David Barron and Jennifer Shering, Legal & General Investment Management
The first generation of ESG strategies excluded whole sectors from investors’ portfolios. Such approaches are still widely used, but investors may be underestimating their impact on portfolio diversification.
We know some investors value the peace of mind that comes from owning lots of different assets in their portfolios, so their risks aren’t too concentrated in any one area. They want diversification, in other words.
But equally we know that some investors want to reflect environmental, social, and governance (ESG) considerations in their portfolios.
For some, that can mean excluding fossil fuels – typically meaning the entire energy sector – from their portfolios. In a sense, these two desires – building a diversified portfolio and avoiding vast swathes of the economy – are mutually exclusive.
We wanted to investigate this apparent conflict in order to quantify more accurately the relationship between negative screens and portfolio diversification in equities. Put simply, are they friends or foes?
SECTOR INSPECTOR
As a starting point, we looked at the correlation of each sector in the MSCI World index to that parent index. This gave us a longterm picture of the diversification dividends yielded by each sector, as illustrated in Table 1.
We see here that some sectors have consistently been diversifiers. These include consumer staples (which encompasses tobacco, of course), healthcare and utilities.
However, we must remember that correlations between sectors are dynamic, not static. For example, energy was a diversifier through the 2000s; technology and then financials have been highly correlated to the MSCI World index for long periods, but this has changed at inflection points rather than staying fixed. Correlations can switch unpredictably at key moments and so excluding sectors can deprive investors of diversifying assets unexpectedly or expose them to greater risk if the retained sectors converge in periods of market stress.
WEIGHT WATCHERS
This possibility prompts another question: when sectors are omitted from a marketcap portfolio, how is their index weight redistributed among the other sectors? This can obviously lead to unintended risk exposures if it concentrates a portfolio in sectors that are either more or less correlated to the index. In the former case, the portfolio could end up with a higher beta than desired; in the latter scenario, the portfolio may not offer the required market performance.
As Table 2 displays, when energy is excluded the largest overweights have tended to be to consumer discretionary, financials and technology.
Comparing this with Table 1, we see that the overall effect of rebalancing away from energy and into these three sectors – each of which has a relatively high correlation to the MSCI World index – is likely to be an equity portfolio with an above-average beta. The consistent diversifiers – consumer staples, healthcare, and utilities – receive more modest upgrades.
Again, though, we have to reiterate that these weights will vary through time – not always to the investor’s advantage. The overweight to financials, for instance, reached its zenith just in time for the financial crisis.
Turning to the present day, the most significant overweight in the MSCI World excluding energy is information technology at 0.98%. This additional exposure to tech stocks has important consequences for investors, not least for those who have already chosen to overweight technology elsewhere in their portfolios.
An additional point is that we have focused on global developed market-cap exposure here, which has well over 1,000 securities across more than 20 countries. For investors thinking about regional allocations, the impacts of reweighting can be even more pronounced. In the UK equity space, for example, three energy stocks – from just two issuers – make up more than 15% of the FTSE 100. Exclude these and the redistribution effect can lead to an overweight of almost four percentage points to financials within that adjusted index.
MATTER OF FACTOR
We can also look at the factors – or risk premia – that the energy sector has contributed over time.
The decline in the oil price from 2014 clearly left energy heavily overweight the value factor, although this has moderated of late. This has led some to the erroneous presumption that such negative screens systematically underweight value.
This is not the case. Excluding energy in recent years has certainly left portfolios underweight the value factor, but not so long ago, quality and momentum were major forces in the energy index.
Investors may have been willing to forgo value exposure over the past few years as that factor has underperformed, but would they have been so happy to minimise the quality and momentum factors under previous market regimes?
Although there isn’t a formally recognised dividend or income factor, we would also note that excluding energy – and tobacco – is likely to have impaired a portfolio’s yield through this period.
PORTFOLIO PERMUTATIONS
With all this in mind, traditional negative screens may be most appropriate for investors who are obliged to avoid certain sectors. But other investors may be able to preserve the diversification benefits from sectors like energy without sacrificing their ESG criteria by integrating those criteria into their investment process in more nuanced ways.
At Legal & General Investment Management, we believe ESG scoring gives us a framework for engaging the companies in which we invest and also allows us to tilt portfolios to reflect ESG criteria while maintaining diversification.
