Whatever happened to the ‘index effect’?


Almost since indices were invented, primarily as a form of measurement, their power to influence the behaviour of market participants and therefore the whole market as defined by those indices has been well understood and sometimes criticised, particularly plain vanilla market-cap indices.

But in the past five years this power has considerably weakened, according to a research paper, ‘A Golden Goose Goes Missing – The Curious Disappearance of the Index Effect’, published by Axioma Inc, a US-based risk management and portfolio construction solutions provider.

The paper says that the change began in 2013, especially impacting large and mid-cap stocks. It observes that this change has run concurrently with the significant increase in the popularity of index investing, which is itself curious. One would have thought the converse would be true.

Anthony Renshaw, the director of index solutions at Axioma, says that one potential explanation for the effect’s weakening is that ETF market makers trade on price disparities as soon as they occur, eliminating any sustained positive or negative price movements. This is possibly evidence that ETFs add liquidity to market.

The index effect is the phenomenon whereby stocks added to – or deleted from – an index experience positive – or negative – excess returns in the days immediately before they are officially added to (removed from) the index. From the perspective of 2018, meaning a time when passive investing is a sizeable market segment, an explanation of the index effect seems obvious, the paper says.

“Managers of equity index products are required to track the official index closely and have little leeway to take advantage of announced index changes in advance. Other traders know that there will be demand to buy or sell the additions/deletions) on the official rebalance date, so they buy or sell the additions/deletions beforehand with the intention of profiting when the passive indexers are forced to update their holdings.”

One interesting finding de-bunks the theory that less-liquid names will be impacted more by the index effect. Axioma found that this was true only for deletions from the index, not for additions. Similarly, the researchers found that the index effect was still, in the past, present for unscheduled deletions but not for unscheduled additions. Most unscheduled changes for market-cap indices are for deletions and usually after a period of stress for a particular stock.

Of interest to some smart-beta providers and sector ETF promoters is that paper finds the index effect has also been present among sector sub-indices, as measured by changes within the US market GICS (Global Industry Classification Standard).

Sadly, the latest research does not prove any particular cause of the weakening index effect, but, rather, presents some plausible explanations, such as the impact of ETF market makers. “Of course, there are other potential explanations for the weakening of the index effect. Since 2013, most markets have experienced low volatility, reduced trading costs, and bullish returns. These conditions may well explain in part why the index effect has weakened,” the paper says.

“Perhaps the most important takeaway from the present research is the recognition that the index effect has evolved substantially over the last few years. The excess returns observed one or two decades ago may no longer be achievable, at least for large and mid-cap stocks. One should be cautious about index effect results that report over, say, a 30-year time window. Such an approach will undoubtedly provide the researchers with a high degree of statistical significance for their results, but the benefits they describe may only be available to portfolio managers who can travel back in time.”

– G.B.