Portfolio Construction: Calculated risks
Factor investing promises to outperform both passive and active management. Carlo Svaluto Moreolo discusses the issue of implementation
At a glance
• Factor investing offers a way to focus on risks that are rewarded systematically.
• The strategy promises to provide investors with systematic outperformance and to be cheaper than pure active management.
• The main benefit, however, is that is gives pension funds a clearer view of the risk profile of their portfolio.
• The success of the strategy may depend not just on how the allocation is made but also on the time horizon.
The idea behind factor investing is simple yet persuasive. To the extent that rewarded and unrewarded risks in financial markets can be successfully distinguished, investors should minimise portfolio exposure to unrewarded ones. This, in turn, should enhance long-term returns.
Felix Goltz, head of applied research at EDHEC-Risk Institute emphasises the importance of the distinction between the two types of risk – rewarded and unrewarded. He explains: “There are some risks that are not easy to diversify away from. By taking exposure to these risks, a portfolio may experience losses at certain times, but the investor is rewarded for that exposure in the long term.”
From a pension fund’s perspective, the concept is undoubtedly powerful. Pension funds are forced to take several unrewarded risks, such as longevity and regulatory risk. The possibility of maximising exposure to rewarded risk from the asset side of the balance sheet is a compelling proposition.
But what are examples of risks that are not rewarded systematically in financial markets? The best example is stock-specific risk, according to Goltz, as it can be cancelled out by not investing in that stock. Because of that possibility to diversify away that risk, investors should not assume they will be rewarded for holding it.
The problem, of course, is whether there are risks that are compensated in a systematic way, according to a well-defined and regular pattern. The premise of risk-factor investing is that certain risks are indeed rewarded systematically, and that it is possible to isolate them. Decades of academic research suggest that these systematic risk factors exist, and that there is a clear distinction between traditional and alternative ones.
At BlackRock ’s factor strategies group, traditional ones are known as macro risk factors, whereas alternative ones are known as style risk factors. What the two categories have in common, according to David Gibbon, head of EMEA investment strategy for the group, is that they are drivers of return that persists over time and are broadly observed across asset classes.
Examples of macro factors are economic growth, real interest rates, inflation, default risk, political risk and liquidity. To different extents, all these factors will drive the performance of different asset classes.
Gibbon says: “They are the lens through which we view a traditional asset allocation. Most asset classes have a positive exposure to them. Factor risk is why risky asset classes earn a return premium, and if you start from that perspective, you can build the portfolio in a way that allows you to access the breadth of factors in the market. You should also make sure you’re not over-reliant on one or two factors.”
The alternative risk-factor world is populated by well-known concepts such as value, momentum, low volatility, growth, quality and carry. The distinction between traditional and alternative factors started to emerge as academics noticed the existence of ‘anomalies’ in the equity markets more than four decades ago. Groups of stocks with one of the attributes mentioned above appeared to outperform cyclically.
Concepts such as value or momentum were therefore also identified as systematic drivers of returns, much like equity risk or credit risk. But if investing in equities or credit also earns returns over time, as evidence suggests, why bother with alternative and more obscure risk factors?
The point is that investing in a portfolio that focuses on alternative risk premia protects you from those macro factors. Alternative risk factors tend to be uncorrelated with macro events such as negative growth surprises or liquidity crises.
The entire proposition of factor investing is based on the existence of systematic returns from alternative risk factors. That value comes from their outperformance and decorrelation from macro factors.
But what can a systematic risk factor allocation specifically achieve for an institutional investor? The straight answer, which is validated by scientific evidence, is that a portfolio built around systematic risk factors can provide positive risk-adjusted returns net of costs over time.
Risk-factor portfolios should outperform pure passive ones, as they focus on uncorrelated risks. More often that not, risk-factor returns should also exceed active returns, as they avoid stock-specific risk. This is excluding active managers that constantly outperform, which, unfortunately, are few and far between.
Naturally, while both theory and evidence support this view, the applications of risk-factor investing in real life are problematic. The complexity arises because isolating rewarded risk factors, both from a theoretical and practical point of view, is fraught with difficulties.
This seems like a tautology, but the problem with risk-factor investing is just as described – risk factors exist, but it is impossible to know with absolute certainty what form they take.
When discussing risk factors, academics and practitioners refer to concepts such momentum, value, low volatility, size and quality, which are ideal characteristic of securities.
There are many ways, however, of bringing these ideal characteristics to life. This is done by identifying signals, indicating that a group of securities carries a particular risk factor. Which signals are indicators of risk factors is entirely subjective.
Ashley Lester, head of multi-asset and portfolio solutions research at Schroders, a firm with an historical bias towards active management, advocates a “less religious” approach in selecting factors: “In the last decade, several dozen new equity return signals have been published. It would be truly amazing if all of those were real signals, and it would be amazing if they all fitted neatly into four buckets which happened to correspond to the original factors described by Fama and French in 1992 [Eugene Fama and Kenneth French, The Cross-Section of Expected Stock Returns, Journal of Finance, June 1992].”
“To us, the art of constructing a good factor portfolio is about selecting families of signals and keeping an open mind to what has been a good signal but may be tailing off, and to what is likely to become a new investment theme. The key question is, what is likely to produce a predictable return profile that sits well in the portfolio?”
But the real advantage of building a risk-factor portfolio is that it allows investors to improve their understanding of their portfolio’s true risk profile and to take greater control of it.
Lester takes this view, saying: “The existence of factor-based investing lends greater transparency to investors like pension funds, as well as greater capacity to control the risks that they are taking in their portfolio. The promise of factor investing for a pension fund is that, if they can think more clearly about the particular feature that makes them different from other pension funds, they can work with asset managers to tailor their exposure to collections of factors. That allows them greater insight into what are the best ways of maximising the expected returns consistent with their risk profile.”
In this sense, Goltz says that risk-factor investing represents a step ahead of traditional active management. It empowers investment organisations with the ability to decide on what risk factors they should be exposed to. The decision should be consistent with their investment beliefs and their overall risk budget.
Goltz says: “Traditional active managers also promise outperformance compared to cap-weighted indices. But the asset owners don’t control what the active risk exposure of the active manager is. This could lead to a mismatch, where the active manager is tilting towards some factors that are not in line with the investment beliefs of the asset owner.
“Thinking about risk factors allows asset owners to make sure that their investments are truly aligned with their investment beliefs, because they can explicitly target the factors they’re going to tilt towards.” In short, risk-factor investing puts asset owners in control.
The other key feature of risk-factor approaches is their potential to reduce cost. Because they are systematic and therefore repeatable strategies, asset managers can charge lower fees for them. .
All the above does not mean risk-factor investing is a free lunch, as investing in a risk-factor strategy introduces several complex problems.
First, there is evidence that certain factors underperform for extended periods of time. Goltz points out that the awareness of factor tilts in an investor’s portfolios means the investor will be able to explain underperformance to stakeholders.
But that does not solve the problem. Some practitioners propose that the solution is investing in a multi-factor portfolio. Pathmajan Rasan, CEO and CIO at Midtown Investments, a firm that specialises in quantitative strategies, makes the case for multi-factor portfolios. He says: “There are risks in allocating to a limited set of alternative-risk-premia strategies. Investors allocating to only one or two strategies are essentially taking an active view on those specific risk premia. In the short term, those strategies may perform poorly. The real value of this approach is in combining a range of lowly or negatively correlated alternative-risk-premia strategies to create a diversified multi-factor portfolio.”
But the benefits of a multi-factor portfolio should increase over time. Jamie Forbes, director of institutional client solutions at Russell Investments, advises that investors maintain exposure to multi-factors over a full market cycle of three to five years.
Forbes says long-term exposure brings added value as it introduces the benefits of diversification. She says: “If investors have the discipline to maintain that exposure, they can achieve the diversification benefits linked to how the different factors behave relative to one another.”
For others, the time horizon does not necessarily dictate the success of the strategy. Gibbon says: “We do not have a specific investment horizon in mind for risk-factor investing. We make the case that elements of a factor investing framework are all hallmarks of good security selection and portfolio building. Adopting a risk-factor framework makes sense almost regardless of your investment horizon.”
Instead, he adds, the fact that certain factors underperform in given times reinforces the idea that diversification is critical. “Because factors are differentiated drivers of returns, their bad times are not coincident. History suggests that a portfolio will enjoy enhanced returns by spreading risks among the various factors. Diversification is critical, especially because the periods of drawdown can be extended in some cases.”
Investors should therefore exploit their competitive advantage of having a long-term horizon, as well as focusing on multi-factor portfolios. One problem with multi-factor solutions is that factors could be correlated to an extent.
BMO Global Asset Management addresses the problem by focusing on ‘true’ factors. Erik Rubingh, head of systematic strategies at BMO Global Asset Management, explains: “In a security that presents a factor signal, we assess what part of that signal can be attributed just to that factor and what can be attributed to other factors. We then seek exposure to the pure factor.”
This is a pragmatic solution to the correlation problem in multi-factor portfolios, which introduces discretion in assessing how securities behave in terms of their factor exposure.
For investors who do not shy away from complex strategies, there is no shortage. Rasan argues that the way forward in risk-factor investing is applying multi-factor strategies to different asset classes. He says: “A multi-asset, multi-factor alternative-risk-premia portfolio enjoys a great deal of diversification benefits. But the ideal approach would be one that employs all the flexibility that factor investing provides to create a multi-asset, multi-strategy portfolio.”
Further ahead in terms of complexity lie market-neutral portfolios, a technique employing long and short exposures used to strip out unwanted factor risks.
At the opposite end is smart beta, which focuses on transparency and simplicity, but Rasan says this is at the expense of efficiency: “Smart beta strategies that are distributed through indexation are designed to be simple, transparent and with low turnover. Constraints such as these mean that the strategies used are only a sub-set of the broader factor-investing universe of strategies.”
He adds: “By loosening the indexation constraints, those investors and asset managers who are adept at factor investing can have a wider range of strategies at their disposal. This will allow them to capture the established investment themes like momentum and value more efficiently and optimally than their indexation counterparts.”
But although strategies such as multi-factor and market-neutral portfolios seek to increase the efficiency of risk factors, they undoubtedly increase complexity. One thing is building exposure to individual factors; another is having a framework for different factors.
This links to a more general point on pension fund governance. Proponents of risk-factor investing will say that their strategies are simpler and more transparent than stock picking. A pension scheme trustee may also find that assessing the quality of a quantitative strategy is safer than taking a bet on a stock picker’s skill.
However, both are optimistic assumptions. First, is it really true that risk factor approaches are simpler and more transparent?
Second, how likely is it that pension fund trustees will have the capacity to assess these strategies properly?
Investors should keep in mind that risk-factor investing is not about being objective. It involves plenty of subjective decisions. It has to do with being systematic. The difference with traditional active management is that stock pickers might allow for unsystematic judgement of what constitutes an unrewarded risk. Quantitative managers instead strive to make that judgment as systematic as possible. The question, therefore, is: how systematic do you want to be?
Case study: PKA’s multi-asset, multi-factor portfolio
PKA, a Danish pension administrator that runs three pension funds for employees in the healthcare and social-care sectors, has developed, over the years, a successful multi-factor, multi-asset-based approach. The DKK235bn (€31bn) fund allocated about 15% of its total risk budget to the strategy as of February.
The strategy, which explicitly targets alternative risk premia, has provided annual returns of about 7% over the past few years. Jannik Teigen Hjelmsted, senior portfolio manager at the fund, says that at the core of the strategy is a belief in the existence in alternative risk premia. “We truly believe that there exist risk premia other than the traditional ones, such as equity, credit or rates. We believe they exist due to structural, behavioural or institutional reasons and that those reasons are likely to persist.”
PKA’s allocation to alternative risk premia is done in a long-short format to extract the risk premia in their purest form. Teigen says: “We implement the traditional risk premia primarily through passive long exposures, whereas the alternative beta is implemented through long-short strategies. This is a way of decomposing the different equity returns into the traditional risk premia and the alternative risk premia, which we think requires a smarter implementation.”
Søren Grooss, portfolio manager at the fund, adds: “By allocating to alternative return sources we can build a more robust total portfolio, as risk events for alternative risk premia are different than the risk events for traditional assets. To us, alternative risk premia is just a natural evolution in understanding returns.”
The fund allocates to value, carry and momentum in equities, rates, currencies and commodities. But its alternative exposure goes beyond the well-known risk premia mentioned above. It also invests in merger arbitrage, insurance-linked securities as well as liquidity event and volatility strategies. It also implements unique derivatives strategies, one of which is based on an equity-index repo. Several strategies are run internally, where the effort required for maintaining them is compatible with the fund’s resources. ‘High-maintenance’ strategies are run in partnership with external providers.
How does the fund integrate such complex alternative strategies within a multi-asset portfolio? “Every investment decision made at PKA is allocated out of a risk budget, rather than an asset weight,” says Grooss.
At first sight, the complexity of the overall strategy seems more apt for a hedge fund rather than a pension fund, particularly because of how labour-intensive it is. But that is not how PKA sees it. Teigen says “most of the strategies we have are actually fairly simple. The implementation, though, can be complex and labour-heavy. Our responsibility in PKA is to find the strategies that best fit the portfolio and not necessarily to run each strategy ourselves.”
Grooss says that it is precisely because PKA, as a pension fund, can act with a long horizon, has a large risk capacity and is cash-rich that allows it to follow such a strategy. “We still see ourselves as a long-term investor. Many of the risk premia we target are actually very slow premia that can only be harvested over time. Being a pension fund is our competitive advantage with regard to risk premia.”