Portfolio Institutional: Why are none of your schemes using factor investing, Andy?
Andy Scott: There are lots of things affecting private sector defined benefit (DB) schemes at the moment, such as hedging and cash-flow considerations.
I am well informed for a trustee but, like liability-driven investing (LDI), getting trustees to understand that something is a good idea and that they should take it up is the challenge.
I do have a defined contribution (DC) scheme that is interested because there’s more of an equity allocation involved, but I am still to appoint a factor investing manager.
Andy Peach: Trustees have had a lot on their plate. LDI and diversified growth portfolios have been big things and now trustees are getting around to thinking about their residual equity allocation.
Compared to LDI, which can be a complex topic, factor investing’s quite a simple story when you get to the nub of it and there are different benefits to different investors. If a client is investing passively, they could benefit from factor investing’s diversification and better risk-adjusted returns over the long term. If they are coming from an active starting point, quite often with a shrinking active equity portfolio, the question is, can trustees justify the governance spend on what used to be a big part of a portfolio but is probably only 5% or 10% now. Factor investing is a potential alternative to consider.
“It’s no surprise that multi-factor equity strategies have lagged their market cap benchmark in the past couple of years.”
Yvette Murphy, State Street Global Advisors
PI: How are institutions using factors?
Yvette Murphy: They are using them to understand their current investments, more so than they did in the past. That has prompted a general awareness of the factors inherent in active strategies, what the total exposure is across multiple equity allocations and a better understanding of risk metrics.
We have been seeing, certainly in the conversations that I have had with clients, a lot more decomposing through the lens of factors.
We are also starting to see big pension plans in the US incorporating factor benchmarks into their global policy. Whether there is a recognition that these are part of the opportunity set now or whether there is a dollar allocation, they are incorporating these themes into their governance framework. This helps contextualise another managers performance.
David Barron: The industry has done a good job of tripping itself up in this space. Academic research said that factors can do something for your portfolio and so there was a mass rush to create product. We are finally making this objective-based by asking the client what they want to achieve.
We approach factor meetings by saying: “Market cap’s not broken, it does exactly what you want it to do and it’s low cost, it’s transparent and it delivers benchmark return. It is fit for purpose if that’s your objective, but factors are another way of thinking about diversifying your portfolio.”
The challenge for asset managers is that it is potentially the end of pooled funds, which is a trend we are starting to see. Clients have choice, they don’t want too much value, or they want low volatility, so no investor’s experience is the same and so we are tailoring our offering to the needs of a client.
Erik van Leeuwen: Only 10% to 15% of factor investing is fixed income. There are 50 years of academic research in equities, but in fixed income the research only extends 15 to 20 years.
We started doing our fixed income factor research in the late ‘90s but that’s still pretty recent compared to equities. We have seen a similar pattern in products coming to the market. There is a lot of interest. We have equity experts joining our fixed income meetings because they know which factors work on the equity side and are transferring that knowledge into fixed income.
Scott: Is this in the UK?
Leeuwen: It’s worldwide. Pension funds, sovereign wealth funds and some wholesale funds have started to adopt it. Scandinavia, the Netherlands, the UK and Germany are quant oriented as are Australia, the Middle East and Singapore. In the US they are aware of the factor equity story, but in fixed income it is still relatively early days.
Data and academic research are sparse, that’s why it takes time for people to understand that there is also an economic rationale for factor investing in fixed income. You can show empirical data, but it easily takes two or three years for people to get comfortable with it, which is understandable and reasonable.
PI: Is factor-based investing in fixed income just about generating alpha?
Leeuwen: The main goal is diversification. We wouldn’t say that factor investing is better than fundamental investing. But if you look at how fixed income assets are managed on the credit side, it’s a 99% fundamentally driven market. When we analyse global fundamental managers, there is typically quite a strong correlation between them. In fact, regardless of domicile, a 20% or 30% correlation is not unusual.
Factor investing is a different way of creating alpha. It is active management; you are trying to harvest alpha but from a different perspective. If you look at the global peer group of asset managers, this is the only strategy that has a negative correlation in terms of outperformance versus fundamental managers.
We do not claim that it’s better. It is a different style of harvesting alpha and therefore a nice diversification for your portfolio.
PI: How have factor strategies performed in recent years? Have they done what it says on the tin?
Leeuwen: We started to blend fixed income into traditional portfolios in 2005 and launched our first dedicated product in 2012. If we look at the past three to five years’ real-life track records, the strategy has performed as expected based on the academic research.
The challenge is in how to implement factors in fixed income. You have sparse liquidity and high transaction costs. For me, the challenge is not coming up with research with information ratios of one or oneand-a-half, I can easily come up with nice stories. For us, it’s important to do realistic back-testing and build real-life track records, to bring the stories to life.
Scott: Is your mandate purely to get alpha and you are giving it unconstrained, or do you have parameters around what you are doing and how you achieve it?
Leeuwen: Our preferred way to implement factors in fixed income is to look at the global credit space, where the aim is to outperform a global index. There is no interest rate or currency exposure, the objective is purely to take credit exposure in a different way than a fundamental manager would.
What is important for a strategy like this is that you need to have a large starting universe. It is difficult to apply factors to sterling credit because such a strategy has a lot of idiosyncratic risk. You need to have a bigger spectrum; global credit is perfect for that. From this broad base you can tailor to the needs of individual clients.
“Trustees need to understand that this is a long-term game and should not be looked at quarterly.”
Andy Scott, Dalriada Trustees
Barron: One of the interesting things about equity indices is that performance depends on which benchmark you choose.
We have seen mass product proliferation with more than 3 million indices, which is 70-fold the number of equity securities – and they keep coming. The scary thing about indices is a shifting of the burden onto the asset owner and their consultants to make decisions. When it comes to implementing that chosen portfolio, I deliver the return that the index creates and so it shifts that responsibility. Understanding what goes into construction is critical.
Traditional factors perform the way you would expect. Some happen to be procyclical, some are not. When markets pull back, implementing those factors in a long-only equity portfolio can deliver drastically different results. A simple decision like sector neutralising a value score is the difference between having Apple feature majorly in a US value index or not having it at all. Obviously, the one with Apple in over the past five years did well, the other didn’t.
You see double digit differences in returns from indexes that say the same thing. That is the scary thing with this space. You need to tie the objectives to index methodology to ensure that the outcomes are going to match what your expectations are.
Murphy: It’s no surprise that multi-factor equity strategies have lagged their market cap benchmark in the past couple of years. These strategies either outperform or underperform depending on the market environment, such as a high beta-driven growth rally. If single factors underperform simultaneously, then the multi-factor diversification may not hold up as you would expect and the portfolio might lag the market cap benchmark. We saw momentum and quality behave in a similar way throughout 2018. Stocks that were ranked strongly on quality factors also happened to demonstrate strong momentum, such as some of the information technology stocks. When it comes to a smart beta allocation decision, it will likely be a bit more expensive than a traditional index and will be anchored to the market cap index from a performance perspective. Whether it is performing as we would expect, or out/underperforming the global benchmark, we need to understand why.
Peach: Which is what the client will focus on. They will have moved potentially from an active portfolio with the MSCI World as a benchmark, so they will be thinking about what they have moved from. A lot of multi-factor products in the past 18 months have struggled. But if you look at the market leaders over that period, they are companies with significant negative cash-flow, which has not been the case since before the dotcom crash 20 years ago.
If you are looking to allocate towards factors that use fundamental data to find companies that are profitable or cash-flow generative, by definition that strategy will not do as well when companies generating negative cash-flows are the ones outperforming. They have underperformed market cap, but they have performed as expected in that environment.
If factors always outperformed they would be a free lunch. So this period of underperformance is actually helping to tell the story about factors, that they can underperform over a period is the principal risk related to a market cap benchmark.
Whilst some of these multi-factor products have underperformed market cap in the short term, most have outperformed over three to five years. Our research showed that there’s a better than a 50/50 chance that a multi-factor product could outperform over shorter periods. Over three or five years you are looking at a much more significant chance of outperforming the market cap.
Scott: When there is an underperformance for good reason it is difficult to tell a trustee that this is a good idea. That must be one of the difficulties in selling it. It is the same with fiduciary management. A lot of the decisions taken have been defensive at a time when markets are going up. When they went down in the fourth quarter some fiduciary managers didn’t pick up on that.
Trustees need to understand that this is a long-term game and should not be looked at quarterly.
Leeuwen: We have a model and a fund that follows duration decisions as well. We have applied that strategy for 20 years taking long and short duration positions in a global government benchmark.
We haven’t once overridden the model. Even when everybody was doubting quant models, when they saw negative rates and quantitative easing. It’s not the Holy Grail but it adds value in the long run to a portfolio when you complement it with other strategies.
Scott: It is about sticking with it in the long run.
Leeuwen: Generally, in back-testing and academic research you will see lines trending up. The biggest risk is that people believe that factor credit is always a winning strategy, but that’s not the case. There is no free lunch. Everybody needs to realise that different investment styles can lead to underperformance, but you must stick with your belief.
“Factor investing is a different way of creating alpha.”
Erik van Leeuwen, Robeco
Stuart Trow: As a strategy you have two problems. One is a portfolio manager curve-fitting, where one name can completely skew the research from the strategy.
The quarterly reporting issue is quite big, but at what point, especially in credit, do you decide it’s not working? One problem we have is deciding when to take a loss on credit. If the strategy’s not working because you have curve-fitted it and you have replicated something that happened in the past, at what point do you pull the plug?
From my own trading experience, it is sobering because I had a couple of good years and thought I was good at picking stocks, but it wasn’t. It was because I was prepared to chuck the stuff that wasn’t working.
Barron: The regime shifts that pundits are starting to point to is interesting. All the academic research is from the 1970s, ‘80s and ‘90s but companies are a little different now. Intangibles don’t necessarily make their way into book value anymore.
We have some clients in single-factor value strategies who are anti them right now, because the drawdown is significant relative to market cap and they are paying more in transaction costs for that product. So, when do you pull?
Trow: QE has squashed factor investing for credit. If you are investment grade in euros you are at the party; if you are not, you are not.
Murphy: There is a timing element, too. People are recognising the benefit of multi-factor diversification and say, not allocating all of their exposure to value. But at what point do they shift their exposure? Investors may want to move towards a more multi-factor approach, but in taking some of that value exposure off the table, they may miss that inflection point where value rebounds, which is historically where we have seen a lot of the outperformance.
Leeuwen: We would advise clients to diversify as much as possible, using different styles. When we are implementing factor investing in credit our approach is to use equally weighted portfolios, small bets and a large number of names to avoid idiosyncratic risk. It is not the Holy Grail but there’s added value in it and the way to harvest it is to diversify all the styles. It is key to stabilising your results.
Scott: There are so many other things that are coming into it, such as cash-flow investment, LDI and defensive structured equity. This is another one. They are getting swamped with ideas. At the end of the day, performance is what we as trustees look at and when that is not happening…
PI: Should investors be concerned about overcrowding?
Barron: In the long-only equity space there is a lot of capacity left in factors. In pure factors, we have a long way to go. History tells us that there are some concerns. Small cap disappeared in the 1980s and then came back.
The point is that factors have a lot of room, but some of the strategies might not. Some of the more concentrated smaller capitalisation strategies have an extremely high turnover that can erode the potential benefit of the factors you are incorporating in your portfolio.
An interesting dynamic against index factors being in that space is it’s not just my capacity, it’s also State Street’s and BlackRock’s. Anybody can have an index strategy but once it is on the shelf you must keep an eye on aggregate interest in that strategy. That makes it tricky. From a pure market cap perspective, there’s wealth erosion that happens on index reviews. You can see that with price dislocation in the last 15 minutes of trading because of the wealth of assets that are benchmarked there.
Factors have a little bit more time before that becomes a problem, but investors should keep an eye out.
Trow: Another way of looking at the same thing is with green bonds. You might want to invest in that field, but there simply isn’t enough product out there to satisfy investor demand.
Murphy: There has been a proliferation of factor indices as well, and they are all constructed differently. It’s not like there’s separate capacity for ‘factors’, distinct from capacity in ‘active quant’. Both strategies are investing in similar themes and those core factor exposures have been a bedrock of traditional active quant and fundamental strategies for a long time now, but we are a long way off a natural limit.
PI: What is the best way to create a multi-factor portfolio?
Murphy: There are a number of ways. It depends on the objective that you are trying to achieve and your willingness to take risk relative to the benchmark. We would look at trying to be risk efficient, such as maximising exposure per unit of active risk, and then implement those exposures cost-efficiently.
Part of the big industry debate is around how to build a multi-factor portfolio and combine factors in a single solution. There is just as much research and support for a bottom-up approach as there is for a top-down approach.
From an investor’s perspective, I would look at what am I getting for the cost I’m paying, and the additional risk that I am taking, for deviating from the market cap index.
We are fans of a bottom-up approach and acknowledge that the correlation between factors can impact the amount of exposure that we harvest, but we also acknowledge that that correlation can change through time. Being able to dynamically adjust the portfolio construction and the exposures that you are taking on with that change in correlation is something that we consider.
“The big question is are the top 10 stocks today going to be the top 10 stocks in 10 years’ time. If you are expecting a changing of the guard, then a multi-factor product may deliver something superior.”
David Barron, Legal & General Investment Management
Scott: Risk is always an interesting word. There are so many risks that things don’t work out the way we thought.
As a trustee, we want to beat the market cap and will pay more to do that. The more that you move away from a market cap strategy to fit your factor strategy, the more you risk that you are doing what you said you were going to do but trustees look at results and say: “Here’s the market cap result and here’s your result and we are paying you more money.” Is that a risk that you analyse?
Murphy: We look at any performance deviation, positive or negative, to make sure that it is consistent with our expectations. We keep a pulse check on several strategies that are similar by name and objective, but not necessarily by outcome. Particularly with underperformance, we want to understand these drivers, and whether it’s driven by the factor exposures targeted as opposed to unintended outcomes influencing performance.
Peach: When building a factor or a multi-factor portfolio you want to make sure that you are getting the exposures you expect and not unhelpful ones that are working against each other. There are a number of ways of achieving that with appropriate index construction.
Murphy: We have been doing more research on the implementation properties of factor strategies, moving from academic factor definitions or portfolio construction techniques to looking at the relationship between exposure, risk and cost. If you have a portfolio that includes emerging markets, how do you calibrate allowable positions in less liquid countries such as Malaysia? Even if that Malaysian stock has great factor characteristics, is will likely cost a significant amount to trade. These trading costs may erode any premia that was harvested.
Things like turnover and other capacity issues are starting to be built into portfolio design research.
Leeuwen: On implementation in fixed income, there are a lot of similarities when it comes to risk. In credit you don’t have the upside; you only get downside, so diversification is key. Risk here is two dimensional. It’s not only versus the index but also absolute risk. We would rather have equally rated portfolios independent of how big the issuers are in the index as, from a trustee perspective, it is a better composition of a portfolio.
Where you see the biggest difference versus equities, is liquidity and the availability of bonds. Life is easy for equity investors because once you have the modelling and a wish list, you know you can trade all equities. You can undertake different types of program trades and see how to achieve a lower transaction cost, but at least you know you can buy everything.
However, liquidity in credit is sparse. This must be built in from the research in the beginning. If you assume that you want to buy the top 20 names and you assume that you can buy all of them, then you are overstating the available factor exposure. In practice you are happy if you find half of the bonds and at a reasonable transaction cost.
Implementation is key. Because liquidity is sparse you should be aware that once you start implementing you need keep turnover low. You need to be mindful of where you can find broker access and implement these names.
We are often asked if we can create factor indices in fixed income. Our stance so far is that you can’t. If you want to create these indices in fixed income you risk creating indices which have a lot of turnover, high transaction costs and names that look good, but you can’t be tracked, or you must build complex indices, which are less transparent.
For us, if you want to harvest factor investing in fixed income or credit you need to apply an active strategy where you buy attractive names and hold them for a long time. The translation from research into how to realise good results in practice is still something where there’s a lot of skill needed.
Trow: That has fundamentally changed the way we look at a credit portfolio. We used to aim to turn it over regularly, but the secondary liquidity’s just not there to manage a crisis. If you put something into any strategy you have to love it because you are going to have a problem getting out of it.
Scott: How easy is it to explain that to trustees? If you have not been able to achieve something because you do not have the liquidity or a particular stock is not available, that is fundamental to the performance you achieve. You are in the same situation as everyone else, but you have under-achieved because of that premise.
Leeuwen: It’s researching factors and how you implement them. Once you have realistic simulations and compare them to what you do in real life you can see that there are a lot of similarities. That is how you build trust.
There’s a good economic rationale and you can show that it works, but it takes time for people to understand it.
Scott: The other thing is quarterly reporting. In meetings you have maybe 20 minutes to explain it and you are one of four managers. There are all these things that stop the message getting fully through to the trustee board so that they can understand it. I understand where you are coming from, but the problem is often explaining it to the trustee board in a way that they can understand it. With all these different types of investing, such as LDI, the explanation and understanding are the most challenging.
PI: How does ESG fit into factor investing?
Leeuwen: RobecoSAM, our daughter company, provides an ESG score on every company in our universe. We want to make sure that the portfolios we create have at least as good an ESG score as the average universe.
Our strategies are 90% rules-based, 10% fundamental. We believe in our models, but we still ask an analyst to review the names that pop up which includes a look at ESG risks.
We use quantitative analysis as well as fundamental screening to make sure that we have a good ESG profile in portfolios.
PI: Is ESG something that clients are asking for generally, not just in factor investing?
Scott: It’s becoming huge for trustees. The regulator is demanding that you put it in your statements, so everyone’s talking about it and having to comply. It used to be a small statement in reports saying that we don’t have any views on this, and we leave it to the managers. That is no longer good enough and is where factor investing can shape a portfolio to be more ESG compliant. So it could be a big driver for factor investing.
Peach: Factor investing in its purest form is not trying to do anything in an ESG-friendly way per se. The first problem is defining what ESG is. It can be strongly driven by clients’ beliefs and everyone’s beliefs are different, so trying to make that into a single all-encompassing product is difficult. So it would not necessarily pass the test of being a factor because it is not robust enough as a definition.
One of the more objective bits of available data is around carbon emissions. There is something that can be done around carbon that is less subjective and can be incorporated into a factor product. It is not a factor in itself but it can be incorporated without diluting the strength of the factors or negatively impacting expected returns.
When it comes to ESG, you can do something on the ‘E’ with a carbon screen. For the ‘G’ you could argue that you are getting good governance within the quality companies that you are buying. However, the ‘S’ is a little harder to define.
Barron: That’s the direction of travel. Nine out of 10 meetings are now ESG focused and the tenth ends up being ESG focused. It is the trend that we are seeing.
We have a whole host of multi-factor products where clients are asking what is the ESG score of this? What’s the carbon footprint of this? The big question is could you do it without derogating your factor intensity.
ESG-plus factors in the index space work well because factors are going to drive performance, that’s why you allocated them, or at least for the diversification benefit. Anything else, highly concentrated ESG leaders, the top 50, that’s a big leap of faith and you must have a lot of trust in the vendor of that ESG data to want to concentrate a portfolio just on ESG.
Factors help you determine if the stock is valuable and you can screen it with ESG but index-only or cap weighted ESG products, for example, can be tricky. You are tracking something that you don’t have a long history showing that it’s going to deliver some return benefit.
“If factors always outperformed they would be a free lunch.”
Andy Peach, Aon
Scott: Trustees must take themselves away from return if they want genuine ESG. Charities and churches must accept that by moving away from market cap they are asking me to do things that are not in line with the general market. The result may be that you underperform, but you keep within your ESG beliefs.
Murphy: The extent of ESG integration can impact what you are trying to achieve. At a basic level, negative exclusionary screens or even positive screens can tick the box of ESG integration.
We keep talking about ESG as a factor, but is it in the traditional risk premia sense? It’s too early to say. One of the pillars of factor investing is intuition. It is intuitive to invest in well governed businesses that contribute to long-term shareholder value. Whilst we do not have a long history of empirical support for ESG as a factor, like we do with financial metrics for style factors, an ESG investing philosophy certainly aligns with that intuition pillar. So how do we build it into portfolios?
If you subscribe to the belief that you want exposure to ESG, like targeted exposure to value or momentum, then you can incorporate it into portfolio construction in a similar way. This could include creating a multi-factor + ESG score or requiring a minimum level of active ESG exposure in a portfolio optimisation, for example. We view this as a full ESG integration as opposed to negative screens, such as removing controversial weapons, for example. As more research comes to bear showing the interaction effect between factors and ESG or climate, I suspect we will see more and more of those dual objective strategies come to market.
“QE has squashed factor investing for credit. If you are investment grade in euros you are at the party; if you are not, you are not.”
Stuart Trow, European Bank for Reconstruction & Development
Trow: In credit, even if you are not actively targeting ESG, the people originating the loans determine what that universe looks like. Banks are much less prepared to arrange loans for coalminers. So even if you are not pursuing an ESG factor your universe has changed, perhaps more so in credit than in equity. Equity still exists, whereas with credit a loan physically might not exist because nobody’s lending to a coalminer.
So things are improving towards ESG without necessarily opting to do it consciously.
Scott: I read somewhere that a company issued a green bond and the price went through the roof. It shows that the demand is there. The thing is, if everyone’s buying green bonds is there a danger that they are overpriced? This, again, is a quandary for trustees. If you want a return and everyone is going for the same stocks it is not going to achieve a return, even though it’s something that is environmentally desirable. Is there a danger that there’s going to be a herding in demand for the best stocks?
Leeuwen: The downside risk, especially for credit investors, is huge. We started more than 10 years ago with ESG as a filter to look at downside risk; now it has become far more important for trustees to show that they are doing something on sustainability. That can be for good reasons, but it could also be just to show that they have done something. Green bonds are sometime seen as an easy one – putting 10% or 20% of green bonds in your portfolio is an easy tick in the box.
A green bond has the same issuer risk. Relative value-wise, any basis points that a green bond is issued at lower than a normal bond means it’s too expensive. There is no premium, yet it’s the same issuer risk. From an economic standpoint it should be priced at least equal to the normal bond.
We would rather have issuers pursue a sustainable policy instead of issuing green bonds. When we invest in green bonds we pay close attention to the possibility of green washing; we look for issuers that do what they promise and pay a good relative value. But the more you see clients wanting to tick this box by just adding green bonds to their portfolio, there is the risk of overpricing.
PI: Will returns from factor strategies improve going forward?
Barron: The tech rally continues to see multi-factor underperform, but as we move through the cycle the hope is that we will see a better risk-adjusted performance from multi-factor products.
The big question is are the top 10 stocks today going to be the top 10 stocks in 10 years’ time. If you are expecting a changing of the guard, then a multi-factor product may deliver something superior.
Peach: Multi-factor has struggled in the past two years, but over the long term I believe it will beat the market.
A lot of multi-factor products have handsome back-tests, so we always view them with a healthy cynicism. Factor and multi-factor products will not outperform to the degree that the back-test say they will, but in the long term we believe they do better than their recent real-world performance.
Leeuwen: It is still early days in fixed income, but if you have realistic expectations these strategies will perform well.
