No excuses should be needed by investors to think seriously about the promises of smart beta – returns in-line with or even better than the market portfolio but with substantially lower levels of volatility. But it is easy to see why these promises resonate with European pension funds in particular.
Many have been de-risking for a decade or more. Some do so because their membership is maturing; some have closed to new members and started crystallising those liabilities; most face pressure from regulators, sponsors and shareholders. De-risking necessarily shrinks the capital allocation to that part of the portfolio whose aim is to outperform liabilities. If there is a way to spend that growth-portfolio risk budget more efficiently, to get the same juice from that shrinking capital allocation, it solves a very visible problem.
But the UK Environment Agency’s Active Pension Fund isn’t in that position. It has been reducing its equities allocation (it is on its way to 50%, down from 63% a year ago), but it hasn’t been de-risking in the traditional sense. It is slashing its government bonds allocation in favour of corporate bonds, for example, maintaining its private equity allocation and more than doubling its assets in illiquid real assets like property and infrastructure. As its name suggests, the fund is open to new members, and it enjoys a healthy 90% funding level – its modest target is to achieve 100% funding by 2031 and limit the likelihood of falling below the 80% level.
“When we did our strategy review in 2011, the scope for de-risking was limited, given our return requirements,” says CIO Mark Mansley. “The real objective was simply to go through the entire portfolio and make sure that every allocation was working as hard as possible, squeezing every drop of return from the risks we were taking – whether that was active managers, their benchmarks, or passive mandates. It was in the course of considering that question that smart beta, in various forms, started to suggest itself.”
It helped that Mansley himself always had an interest in quantitative portfolio management and was aware of smart beta ideas, and also that its investment consultant, Mercer, had done quite a lot of work in this area. But perhaps more importantly, the fund is unsurprisingly a leader in integrating environmental, social and governance risk management into its investment processes – which has reinforced the habit of thinking about the total portfolio in terms of risk exposures, hidden as well as obvious, rather than simple asset class allocations.
“We had become increasingly aware of the importance of the style biases in the active mandates that we had,” Mansley recalls. “During the crisis we noticed that managers who did particularly well for us were those with something of a quality bias; and on the other side of the coin, we were becoming aware of some less attractive biases we had in the portfolio, especially a bias away from value.”
Two objectives underlying the basic desire to maximise the portfolio’s Sharpe ratio, then, were to reduce risk against the fund’s liabilities, where possible, and to correct the portfolio’s unwanted style biases.
Environment Agency Active Pension Fund
• Defined benefit
• Members: 23,000; open to new members
• Assets: £2.13bn (March 2013)
• Bonds: 27%
• Equities: 68%
• Property, infrastructure, timberland: 5%
What has it done?
The fund has implemented a new benchmark, with the weights of single-stocks capped at 2%, for its active UK equities mandate with Standard Life Investments; it has awarded two new actively-managed low-volatility global equity mandates to Quoniam and Robeco; and a new passive, fundamentally-weighted global equity mandate to Legal & General Investment Management (LGIM), using a FTSE RAFI benchmark.
Other asset managers
• Legal & General Investment Management: Index-linked Gilts
• Royal London Asset Management, Pimco, LGIM: corporate bonds
• LGIM: passive market cap-weighted (MCW) UK equities
• LGIM: passive MCW global equities
• Sarasin & Partners: active global equities
• Generation Investment Management: active global equities
• Impax Asset Management: active global equities
• Comgest: active emerging market equities
• First State Investments: passive MCW emerging market equities
• Robeco: rivate equity
• Aviva: UK property
Portfolios exploiting the low-volatility anomaly seemed apt to meet all three of those objectives: they clearly exhibit lower volatility; they nonetheless appear to outperform the market; and they tend to have a tilt towards quality stocks. Fundamental indexation which, under the well-known Research Affiliates formulation, weights stocks by their sales, cash flows, book value and dividends, seems to deliver a better Sharpe ratio than the market and the tilt to value that the Environment Agency fund was lacking.
Mansley says that the fund considered most other smart-beta strategies, but found that they were either too similar to the active management it already had or not clearly relevant to the portfolio problems it had identified. The result was two low-volatility global equity mandates of about £90m (€109m) each for Quoniam and Robeco, and a RAFI index-tracking global equities mandate for Legal & General Investment Management worth about £100m.
But actually, the fund’s real first encounter with smart beta came from grappling with a specific benchmarking problem rather than a general portfolio problem. It had a £160m active UK equities mandate with Standard Life Investments benchmarked against the FTSE All-Share index. That index is highly concentrated, with the top 10 stocks accounting for 40% of the market CAP. Oil, mining, healthcare and banks swamp all other sectors.
“I do believe in having an allocation to our domestic market, but I’m also very aware that the FTSE is desperately flawed,” says Mansley.
Taken together, those flaws mean that taking active positions against the index can involve an excessive amount of active risk. And indeed, when the fund became concerned about the manager’s performance and took a closer look at what was going on, it found little wrong with the stock picking but some issues with managing benchmark risk.
“One of the difficulties, given the concentration of the index, was the impact of an incorrect weighting in a large stock,” Mansley explains. “Incorrectly underweighting Shell or BP in the UK index can be very significant. We recognised that Standard Life had a lot of strengths in stock picking, and in areas such as corporate governance. Rather than changing the manager, we thought it worth looking at changing the benchmark, and after some discussion we arrived at a trade-off between theoretical efficiency and practical implementation in the form of a benchmark with capped weights for single stocks.”
The result was what is sometimes described as ‘diversity weighting’: single stocks are capped at 2% of the FTSE All Share index. At the cost of slightly higher beta, exposure to smaller companies is increased and some of the sector biases are ironed out.
Since the change in January 2012, the difference in performance has been startling. The new benchmark weights have themselves added seven percentage points to performance but that is with the tailwind of a mighty small-caps rally. More intriguingly, the manager’s stock picking has added a further 13 percentage points.
“The manager has been revitalised,” says Mansley. “They are able to focus on what they are good at, picking the best small and mid-cap stocks, and they have felt free to go up to something like 70-80% active share against this new benchmark. We think this is interesting as an implementation of an alternative beta idea in an active management context.”
In that respect, the UK equities solution overlaps with the global low-volatility solutions the Environment Agency fund has implemented. For here it has chosen to benchmark two active managers against a market cap-weighted index rather than simply track a systematic low-volatility index. (Similarly, although the fund eventually opted to track a RAFI fundamentally-weighted index, Mansley says it was “a tight call” against simply hiring a traditional active value manager).
First of all, that throws up benchmarking questions. Why not measure these mandates against a more suitable index, in order to limit the confusion of a huge tracking error?
“We debated that a lot and just took a pragmatic approach,” Mansley says. “We expect lower volatility than the benchmark, of course, but that’s something we can only really assess over the long term. So we do keep an eye on the MSCI Min-Vol index and other indices on a shorter-term basis, just to help us monitor quarter-to-quarter performance. If the long-term returns from equity are 6-7% per year and the manager reckons that they will deliver two-thirds of the benchmark index, one other shadow benchmark we can use is two-thirds of the cap-weighted index plus the 2% extra that would bring us to the cap-weighted index return.”
Second, it throws up the question of how committed the fund is to the low-volatility excess return as an investment belief. It is analogous to the distinction that some investors draw between the risk premia that are paid to risks like value, smaller companies and illiquidity that everyone agrees on, and the rather more mysterious excess returns that accrue to strategies like low-volatility, that are more difficult to explain as risk premia within the framework of conventional finance theory.
If you are 100% committed to the idea that some kind of risk premium is paid to investors that hold low-volatility stocks, that raises the bar for rejecting a ‘passive’ exposure to that risk in favour of an active manager. Mansley articulates a cautious position.
“The investment rationale is important and we do think there is logic for these approaches beyond something someone has observed in a spreadsheet,” he says. “It is important that we can explain low-volatility intuitively, and with the behavioural-financed models, by the fact that benchmarked managers dominate the market and they see low-volatility stocks as high-risk stocks, even though they probably aren’t. But we also acknowledge that we can’t rely on this 100% and that funny things can come up out-of-sample. That’s why we get a lot of comfort from the idea that there is a certain amount of human judgement standing behind these things.”
The cost differential for adding that human judgement was modest, and fortune dictated that the benefits should become almost immediately clear: as the mandates were being funded in the first quarter of 2013, the market was nearing the end of a long love affair with quality companies that had inflated the valuations of many stocks that are overweight in low-volatility portfolios; soon afterwards, as tapering of the Federal Reserve’s quantitative easing appeared on the horizon, a good deal of that deflated again. Mansley says that the more active approaches avoided some of the worst losses of the low-volatility indices. In exchange, the Environment Agency fund has taken on some idiosyncratic manager risk, which is one reason why it mandated two firms and put a third on reserve.
One thing becomes clear when discussing what the Environment Agency fund has done in smart beta: it is not ideologically wedded to it as an abstract idea. Smart beta – or rather, the right types of smart beta – is one tool among many for solving specific portfolio problems. This comes through in what Mansley describes as his “frustration” that the industry has yet to come up with a range of smart-beta products in the fixed-income world comparable with those proliferating in the equities world.
“The anomalies in bond and credit markets are even greater than in equities,” he observes. “We have tried to get away from the benchmark mentality in bonds to some extent, and have considered buy-and-hold strategies, for example, but we would like to see more products and ideas out there for us to harvest. Perhaps we are still in the early days of this: the bond market has had things its own way for 20 or 30 years, and now that the tide is turning perhaps more innovative thinking will come to the fore.”
That sounds like a challenge. The Environment Agency Active Pension Fund is ready to hear about bond-market smart beta: for those willing to put in the effort to innovate, it will come with very specific requirements, but a genuinely open mind.