Why It Pays To Know Your Index

It sounds like one of the most obvious questions an investor should ask when investing in ETF products – does the underlying index of my fund track the asset class I want exposure to?

If you choose to invest passively, the index is the direct reflection of your view. Despite being a fundamental factor in determining an investor's outcome, index selection is often a lesser priority for investors when choosing an ETF: it can come behind below tracking performance, replication methodology and it can even fall behind fees.

More Choice Means More Due Diligence

Most asset classes within the European ETF market now offer multiple index choices for each asset, and this is no bad thing for investors. The freedom to choose between indexes offers investors an important degree of flexibility; but differentiating between the indexes can be a much trickier task.

At Nutmeg, index selection is a critical driver in our selection of ETFs. Whilst most investors will, at a very minimum, be aware of the basics of the index (such as geographical exposure, sector bias and the individual constituents), determining how an index may perform versus its counterparts, and thereby how the index exposure reflects the underlying asset class, will be critical to the success of any allocation.

Wide Dispersion Of Returns

An index that does not truly reflect its asset class sounds like a misnomer. However, the growing range of indexes introduces a widening range of potential returns, even among those that track the same asset class.

Whilst there are many examples of this, we'd highlight the example of high yield indexes.

Below, we've compared the performance of two of the largest European-domiciled ETF strategies tracking the performance of the 0-5 year USD high yield market over a five year period. What is immediately apparent is the difference in performance. The Markit iBoxx index underperforms the BofA Merrill Lynch index by the equivalent to 12 months yield from either index – a significant deviation over a short timeframe.

The Markit iBoxx index is a low beta representation of the asset class relative to its peers. Compared with another – the Barclays US High Yield 0-5 Year Index – it is clear, even over a short period, that the Markit index captures much less of the upside risk than both the Barclays and BofA Merrill Lynch indexes.

Make Sure You Are Paid For Risk

And whilst there are multiple reasons behind this performance differential, including differences in fundamental characteristics of the indexes, and the potential premium paid for 'liquid' securities in the iBoxx index, this presents both tactical and long-term asset allocators with a serious issue. Given the high yield market is considered a high beta asset, an investor should expect a return consistent with that of the risk presented by the asset class. In this instance, both tactical and long-term investors face the prospect of a return that does not adequately compensate them for the risk taken, potentially underperforming alternative indexes and their return assumptions.

Equally important for investors to consider is how the index lends itself to passive management i.e. how the ETF tracks the index, turnover costs within the ETF and how these issues affect performance.

Index turnover should be a key consideration here, particularly in non-market cap weighted strategies, but also in those market cap indexes which rebalance more frequently (such as fixed income). Monthly rebalancing in less liquid environments will give rise to a sharp increase in transaction costs, reducing returns for the investor and widening the performance differential of the product and the index.

More Transparency Needed From the Industry

One of the biggest hurdles for investors is the lack of high quality information from index providers. Finding easy access to index return data and fundamental characteristics of indexes can be very difficult, and can require multiple index licences, even when using premium data systems such as Bloomberg.

In fact, away from the major equity indices, there are several large products tracking building block assets where there remains no index factsheet available. If we can't adequately analyse an index, why would we want to track it?

Given that index fees are a core component of ETF expense ratios and are ultimately paid for by investors, not the ETF issuer, we believe it is high time the index providers recognised the end-investor as their true customer and made index information readily available at no extra expense.

Shaun Port is chief investment officer at Nutmeg