How indices change behaviour

by Hansi Mehrotra

6th June 2016

In the evolution of investment management, the lines between active and passive are becoming blurred. Indices no longer reflect “the market”, or not many of them do. Investors should be aware of the implications.

Kathryn McDonald, director of investment strategy at AXA Rosenberg Investment management, gave an interesting series of presentations in Australia recently on the subject ‘Indices, Benchmarks and the Long-Run Investor’. The key to understanding and utilising the emerging differences between old terms is to focus on the long term.

McDonald points out that index construction began as an attempt to reflect the market as a whole, going way back to the late 1800s. “The Dow [Jones Industrial Average] allowed us for the first time to start talking about equities as an asset class,” she says.

The oldest indices, such as the Dow and Nikkei, are price weighted, while most others that are used as benchmarks are market-cap weighted. They met, and still meet, some very basic needs:

  • they are barometers of markets
  • they define security selection universes and are proxies for asset classes
  • they are performance benchmarks and, importantly, measures of opportunity cost, and
  • they are the basis of index-replication investments – ‘indexing’ and the plain vanilla ETFs.

With the increasing power of the major index providers, coupled with the early trend to replication followed by the more recent trend to ‘fundamental indices’ and ‘smart beta’ – scientific strategies which systematically exploit market anomalies and better reflect an investor’s actual portfolio – the truth is that indices are now shaping reality instead of attempting to reflect it, McDonald says.

She questions whether anchoring to traditional indices is at odds with long-term investing goals. “Benchmarking should work to help ease friction between asset owners, investment managers, trustees and consultants. The typical investment mandate contract attempts to point all parties in the same direction by specifying the universe and limiting the risk to be taken.

But Paul Woolley, who is well known in Australia for founding the UTS Paul Woolley research centre with Ron Bird, and other researchers have shown that managers are forced to buy stocks that have already run up in price. In a paper published this March by the London School of Economics, Vayanos, Woolley and Dimitri say: “The majority of trades bear no relation to fundamental worth and are focussed instead on window dressing and gaming.”

There are also other pressures on investment managers which result in an emphasis on shorter-term returns, such as asset-gathering pressures related to short-term alpha. “This is obviously in stark contrast to the goal of long-run value creation,” McDonald says.

While the idea of having “an independent third party, such a traditional index, is very attractive, we must remember that no index is unbiased, she says.

In her presentation McDonald shows the steps which can be taken toward long-run investing as related to the use of indices of all sorts. These include:

  • Distinguish between index and benchmark – the terms should no longer be used interchangeably
  • Whether a benchmark remains cap weighted or something more bespoke (more like a “normal portfolio”), AXA Rosenberg advocates longer evaluation cycles
  • Resist the temptation to reduce manager analyses to comparisons along simplistic dimensions, and
  • Find ways other than short-term performance to assess whether a manager is on track – good investment ideas are often met with lumpy returns.

AXA Rosenberg also believes, with bespoke or smart-beta-style index strategies, investors should invest with managers which both design and implement the strategy for better alignment.

The good news for investors is that McDonald also predicts a continuation of downward pressure on fees, especially for scientific investments. “We will probably see fees for smart beta or risk premia strategies, or whatever you want to call them, come down much further.”

This article originally appeared at http://www.ioandc.com/the-old-and-new-of-indices-how-they-change-behaviour/