions (CMAs). These comprise sets of risk and return expectations for a wide range of asset classes (equities, bonds, commodities, property and so on).

They are an attempt to give an idea of the characteristics and behaviour of certain investments over the long-term (the next decade), both in isolation and when considered in relation to each other.

Our CMAs are formed using data-driven models and by leveraging academic and industry research. This approach enables us to combine both historic trends and forward-looking projections.

Different asset classes can be assessed on different attributes within a shifting macroeconomic environment. For instance, the return outlook for equities accounts for both income (dividend yield) and capital appreciation (earnings growth and valuation changes) over time.

When it comes to risk – or ‘volatility’ – we use a two-step methodology. The first step is to analyse multi-decade historical data to see how volatile each asset class has been. This enables us to account for different business cycles, including booming markets as well as recessions.

The second step is to combine historical data with projections for asset class volatility over the next decade. Doing so relies on understanding the drivers of uncertainty whilst integrating statistical theory to make inferences about financial relationships. In essence, risk can be forecasted thanks to some persistent statistical properties.

In order to build a diversified multi-asset and goal-oriented portfolio we need to evaluate not just the risk of each asset class but the correlations between them.

Correlations – the extent to which asset classes move in the same direction – play an important part in constructing a portfolio that seeks to maximise the benefits of diversification.

We calculate correlation coefficients using multiple decades of historical monthly return data covering a large pool of investment indices. This is to include a range of business cycles and market conditions, which offers statistical benefits. We opt for a minimum data span of 20 years; where there is insufficient data, we use a back-filling algorithm to estimate index movements.

 

Combining assets

Each asset class behaves differently, as defined by their long-term returns, short term variations in value (volatility) and their behaviour relative to each other (correlation).

We use Modern Portfolio Theory (MPT) to analyse which different mixes of assets will create the most efficient portfolios (i.e. the highest return for a given level of risk). This analysis is based on our long-term CMAs.

MPT assumes that no rational investor would want to take on any more risk than they strictly need. As such, higher risk must be compensated with higher returns. Crucially, in determining this trade-off, the MPT approach focuses not just on the individual risk of each asset class but the combined risk of all assets in a portfolio.

The process is sometimes known as Mean-Variance Optimisation. It shows that adding an asset that is individually very volatile can actually reduce the overall volatility of a portfolio if it has a low correlation to other constituents.

This is an extension of the concept of effective diversification. Holding a range of investments for the purpose of reducing risk is only useful if their returns counter-balance each other in some way. The classic example of this being stocks (which do well when times are good) and bonds (which do well when the market turns sour).

 

Strategic and Tactical Allocation

Having amassed our historical and forward-looking data, and calculated the most efficient mix of assets, we are left with a broad target allocation. Due to its long-term focus, we call this our Strategic Asset Allocation (SAA).

The SAA for an investor with a high return objective, and a long time horizon, might contain more assets that are higher risk (e.g. stocks or equities) and fewer lower risk assets (e.g. bonds). In reality, our toolkit contains a host of major global asset classes including equities, fixed income and alternative assets.

The SAA determines most of the allocation underpinning our strategies. But to bridge the gap between long-term assumptions and current market conditions, we also make shorter-term decisions within our Tactical Asset Allocation (TAA).

The tactical element acts as an overlay to the main strategic portfolio. It is implemented by investing in particular asset classes, regions, sectors, investing styles, and themes. The aim is to capture opportunities created by changes or trends in financial markets.

Our approach is best illustrated with a recent example: we moved part of our TAA into large-cap developed market equities shortly after the worst of the pandemic-driven volatility. This was based on the observation that developed nations were better placed to spearhead the global economic recovery thanks to access to capital and medical infrastructure. We have recently taken the view that the rebound is set to broaden, so we are increasing the tactical exposure to Emerging Markets.

 

Selecting investments

Knowing which asset class to invest in is only half of the story. Once we have determined a target SAA and TAA we implement our investment selection process.

Data stays at the heart of the decision making. We assess best-in-class asset managers with proven track records against a range of quantitative and qualitative metrics. This rigorous selection process identifies investments that can fulfil our required contributions to overall performance and risk.

 

Summary

Each year, we combine many years of historical data with forward-looking analysis to generate the risk, return and correlation profiles for a range of asset classes. Using statistical theory, we calculate the best mix of those assets to maximise the return at different levels of risk, and new strategies are formed. In this way, we can adapt to changing market conditions and expectations by periodically adjusting the sails, keeping us on a path to achieving long-term objectives.

 

 

<< Back to Insights