UK: How to aim for the impossible
Rudyard Ekindi, director at NEST, discusses how the scheme hopes to achieve long-term and strong performance
When I was a child my mother used to ask me if I wanted a vanilla or strawberry ice cream. Naturally, my answer was both. As investment professionals we are often faced with multiple choices. Investment practitioners, academics and the press have long found themselves involved in debates on a single solution to a single problem. DB or DC? Hedged or unhedged? Market cap indices or fundamental indices? Sometimes the answers to these questions are not a case of choosing between binary opposites.
NEST is a new initiative that aims to provide a pension scheme suitable for a large range of savers at an affordable price, over many years and economic cycles. It is important for us to ensure we look at the best ideas from all segments of the financial and insurance industries, without excluding anything as a matter of principle.
The June 2012 issue of IPE presented a special report on smart beta and the alternatives to market-cap weighted indices. We challenge the notion that an investor should have to choose one over the other.
Here we look to address the issues at the total portfolio level and consider the choice of a type of index or benchmark as a strategic decision that needs to reflect the risk/return profile of the investor. For NEST, this is a top-down decision rather than bottom-up decision based on the methods of construction of one benchmark or another.
To address the question of the techniques and methodologies to consider when making global risk/asset allocation, we focus on the two mainstream categories - mean variance optimisation (MVO) and risk parity (RP).
Typically, market cap weighted indices support the use of MVO asset allocation techniques, while risk efficient indices (smart beta) support techniques such as RP.
When making asset allocation decisions - we prefer to talk about ‘risk allocation' decisions - we aim to identify the risk factors that are more likely to be rewarded over the investment horizon and try to avoid or minimise exposure to risk factors that are less likely to be rewarded.
The first step in this process is to identify the relevant risk factors and assess which ones will be better rewarded from a risk-adjusted perspective.
What do we learn from looking at market behaviour in different economic environments?
Empirical evidence shows that there are three types of market behaviours:
• Clustering - When all assets provide comparable risk/reward characteristics
• Divergence - When assets' characteristics move away from each other
• Convergence - When assets' characteristics move closer to each other.
Our analysis shows in a diverging market, a market cap-weighted MVO approach is likely to outperform, in a converging market an RP approach is likely to outperform and in a clustering market, an RP approach is also likely to outperform.
From this level of understanding, NEST has established a strategic approach to risk management, and an in-house asset allocation system developed by our investment team to implement this strategy. The aim is to provide our members with sustainable fund performance over the long term.
We have back-tested returns and volatility over the last 17 years for three portfolios:
• A portfolio rebalanced each year using MVO
• A portfolio rebalanced each year using RP
• A portfolio rebalanced each year using 50% MVO and 50% RP.
Both the RP and MVO portfolios each had periods of outperformance. Out of 17 years, there were nine years when RP outperformed and eight years when MVO outperformed. One might also notice that the swings between RP versus MVO domination are quite frequent. In the end, for this particular period from 1995-2011, on average RP performed significantly better in terms of return, but it also carried a lot less volatility. This means a much improved risk/return trade-off.
But if we were to look at the last 20 years of the previous century we would have an overall better performance of MVO.
We have chosen to blend MVO and RP at 50/50 to provide a more sustainable risk/return profile that aligns with our long-term focus. A further benefit to this approach is that we do not have to switch our investment process every couple of years. We are also less vulnerable to estimation error, which is inevitable when using financial models/techniques.
Just like the ice cream dilemma, we understand that when seeking the best outcome - when it comes to questions of choosing between one option or another, the answer is actually to have both.
Rudyard Ekindi is director of investment research at NEST