How Not To Invest By Index

Dividend-weighted ETFs have been demonstrating to us how not to construct and manage a portfolio.

Over the last few months my email inbox has been filling up with news announcements from DJ Stoxx, and a great number of them relate to changes in their dividend indices.  As companies cut or suspend their payouts in response to the deteriorating economic environment, they are automatically replaced by others in benchmarks like the DJ Stoxx Select Dividend 30 Index.

As a little exercise to assess the impact of such changes on those who have invested in ETFs tracking dividend-weighted share indices, I took the last month’s index deletions from the Stoxx website news section, and looked up the performance of the relevant companies’ shares over the year leading up to the deletion date.  I also worked out the share price change (where available) from the deletion date to yesterday’s market close.

Here are the results.

For the seven companies that have been taken out of different Stoxx dividend indices over the last month, for having cut (in the case of Citigroup) or cancelled (the six others) their ordinary share payouts, the average return over the year to the date of deletion from the indices was a pretty appalling -82.37%.

To add insult to injury, after the companies’ respective deletion dates, their share prices then rallied by an average of 22.35% to yesterday’s close.

So investors in dividend-weighted ETFs, having seen a series of such index changes over the last year, have found themselves facing the worst of all possible worlds.  Having (presumably) invested for income, they found their income vanishing, suffered a loss of four fifths in capital terms on the affected stocks and, ironically, would probably have been better holding on to the relevant equities once the dividend cuts were confirmed.

If some ETF providers claim to be offering “intelligent” indices as the benchmark for their funds to track, dividend-weighted funds must have been at the other end of the spectrum, as an example of how not to do it – at least since the onset of the credit crunch.

Of course everything is easy in hindsight, and there will no doubt come a time when these ETFs will shine.  But there’s surely a serious point here about the suitability of this index construction methodology as a way for investors to allocate their capital.   Just as the proponents of the various model-driven indices (such as RAFI) point to the drawbacks of capitalisation weighting (overweighting overvalued stocks, and underweighting undervalued stocks) as a selling-point, so the designers of indices that are meant to offer yield to investors must be wondering how to do a better job than this.