While there are no shortage of limitations and no “right” answers, portfolio theory nonetheless remains, rightly, the bedrock of traditional multi-asset portfolio design.
In this series of articles, I look at some of the key topics explored in my book “How to Invest With Exchange Traded Funds” that also underpin the portfolio design work Elston does for discretionary managers and financial advisers.
Portfolio theory in a nutshell
Portfolio theory, in a nutshell, is a framework as to how to construct an “optimised” portfolio using a range of asset classes, such as Equities, Bonds, Alternatives (neither equities nor bonds) and Cash. An “optimised” portfolio has the highest unit of potential return per unit of risk (volatility) taken.
The aim of a multi-asset portfolio is to maximise expected portfolio returns for a given level of portfolio risk, on the basis that risk and reward are the flipside of the same coin.
The introduction of “Alternative” assets, that are not correlated with equities or bonds (so that one “zigs” when the other “zags”), helps diversify portfolios, like a stabilizer. Done properly, this can help reduce portfolio volatility to less than the sum of its parts.
Whilst the framework of Modern Portfolio Theory was coined by Nobel laureate Harry Markowitz in 1952, the key assumptions for portfolios theory – namely which asset classes, their returns, risk and correlations are, by their nature, just estimates.
So using portfolio theory as a guide to designing portfolios is only as good as the quality of the inputs assumptions selected by the user. And those assumptions are ever-changing. Furthermore, the constraints imposed when designing or optimising a portfolio will determine the end shape of the portfolio for any given optimisation. And those constraints are subjective to the designer.
So portfolio design is part art, part science, and part common sense.
Whilst there are no shortage of limitations and no “right” answers, portfolio theory nonetheless remains, rightly, the bedrock of traditional multi-asset portfolio design.
What differentiates multi-asset portfolios?
A portfolio’s asset allocation is the key determinant of portfolio outcomes and the main driver of portfolio risk and return. Ensuring the asset allocation is aligned to an appropriate risk-return objective is therefore essential. Getting and keeping the asset allocation on track for the given objectives and constraints is how portfolio managers – whether of model portfolios or of multi-asset funds – can add most value for their clients.
There are no “secrets” to asset allocation in portfolio management. It is perhaps one of the most well-studied and researched fields of finance.
Stripping all the theory down to its bare bones, there are, in my view, three key decisions when designing multi-asset portfolios:
Strategic allocation is expected to answer the key questions of what are a portfolio’s objectives, and what are its constraints.
The mix of assets is defined such as to maximise the probability of achieving those objectives, subject to any specified constraints.
Objectives can be, for example:
Strategic allocations should be reviewed possibly each year and certainly not less than every 5 years. This is because assumptions change over time, all the time.
Static vs Dynamic
One of the key considerations when it comes to managing an allocation is to whether to adopt a static or dynamic approach.
A strategy with a “static” allocation, means the portfolios is rebalanced periodically back to the original strategic weights. The frequency of rebalancing can depend on the degree of “drift” that is allowed, but constrained by the frictional costs involved in implementing the rebalancing.
A strategy with a “dynamic” approach, means the asset allocation of the portfolios changes over time, and adapts to changing market or economic conditions. Dynamic or Tactical allocation, can be either with return-enhancing objective or a risk-reducing objective or both, or optimised to some other portfolio risk or return objective such as income yield.
For very long-term investors, such as endowment funds, a broadly static allocation approach will do just fine. Where very long-term time horizons are involved, the cost of trading may not be worthwhile. As time horizons shorten, the importance of a dynamic approach becomes increasingly important. Put simply, if you were investing for 50 years, tactical tweaks around the strategic allocation, won’t make as big a difference as if you were investing for just 5 years. This is because risk (as defined by volatility) is different for different time frames, and is higher for shorter time periods, and lower for longer time periods. In a way this is also just common sense. If you are saving up funds to buy a house, you will invest those funds differently if you are planning to buy a house in 3 years or 30 years. Time matters so much as it impacts objectives and constraints, as well as risk and return.
Managers need to make implementation decisions as regards how they access particular asset classes or exposures – with direct securities, higher cost active/non-index funds, or lower cost passive/index funds and ETFs. Fund level due diligence as regards underlying holdings, concentrations, round-trip dealing costs and internal and external fund liquidity profiles are key in this respect. The choice between direct equities, higher cost active funds or lower cost index funds is a key one and is the subject of a later article.
Types of multi-asset strategy
There is a broad range of multi-asset strategies available to investors, whose relevance depends on the investor’s needs and preferences. To self-directed investors, these multi-asset portfolios are made easier to access and monitor through multi-asset funds, many of which are themselves constructed wholly or partly with index funds and/or ETFs.
We categorise multi-asset funds into the following groups (using our own naming conventions based on design: these do not exist as official “multi-asset sectors”, unfortunately):
Relative risk strategies target a percentage allocation to equities so the risk and return of the strategy is in consistent relative proportion the (ever-changing) risk and return of the equity markets. This is the most common approach to multi-asset strategies. Put differently, asset weights drive portfolio risk. Examples include Vanguard LifeStrategy, HSBC Global Strategy and other traditional multi-asset funds.
Target risk strategies target a specific volatility level or range. This means the percentage allocation to equities is constantly changing to preserve a target volatility band. Put differently, portfolio risk objectives drive asset weights. Examples of this approach include BlackRock MyMap funds.
Target return strategies target a specific return level in excess of a benchmark rate e.g. LIBOR, and take the required risk to get there. This is good in theory for return targeting, but results are not guaranteed. Examples of this approach include funds in the Target Absolute Return sector, such as ASI Global Absolute Return.
Target Date Funds adapt an asset allocation over time from higher risk to lower, expecting regular withdrawals after the target date. This type of strategy works as “ready-made” age-based fund whose risk profile changes over time. Examples of target date funds include Vanguard Target Retirement Funds, and the Architas BirthStar Target Date Funds (managed by AllianceBernstein)*.
Target income funds target an absolute level of income or a target distribution yield. Examples of this type of fund include JPMorgan Multi-Asset Income.
Target Term Funds
These exist in the US, but not the UK, and are a type of fund that work like a bond: you invest a capital amount at the beginning, receive a regular distribution, and then receive a capital payment at the end of the target term.
For self-directed investors, choosing the approach that aligns best to your needs and requirements, and then selecting a fund within that sub-sector that has potential to deliver on those objectives – at good value for money – is the key decision for building a robust investment strategy.
The (lack of) secrets
The secret is, there are no secrets. Good portfolio design is about informed common senses. It means focusing on what will deliver on portfolio objectives and making sure those objectives are clearly identifiable by investors.
Designing and building your own multi-asset portfolio is interesting and rewarding. Equally there are a range of ready-made options to chose from. The most important question is to consider to what extent a strategy is consistent with your own needs and requirements.
* Note: funds referenced do not represent an endorsement or personal recommendation. Disclosure: until 2015, Elston was involved in the design and development of this fund range, but now receives no commercial benefit from these funds.
Traditionally, UK pension fund managers and UK private client managers alike would have a bias towards home (i.e. UK) equities. Why is this, what does the research say and what does recent experience show?
Understanding “home bias”
First of all, what do we mean by home bias? We define home bias is allocating substantially more to the investor’s “home” market, relative to its capitalisation-based weight in a global equity index. Given the UK’s weight in global (developed markets + emerging markets) equity indices is now approximately 4% (it has been on a steady drift lower), any allocation above that level can be considered a home bias, from a UK investor’s perspective.
Yet traditionally UK pension schemes and private client managers would split an equity allocation between broadly 50% UK and 50% international (ex-UK) equities. This represents a massive home equity bias, with a UK weight that is over 10x its market-cap based weight.
Why does this home bias exist?
The reasons given for such a massive home bias are typically the following:
We can look at each of these in turn.
Firstly, we would argue that investing in equities is not for currency/liability matching, but for return seeking and inflation beating: in which case, the broader the opportunity set, the greater the potential for returns.
Put differently, a UK only investor is not only wilfully or accidentally ignoring 96% of the opportunities available in equities, by value, but would also thereby miss out almost entirely on the technology revolution led by US companies, for example, or the demographic revolutions of emerging markets.
So whilst a home bias makes sense for a bond portfolio (matching changes in inflation and interest rates), a home bias for equities does not.
Secondly, whilst the largest UK companies within the FTSE 100 are indeed “global” in nature, the broader, and more diversified (by sector and constituents), all share index is not. Furthermore the sector allocation of the UK market is skewed by domestic giants, can be out of step with the sector allocation for world equity markets.
A UK equity bias is therefore a structural bias towards Consumer Staples, Materials and Energy, and a structural bias against Information Technology.
Fig. 1. Sector Comparison UK Equities relative to World Equities
Thirdly, whilst it is indeed true that UK managers will be able to get more access and insight to UK companies than, say, an overseas-based manager, for portfolio managers who focus on asset allocation over security selection, this access to management is less relevant and less valuable.
Whilst we can debate the detail of all three of these arguments, they are not individually or together enough to justify an allocation to UK equities that is over 10 times its market weight. This is not a question of a rational overweight, it’s simply an irrational bias.
What does the research say?
There has been extensive research into why individual investors and professional managers have a preference for creating an equity portfolio with a strong home bias.
French & Porterba (1991) observed the predominantly home equity bias of investors based on the domestic ownership shares (as at 1989) of the largest stock markets. In each case the high domestic ownership of each respective market implies a high home equity bias at that time: US (92.2%), Japan (95.7%), and the UK (92%), for example. In 1990 UK pension funds held 21% of their equity allocation in international equities from just 6% in 1979 (Howell & Cozzini 1990). Now the figure could be closer to 50%, or even higher. The shift away from home equity bias has been steady and pronounced in the UK institutional market, but is still ingrained.
However, it’s worth noting that subsequent home bias research is written in the US. Given the US represents approximately 66% of the world equity market (a share that has been steadily increasing), the central tenet of that research is that home-biased US managers miss out on the diversification benefits and increased opportunity set available from investing in markets outside the US. Hence home-bias for a US manager creates a smaller “skew” vs Global Equities than it does for a UK manager.
What is current practice?
Whilst the institutional UK managers have been gradually reducing home bias within equity allocations, what about UK retail portfolio managers?
We looked at the MSCI PIMFA Private Investor Indices – and predecessor indices – to gain an insight as to what current asset allocation practice looks like for UK-based managers in the retail market. These weightings of these indices are “determined by the PIMFA Private Indices Committee, which is responsible for regularly surveying PIMFA members and reflecting in each index the industry’s collective view for each strategy objective”.
Based on the “Balanced” index (and predecessor indices), within a typical balanced mandate, the allocation within the allocation equities have decreased from a 70/30 UK/international split in 2000, to a 48/52 split today (see Fig.2.). Whilst this reflects a reduction in the home equity bias, it is nonetheless a material bias towards UK equities by retail investment managers.
Fig.2. UK/international equity split within an indicative UK retail balanced mandate
Source: Elston research, FTSE data, MSCI data
In fairness, PIMFA has responded to this through the creation of a “Global Growth” index, which is 90% allocated to developed markets, and 10% allocated to emerging markets – so no UK home bias at all: but this is also a different risk profile to the Growth Index (100% equities, rather than 77.5% equities).
Zimbabwean investors go global – UK investors should too
We would make the case to advisers that if you were advising someone who lived in Zimbabwe, gut instinct would suggest that having the bulk of their equity allocation in Zimbabwean equities would feel like a poor and restrictive recommendation. After all, Zimbabwe makes up only a fraction of the global equity market.
Without wanting to do UK plc down, the same gut instinct should apply to UK equities. If the UK is only 4% of global equities – why allocate much more than that?
If you believe in equities for growth, it follows you believe in global equities to access that growth. Clients benefit from being shareholders in the changing mix of the world’s best and largest companies, not just the local champions.
What does recent experience showing
This debate was largely confined to theory given the relative stability of GBP to USD prior to Brexit. But given the dramatic currency weakness on the Brexit referendum, and the UK’s lack of exposure to the technology “winners” from the COVID-19 crisis, the disconnect between UK and Global Equity performance could not be more acute.
Over the 5 years to 30th June, the FTSE All Share has delivered an annualised return of +2.87%p.a. in GBP terms, and MSCI World has delivered +7.53%p.a. in USD terms. That represents the difference in the performance of the underlying securities within those markets. Adjusting for currency effect too, and MSCI World has delivered +12.79%p.a. in GBP terms: an approximately 10ppt outperformance annually for 5 years.
When expressed, in cumulative terms, the disconnect is more clear: over the 5 years to 30th June, the FTSE All Share has returned +15.22% in GBP terms, and MSCI World has delivered +43.80% in USD terms, and +82.66% in GBP terms: a 67.44% cumulative performance difference between those indexes, and the ETFs that track them.
Fig.3. World vs UK Equity performance, 5Y to June 2020, GBP terms
Source: FTSE All Share, MSCI World, Bloomberg data
Delivering good portfolio returns is less about picking individual winners within each stock market, but making sure you have access to the right asset classes for the right reasons. Index funds and ETFs are a low-cost, liquid and transparent way of accessing those asset classes.
UK multi-asset perspective
From a multi-asset perspective, the performance difference between MSCI PIMFA Global Growth (100% equity, no home bias), MSCI PIMFA Growth (77.5% equity, with home bias) and other risk profiles is presented in Fig.4. below.
Fig.4. MSCI PIMFA Private Investor Index Performance, 5Y to June 2020, GBP terms
Source: MSCI PIMFA Private Investor Indices (formerly WMA), Bloomberg data
The choice whether to embrace a UK home bias or avoid it has been critical and material and the main determinant of differences between multi-asset portfolio and multi-asset fund performance.
The lack of UK home equity bias, is one of the key underpins of strong performance of the popular HSBC Global Strategy Portfolios and Vanguard LifeStrategy range, for example.
A question of design
Our preference for avoiding entirely any UK home bias for equities (but not for bonds) underpinned the construct of multi-asset funds and multi-asset portfolios that we have developed with and for asset managers. End investors in those products have benefitted from that key design parameter.
Whilst we welcome managers launching global-bias multi-asset portfolios – it’s a bit late in the day as it won’t help their existing clients stuck in UK equities claw back the foregone performance of the last 5 years.
The irony is that one of the reasons for the persistence of home equity bias is sustained by asset allocation providers used by wealth managers to construct multi-asset funds and portfolios. A closer interrogation of those research firms’ methodologies, parameters and constraints is required to think what makes best sense for end investors.
Looking to create your own investment strategy?
Watch|Copy|Adapt our research portfolios
 French, Kenneth; Poterba, James (1991). "Investor Diversification and International Equity Markets". American Economic Review. 81 (2): 222–226. JSTOR 2006858
 We define the predecessor indices to the MSCI PIMFA Private Investor Indices as: MSCI WMA Private Investor Indices, FTSE WMA Private Investor Indices, FTSE APCIMS Private Investor Indices
Image credit: Lunar Dragoon
Commercial interest: Elston Consulting is a research and index provider promoting multi-asset research portfolios and indices. For more information see www.elstonetf.com