On the economic consequences of indexing..

There has been some recent comment on a 2010 NYU research paper into the “Economic Consequences of Indexing”.

My take, and it has been my take for a long while, is that passive benchmarking strategies need to pay more attention to significant valuation/allocation differentials and to expand the operational parameters of their benchmarks to manage these risks. This does not mean active management and it is risk focussed rather than return focussed.   To expand further with a few points might be worth while.

  • Number 1buying an index can involve a trade off between between risk (it may be a higher risk allocation) and return (there may be strategies with higher potential return), but this trade-off is, to varying degrees, offset by the lower costs.   For investors with limited low cost active management options, an indexed investment has been shown, over the long term, to more than make up for the additional valuation and allocation risks taken on by index allocation.   
  • Number 2 – believing that markets are efficient and that all times an indexed allocation will represent an optimal allocation to the market is naive: in 2000 we had valuation and allocation issues with respect to large versus small and mid and growth biased technology/media/telecom versus value biased investments.  In 2007 we had over valuation and allocation issues with respect to the financials.  

Passive benchmarking to major indices is designed to reduce the costs of management and to avoid compounding performance risks, but the primary benefit is cost.  

  • Number 3 – The message from the paper is not that indexing creates valuation, allocation and risk/return issues, but that any allocation strategy that overly favours one segment of the market to the detriment of the other creates valuation and allocation risks.   At times, these differences can be significant and need to be assessed in managing allocation.
  • Number 4 – You either decide to manage these risks or not, but you need a rationale as to why and how.

With regard to 4:

The lowest cost option is to ignore these risks and to use broad based indexes that capture as much of the market as possible and rebalance to a strategic allocation over time.  

The problem here, in an out of equilibrium world, is that the type of broad strategic allocations used by many are not equilibrium based structures (i.e., if the market was properly valued what would the allocation to a given market cap, sector, high yield, growth be?), in the sense that you can be underweight in an over valued market and overweight in an under valued market and end up making costly buying and selling decisions.  Again, low cost management options are really there to mitigate these risks and rebalancing is primarily to avoid excessive or insufficient risk exposures. 

The other, higher cost option, involves adjusting allocations to these benchmarks in response to valuation and allocation issues.   In this case you would need an assessment of equilibrium valuation and equilibrium allocation with more dynamic benchmarks .

I tend to believe that it is important to pay attention to valuation issues and to adjust allocations to key index components accordingly.   Furthermore, where you are managing assets to meet financial needs/liabilities over time, paying attention to valuation anomalies allows you to fund these liabilities more efficiently – you are selling relatively high, reallocating to the shorter end of the portfolio, without having to reinvest in a volatile asset class with uncertain return over a given time period. 

The key point though is significance, and significant valuation differentials, in my opinion, should be monitored, managed and taken advantage of.  Historically, the market has tended to ride valuation differentials for long periods of time – the idea is to take advantage of significance and not fall prey it..   

Passive benchmarking strategies need to pay more attention to significant valuation/allocation differentials and to expand the operational parameters of their benchmarks to manage these risks.  This does not mean active management and it is risk focussed rather than return focussed. 

Leave a Reply