Monday, 17 November 2014

That Syncing Feeling

Index funds could undermine diversification.

Update 2 Dec 2014: The events in this post may be less likely. To wit:

The description of the S&P 500 index as a 'cap weighted index' opens two interpretations as to how it works:
  • Constant number of shares - a re-balance sets a constant proportion of shares while the weightings change with price movement. In this case, movement in price 'self-adjusts'.
  • Constant weighting - what I assume, where the weighting remains constant (until rebalancing) and the proportions of shares held must change to compensate for price movement.
I haven't found any definitive detailed answer on which it is. For every article supporting self-adjusting indexes and index funds...

... There is another implying that index funds are weighting-locked.

Additionally, each index holding is multiplied by a float factor to account for outstanding shares, and proprietary divisors are used to generate the index (which affects fund performance in so far as it is a benchmark to be tracked.)

Irrespective of ambiguity or misunderstanding, the workings of indices and index funds are more complex than I describe in this post. The good news is that this complexity seems to make my nightmare scenario even less likely. My thanks to reader GL for patiently working through this with me.

Ultimately too, my points about index funds being too small (currently) for a fundamental defect to swing the market, and that they remain worthy of investor consideration still stand, so you may want to read on.

Imagine this scenario: shares in a major component of the S&P 500 index tank. This has two immediate effects. First, the S&P 500 drops slightly. Second, while the drop in the company's share price reduces its weighting in the market instantly, its weighting is preserved in the index. Investors react by selling out of the market generally, including from index funds that replicate the S&P 500. To redeem shareholders and continue replicating the weighting, index funds have to sell broadly from other positions as well as from the flailing stock - jettisoning the 'good' along with the 'bad'.

How does this affect diversification? Although established in investment-speak as a means to reduce one's exposure to the ups and downs of any one particular asset, diversification does not simply mean to hold multiple shares. These shares must be as non-correlated as possible. Holding, say, Intel and Microsoft is not true diversification as one will follow the other. A hypothetical example of true diversification would be holding a firm that made jet-skis and another that made ski gear (although both would be in the same 'consumer discretionary' industry sector). In the above scenario, both correlating and non-correlating stock are sold off.

Shedding shares across the board may affect diversification beyond the fund itself. If the fund or the weighting change was large enough, selling out of a non-correlating stock would put downward pressure on it at a time when it would otherwise preserve its price - if not see it rise - by virtue of being considered safe. According the proper weighting to each stock would only occur when the index was rebalanced. In the case of the S&P 500, it could be anywhere up to three months, by which time a sinking behemoth could have dragged a smaller contrarian victim down with it.

Of course, non-correlating stock has been dumped during wholesale market sell-offs throughout history. Crashes have happened without the aid of index funds. And though index funds were around in 2008, they played little part in the crisis that ensued.

That is because they are small. Of the USD$18.5 trillion S&P 500 (the remainder of US market cap is negligible), iShares' and Vanguard's tracking products make up 0.008 and 0.182 trillion respectively. They are among the largest, but not the only two, and the index ETF market is growing at about 28% per year.

Index funds also need not dominate to influence the greater market. Though all the companies on the S&P 500 are large, some are far larger than others. As an aside, some complain that the weighting formulas used to compose indexes inordinately favour industry sectors such as finance. I think this may simply be environmental. There is no saying that a counter-correlating sector would not have its turn at the top as business conditions evolve. The 'perfect storm' above would then hamper current index underdogs riding the business cycle upwards.

To illustrate, let's look at how iShares' Core S&P ETF dealt with shares from two non-correlating sectors in the period 7 to 10 November 2014.
SectorHealth CareConsumer Discretionary
Weight Change0.16%0.16%
Market Value$14,334,269$87,142,334
Market Value Change$28,633$298,575
Market Value Change (%)0.20%0.35%
iShares Bought Daily216693150181
3-Day Total Volume9391173521696
% of Daily Total Volume69.22%12.79%

The same tiny weight adjustment in the smaller company formed a larger proportion of the daily volume of its stock traded. iShares replicating the index could have forced PDCO up when really it should have moved counter to Netflix.

Relax. The algorithmic apocalypse is nowhere in sight. For starters, index funds have a lot of leeway in how they adjust their weightings, and as long as they do it at different times in different ways, we need not start hoarding the tinned food. That said, I believe that their composition and how they synchronise with their base index is already affecting volumes, if not prices.

Furthermore, while the interconnectivity in tracking an index leads to systemic risk by nudging component stocks - correlated and non-correlated - towards moving in lockstep, index funds and ETFs are still a simple way to be exposed to a range of shares. Despite the faults of passive index funds, active ones - managed by experts who pick and choose stocks for growth and diversification - fare little better. And even if you were to try and start your own DIY fund, you would not be able to counter a market-wide replication effect on diversification.

For the truly defensive lamenting the conformity of the stock market, you can diversify out of it; property, bonds (although they also froth when there's lots of easing), gold, and even wines and art. But considering that the median 10 year annualized return of the S&P 500 to 2013 is 13.83%, a weakened diversification - seen by some as glorified mediocrity anyway - may be an acceptable price to pay.

No comments:

Post a Comment