Underperformance is acceptable up to a point
AP7 Sweden: Anette Dahlberg, Portfolio manager
• Location: Stockholm
• Assets: SEK198bn (€21bn)
• Membership: 3m
• Swedish government buffer fund for the premium pension system
• Membership: 53,000
“Being a long-term investor, AP7 has the opportunity to take advantage of market deviations and add extra value for our savers. Dealing with underperforming strategies from time to time is a normal consequence of being a long-term investor.
As part of our overall strategy and risk monitoring, we evaluate, on an ongoing basis, whether the arguments in favour of an investment strategy still stand. As long as I believe in a given investment strategy, I am comfortable with underperformance. Underperformance can be an opportunity to buy more into the strategy.
Harvesting certain risk premia is hard unless the investment horizon is long. We have the patience to wait until the risk premium pays off.
In general, we apply a stop-loss that should trigger a decision on whether we want to double the position, scale down or close the position entirely. This mechanism means portfolio managers really have to decide whether the case is still valid.
As an example, I manage a SEK10.5bn (€1.11bn) long/short Swedish equity portfolio and the last couple of months have been challenging for some positions. In one case I was short Ericsson as I found the stock too expensive, but it continued to rise due to the weak Swedish krona. In my risk monitoring and revaluation of the strategy I decided to close that position since new arguments came into play and the Swedish krona went even weaker.
The external managers we employ are long/short investors in equities and currency. They have an absolute return target with an allocated risk budget. We evaluate all managers, taking into account the importance of a long-term approach in asset management. Thus, a manager can underperform as long as it is a reasonable part of the manager’s investment process and within the allocated risk budget.
In monitoring managers, we follow a strict process, with regular meetings. But there will be further discussion depending on the performance of the mandate and use of the risk budget. If the risk budget is fully consumed by the manager, the mandate is closed down.
Underperformance is acceptable as long as the strategy works and we can see the reasons for it. When the manager is no longer able to explain the underperformance or the strategy, it becomes unacceptable.”
ERAFP France: Philippe Desfossés, CEO
• Location: Paris
• Assets: €20bn
• Membership: 4.5m
• Mandatory pension fund for French civil servants
“In itself, underperformance or outperformance of a strategy does not mean much; it depends on the reasons behind it.
Firstly, assessing the performance of a strategy depends on the kind of benchmark you are using. If you are using a market-cap benchmark, maybe the outperformance of your asset manager is just the hidden risk you’re taking –without being conscious of it – by surfing on top of a gigantic wave. As we know, at some point bubbles do explode, and that is when your outzperformance may become an underperformance.
The assessment also depends on the timespan you use. For a long-term investor such as a pension fund, with liabilitiy duration that can reach 20 to 30 years, does it make sense to assess performance on a yearly basis?
We give our asset managers time and do not make rushed decisions to terminate a mandate just because there has been one year of underperformance.
At some point, if an asset manager is underperforming constantly it may become necessary to send tha manager a message, which can be a withdrawal of funds. If the performance remains poor, it might be time to discuss putting an end to that mandate.
Our trustees accept that an investment strategy may underperform in the short term. For this reason, they have been instrumental in the adoption of our SRI charter. All the assets we invest in are selected only after they have passed through our SRI grid, which comprises 45 criteria for stocks.
The trustees know that this approach may lead to greater tracking error, especially if compared with approaches that follow more traditional rules, such as not being too removed from the benchmark.
So far, we have yet to come across a mandate underperforming to the point where it would have become necessary to react.”
INARCASSA Italy: Alfredo Granata, CIO
• Location: Rome
• Invested assets: €8bn
• Membership: 170,000
• First-pillar fund for self-employed engineers and architects
“For both our passive and active mandates we set a reference benchmark and a maximum level of tracking error.
Our custodian bank supplies the technology to monitor these aspects closely and we look at daily performance and monthly tracking-error levels. If we observe either absolute or relative underperformance of a strategy, or we find that a manager is exceeding the tracking-error levels, depending on the extent of the underperformance, we will complain to the manager.
We do not react to every instance of underperformance by managers. Our minimum time horizon for each strategy is 12 months, so we can tolerate some underperformance and non-compliance with tracking error during a similar period. But if we observe an extended period of underperformance or tracking error exceeding expectations we will request a report from the manager on the reasons for that situation. We will also expect the manager to take steps to reduce the tracking error immediately.
If compliance with the tracking error requirement is not restored there are grounds to discuss replacing the manager with the board of directors.
Similarly, if we observe prolonged outperformance and also continued non-compliance with tracking-error requirements, we will ask the manager to review this and to explain why.
If the underperformance occurs but the limits to tracking error are respected, we simply ask for an explanation, and accept that it may be due to a temporary weakness of the strategy. In this case, underperformance may not justify changing the strategy, as it is still be an integral part of the overall portfolio strategy, which may not be affected by the temporary underperformance.
For example, we would accept that an active bond manager has decided to reduce the duration of the portfolio without informing us, so long as the portfolio is not positioned too far from our initial indications.
A real-life situation that occurred concerns an active equity manager which, at one stage, regularly outperformed and underperformed its benchmark index – the S&P500. We realised that situation was due to the manager integrating small-cap stocks into the portfolio. The manager was cautioned against doing that again, as it conflicts with our strategy of giving mandates for special asset classes, such as small-caps, to specialised managers.”
Interviews conducted by Carlo Svaluto Morelo