Two more reasons to be careful with the ETF ‘sausage machine’

You may have decided to invest in a particular type of index-tracking fund, but are you sure you are achieving precisely the kind of exposure you were after?

We recently argued, here on The Value Perspective, the greatest exposure investors receive from a value-tilted exchange-traded fund (ETF) is not, as you might expect, value-oriented businesses but very large companies. This was the basis of our sausage machine analogy and the view that Passive funds, like sausages, are only as good as their ingredients.

The reason for this bias towards large companies, we suggested, might be the business model of groups that create stockmarket indices – after all, their clients are not end-investors but the ETF providers who pay them a fee to replicate an index. This fee is typically a fixed fraction of ETF assets under management and, in order for these fees to be large and everyone to be happy, the ETFs also have to be large. Billions of dollars large.

The easiest and most practical way to do this is by ‘weight’ – that is, by reference to the size or ‘market capitalisation’ of the constituent companies – and that is fine, just so long as you believe markets are totally efficient and you want to track a broad stockmarket index. But what if you had formed the view markets were not totally efficient? What if you thought now could be a good time to buy value stocks?

After all, as the following chart shows, they have just suffered their worst decade relative to growth stocks since 1937. Say, then, you like the look of value and, while you acknowledge our concerns about their bias towards larger companies, you still like the sound of a value-oriented ETF rather than an active fund manager – do we have any other sausage machine-type misgivings?

Annualised outperformance of value over growth (10 yrs rolling)

As it happens we do and, since we just mentioned the word ‘active’, let’s first consider the question of ‘active share’. This still relatively little-known academic metric, as we have discussed in articles such as Active stance, measures the extent to which the weightings of a fund’s holdings differ from those of its benchmark index. The lower the active share percentage, the closer a fund is to being an index tracker.

Now, when we consider the active share of the MSCI World Value ETF in relation to the broader MSCI World index, we find it is just 22.7%. In other words, if that ETF were an active fund manager picking value stocks, it could rightly be branded a ‘closet tracker’. Certainly, you could take the view that, rather than tracking ‘true value’, a market-capitalisation weighted value ETF is a closet tracker of the broader index.

Another problem with buying the MSCI World Value ETF is that, despite index provider MSCI’s detailed description of how the relative value and growth characteristics are weighted and its assurance some stocks are assigned partial weights across both the value and growth indices, the vast majority of stocks that go into the sausage machine are essentially called as either ‘value’ or ‘growth’.

Of the 1,654 stocks in the MSCI World index at December 2016, only 208 – just one in eight – were partially assigned to both indices. With 678 stocks going exclusively into the MSCI World Value index and 768 into the MSCI World Growth index, meanwhile, you might think of the value/growth assessment process as different types of meat being prepared for the sausage machine.

Most stocks enter the mincer and come out into just one container – either marked ‘value’ or ‘growth’. As at December 2016, for example, 90.5% of the weight of investments in the MSCI World Value were only present in that index – as opposed to its growth counterpart – meaning fewer than one in 10 were stocks where there had been been a ‘smart’ splitting of the business to reflects its growth-and-value nature.

What this all means, therefore, is we are left with a value index that is buying roughly the cheaper half of the market while ascribing significantly more capital to the largest stocks. And the big problem in that equation is that, in the eyes of The Value Perspective, the largest stocks in the cheaper half of the market are really not that cheap.

As of December 2016, the largest 20 investments in the index represented 26% of its total These were stocks with market capitalisations between $141bn (£120bn) and $480bn, which on average traded on 22x their 10-year average earnings. Trading on 27x the average profits the business makes, the most expensive of these was the third largest holding – US pharmaceutical and consumer goods giant Johnson and Johnson.

The reason so many expensive stocks are in the index, we suspect, is because ‘value’ is defined as being cheap in the context of the country in which the stock is listed. Thus, if the US market is expensive on the whole – as we believe it is – then a stock could make it into the MSCI World Value index simply by virtue of being less horrendously overvalued than its compatriots.

Either way, the proof of the sausage is in the eating and, as you can see from the following chart, the MSCI World Value index has done a fine job of tracking the broader MSCI World index over the last 10 years. Here on The Value Perspective, we would strongly suggest investors looking to buy value on a global basis should buy stocks that are cheap versus others on this planet – not just versus those of the same  nationality.

Cumulative index performance (Gross returns)

Source: MSCI, in USD ($). Past performance is no indication of future performance and may not be repeated.