AJ Bell's head of passive portfolios Matt Brennan speaks to
deputy editor Tom Eckett on disliking backtests, avoiding multi-factor products and the reasons why the smart beta story has much further to run.
Brennan is responsible for managing the AJ Bell's passive portfolios having joined the firm in 2017. Prior to this, he spent three years at Brown Shipley where he held a number of roles including senior fund manager and head of fixed income research.
Tom Eckett: Do you use smart beta or factor products within your clients’ portfolios?
Matt Brennan: Although it is a fairly clichéd answer, we do not characterise investments as ‘passive’ and ‘smart beta’. Instead, we prefer to just consider rules-based investments.
To that note, we are agnostic if the strategy is straight up market cap weighted or takes a smarter approach to index constitution.
However, one important factor we do look at when using more complex strategies is cost.
Do the extra costs generate better risk-adjusted returns?
Areas where we consistently use smart beta are where the portfolio objective is a bit more complex than maximising total returns.
For example, when the portfolio has an income objective we are more likely to use smart beta strategies. This is because we want to make sure that the income is not just higher than average but sustainable in the long term.
Other areas include thematic investments and when we are trying to achieve alternative like exposure.
How much of your portfolios does smart beta typically make up?
Our ‘standard’ growth portfolios are still largely dominated by simple strategies. We do however hold a small portion of smart beta strategies to enact our property view.
However, over time we expect to see the number of smart beta strategies we use to increase, but it is really just a case of waiting for track records to build (we hate back-testing!) and for costs to come down. For our income portfolios, the majority of the equity component is made up of more complex index strategies.
How do you view smart beta/factor-based ETFs?
Smart beta is such a broad term and probably encompasses both elements.
For example, a fixed income ETF that adds in liquidity screens is certainly closer to ‘pure passive’, whereas a robotics ETF that has a rules-based strategy blended with a discretionary element, such as a panel of experts, is much closer to the active side of the spectrum.
It is therefore very important to get under the hood of the investment.
Which parts of the smart beta/factor-based spectrum interest you most at the moment?
The biggest challenge for a multi-asset fund manager using passive investments only is to achieve alternative like exposure in the portfolio to improve the risk-return characteristics.
In the US, we have seen several strategies looking to replicate hedge fund returns, and in Europe we have seen products using managed futures, going long/short or trying to replicate physical property or infrastructure investments. We feel these are still at a nascent stage but is something we are watching closely.
We also feel it is one of the last barriers that stop people switching from an active strategy to passive, so we are advocates of further development in this space.
On the other side of the coin, we really do not like momentum-based strategies. It feels like it is something that works, until it does not – they usually involve high turnover, which is against our philosophy of what a ‘passive’ investment style should look like.
When you focus on a particular smart beta product to invest in what factors do you take into account?
Cost is a big factor – what is the strategy giving me access to and does this justify the premium I am paying compared to a simpler strategy – could I replicate the factor using a combination of other ETFs.
It is important to us that the strategy is not a ‘black box’. We want to make sure we can understand how the factor is systematically harvested, and more importantly, the factor makes sense (e.g. we would expect less liquid stocks to outperform, however, why should momentum work – just because a stock has performed well recently, there is no justification for further outperformance).
Alongside smart beta, we have also seen the rise of thematic-based investing using ETFs.
Does this interest you?
In our higher risk portfolios, we have a small allocation to an automation and robotics ETF. We find the use of these products interesting on two levels:
• On a standalone basis, it allows us to access strategies that will give us access to long term growth trends.
• On a portfolio basis, they tend to improve the overall risk-return profile
However, a word of caution, something like an E-vehicle ETF often has high exposure to highly cyclical companies, such as traditional car manufacturers pivoting towards electrical vehicles.
It is therefore important to ensure the thematic ETF actually gives you exposure to the theme!
Are you concerned by the recurring accusations of hacking and data mining levelled at all factors and smart beta strategies?
Data mining is one of our biggest concern, and therefore we always ask ourselves before investing if the factor actually makes sense. We usually avoid any strategy that is ‘multi-factor’.
We really see factors as a tool to build more diversified portfolios, we are sceptical about outperformance.
For example, having an allocation to both quality companies and value companies, and thinking of them separately allows us to deliver performance depending on which strategy is in style, lowering volatility. It may not, however, lead to outperformance. In sample backtesting, it is a statistical travesty!
How do you engage with clients about smart beta?
We recently dropped the ‘passive’ name from our fund range to allow us to focus more on ‘rules-based strategies’, encompassing smart beta.
We avoid using this term, and instead focus on education, explaining exactly why we are using the more complex strategies, and what extra value they bring.
Clients are only concerned if the cost of investment significantly increases, hence our focus on product cost.
Are there any specific areas where you would like to see new products emerge?
Income investing is still at an early stage. We really like products that add a quality tilt to the income generation, however, the number of investment options remain few and far between.
Costs still need to come down, and more education is required, however, given low bond yields, we feel this is a growth area.
Does multi-factor investing interest you?
No – our motto is to build simple, transparent low-cost solutions. It does not tick any of these boxes.
By 2025 do you think you will be making extensive use of smart beta products and factor ETFs?
The best analogy I have here (and it still a poor one!) is Google. In the 90s, it was a very crude search engine, returning results for the exact phrase you typed.
Over the next 25 years, it has transformed into a much more complex beast but is still ultimately rules driven. It can suggest results, recognise errors and identify patterns.
‘Passive investing’ is the Google of the late 90s – we have seen strong growth, but the rule sets are still pretty crude.
Over time strategies will develop that are little bit more nuanced and learn from previous mistakes, however, they will still be quantitative strategies.
This is where smart beta will sit. Ultimately, most of our portfolios will have some element of a smarter rule set.
This article first appeared in the Q1 2020 edition of Beyond Beta, the world’s only smart beta publication. To receive a full copy, click here.
To read the previous edition of ’60 seconds with the buy-side’ with James McManus, CIO at Nutmeg, click here.