Successful stock-picking is rare
Environment Agency Pension Fund
Mark Mansley, CIO
• Location: Bristol
• Assets: £2.9bn (€4bn)
• Pension fund for the UK’s Environment Agency
We are very careful how we select active managers, but we do think they can add value in the right markets and with the right mandates. Essentially we look at two key types of active manager, neither of which are the traditional ‘core’ active managers. One type are the smart beta or systematic managers with a disciplined focus on styles and factors. The other type are highly concentrated, high-conviction managers with focused exposure to particular companies. For us it’s not just about the returns, although they are important; we are trying to reduce risk as well as integrating our ESG approach, which is fundamental to us.
Most of our active managers have broad investment universes, but we feel active management works particularly well in more specialist markets, such as small caps and emerging market. In short, we always try to work out what is the best context to apply active management.
We do not feel that in recent years it has become particularly difficult to find good active managers, and we think we can still find opportunities.
I believe good active managers will have a clear and credible philosophy and a sound process for implementing this. Also, good active managers will have size constraints on the total assets in a particular strategy, so they may not be the largest managers. They will often control the size of their fund, so they remain able to access the best opportunities.
In selecting managers, we may assess a particular strategy first, and then consider whether it can be applied actively or passively. But we often find that if we want to cover all the aspects that we consider important, including ESG integration and risk management, it may be easier to go for an active option. Quite often, there isn’t much difference between tailored passive managers and active systematic managers in terms of costs.
For example, we have a value exposure through a RAFI Fundamental index and a larger low-volatility exposure, which we run through active mandates. We believe low-volatility stocks can become expensive, and an active manager can apply a value filter and thus contain exposure to over-priced stocks.
Above all, we are long-term investors with real liabilities, and that is why we favour quality, low-volatility, value if possible and to a lesser extent small caps. Momentum can be also useful at times, which is why both our low-volatility managers look at momentum as a secondary factor.
Wolfram Gerdes, CIO
• Location: Dortmund
• Assets: €11bn
• Pension fund for employees of German protestant churches
As an institution, we do believe in active management, even though we would agree that generating substantial outperformance with reasonable predictability is very difficult. Many of our mandates are active, and the aim is to get a slight outperformance over the benchmark after costs. Many proponents of passive argue that the majority of active managers fail to achieve their goals. But in our long-term experience we have been able to find some alpha left across our 20-plus mandates, adjusting for costs. This year has been particularly good, as we have harvested 0.5% alpha after costs across the entire spectrum of mandates.
There are ifs and buts, however. First, it takes much effort to select and control managers. When there is underperformance, we need to decide whether it’s a case of bad luck or there is something more fundamental going wrong.
Another key criterion that has to be applied to safeguard alpha is making sure you are not overpaying for it. The commonly held knowledge is that good managers can afford to ask for high fees, but our experience is different. If you hire a manager that charges high fees, the alpha gets diluted very quickly, so controlling costs is extremely important. Also, we find that the best managers are those that keep their clients happy, and hence grow their business.
With these managers it should always be possible to negotiate fees. Other common ideas about active managers are that the best ones have low AUM (ie, their strategies cannot be scaled) and that alpha is only available in illiquid markets. I think the two arguments are not necessarily backed by our data. We have ‘plain vanilla’ active equity mandates in US equities or EU equities that generate very good alpha, and the AUM of these managers is not necessarily small. It is likely that some of those managers have created a very aggressive momentum strategy that works very well, at least for the time being.
What makes a good active manager is a very difficult question, but we find two features that are common to all, although they are not necessarily predictive. First, they need to have strong ideas and strong beliefs in what they are doing. Second, good teams tend to have been together for a long time. Significant staff turnover as well as ownership changes inevitably affect the products. That is why as investors we need to re-evaluate mandates often. It certainly takes significant time and effort.
Stefan Beiner, head of asset management and deputy CEO
• Location: Bern
• Assets: CHF37bn (€30bn)
• Collective provider of pension plans for Swiss federal employees
We believe only a handful managers can select stocks successfully and achieve a sustainable risk-adjusted performance above the benchmark.
In our investment process, we start from deciding which building blocks – for instance Swiss equities, US government bonds or real estate – should form our strategic asset allocation, and what should be the relative weight of each building block. We believe this is the most important decision, and it is based on our dynamic ALM process. The next most important step is identifying the most efficient benchmark for each building block. If we believe we can define an efficient benchmark, then we will design a rule-based replication process.
Quite often we use customised benchmarks. We tilt them towards the risk premia we want to extract. In some cases, we cannot define an efficient benchmark, or the benchmark can’t be implemented. One example is Swiss real estate. For that asset class, we have an in-house team that has, almost by definition, an active approach. Another example is local currency EM debt. We think the main current index is not the most efficient way to do it, but we don’t have a better way to customise it, and therefore we are looking for active managers. Yet, we will give quite clear guidelines, so the mandates we will award will not be highly active.
Assuming we can define an efficient benchmark, we struggle to believe we can regularly find active managers that can beat that benchmark through stock selection. As well as the strategic asset allocation, we establish tactical bandwidths. We follow a risk-focused tactical asset allocation process that has allowed us to outperform the efficient benchmark in the last 10 years, by timing the allocation to risk premia.
We have a number of criteria when choosing asset classes and mandates, and one would be net return. Costs matter, but they are not the only focus. For instance, we have just added private debt to the portfolio, which is expensive, but makes sense net of costs. In other words, we try to get the most efficient risk-return profile as possible. When it comes to pure alpha – by pure alpha I mean outperformance that is not due to an exposure to systematic risk factors – even though it would be highly valuable, we do not believe we are able to find it, particularly net of costs and for the long term.
Interviews conducted by Carlo Svaluto Moreolo