Most, if not all, of the commonly used stock markets today are measured by market capitalization-weighted indexes. Many believe these indexes represent the best way of measuring how a market performs. In fact, market cap-weighted indexes have dominated the assessment of equity market performance over the past 30 to 40 years, as they are seen as representing the true measure of equity market return.
Market capitalization-weighted indexes clearly have some advantages. Theoretically, they represent an appropriate broad market benchmark against which to compare the performance of active managers. Also, market cap-weighted indexes are easy and cheap to track.
However, market cap-weighted indexes have three significant flaws.
- Market cap-weighted indexes tend to be backward-looking. Although economic theory tells us that share prices are today’s value of the sum of future expected returns, and therefore forward-looking, it is clear that the shape of the index is driven heavily by past success.
- Market cap-weighted indexes are prone to risk of concentration. This is readily observed at a country level (Japan reaching over 40% of the MSCI World Index in 1989) and at a sector level (the technology/media/telecommunications bubble in late 1999/early 2000 saw these combined sectors peaking at around one-third of the MSCI World Index).
- Market cap-weighted indexes are prone to mispricings or, at an extreme, asset price bubbles. By their very nature, these indexes will capture the full effect of asset price bubbles, and we know that this effect leads to increased volatility and reduced return over the longer term.
The question then is: Can these three drawbacks be addressed by reshaping index benchmarks or passive management approaches? Or should investors rely on ‘unconstrained,’ ‘benchmark-agnostic’ active managers to do this job for them?
Backward or forward-looking bias
The backward-looking bias is a difficult problem to overcome given that share prices are generally regarded as discounting future growth prospects. So, unless we have sufficient foresight or the ability to predict which sectors or shares will grow quicker or for longer than the market expects, it is going to be tough to address the backward-looking bias of indexes.
However, even without accurate forecasting, over the long term both emerging markets and smaller companies should deliver higher earnings growth than developed market large-cap companies. Therefore, strategic over-allocations to these two areas (mindful of not overpaying at the outset) will tilt a portfolio to higher future returns, noting the associated risks.
Concentration of risk and diversification
There are both simple and complex methodologies for constructing more diversified portfolios than market cap-weighted indexes. The development of global equity portfolios on a fixed regional weight basis was a response to the excessive concentration of risk that had arisen in Japan at the end of the 1980s. Capped indexes (e.g., where the weight in any one stock is limited to no more than 10%) were a response to excessive single stock concentration such as the 30%+ weighting that Nortel once represented within the TSE 300. Both the fixed weight and capped index approaches have proved successful, delivering modestly better returns per unit of risk over the period of their existence.
Taking the diversification approach to an extreme would be to create an index that is equally weighted in all the constituent shares. In terms of diversification of stock-specific risk, this is effective since each company will have the same weight as every other company and therefore there will be negligible concentration of risk at a company level. However, if this approach is considered in more detail, some disadvantages emerge. For one, it would not be investable on a large scale because investors could not acquire as much of small companies as they could of large companies. Secondly, there would be considerable concentrations of risk at a sector level: a sector made up of a large number of small companies would have a much bigger allocation than a sector made up of a small number of very large companies.
If equal-weighted indexes are not a realistic alternative to market cap-weighted indexes, there are other, more sophisticated, diversification-based approaches that can be applied. At its simplest, the theory is that an investor wants to gain exposure to all parts of the market while seeking to avoid concentration of risk. To achieve this objective is not, however, straightforward because ‘concentrations of risk’ are a function of share price volatilities and correlations that, unfortunately, are not constant (or even stable) over time.
Notwithstanding these difficulties, there are now a number of products on the market that represent effective ways of constructing diversified portfolios and reducing risk concentrations. The sophistication of these approaches, including the skill of the manager in designing the portfolio construction process and the optimizations behind it, inclines us to regard these more as active management strategies than index-tracking or passive management.
Avoidance of asset price bubbles
The avoidance of asset price bubbles appears more straightforward. Market capitalization-weighted indexes are based on the value of a company being its number of shares multiplied by the price of those shares. But we know that shares can be mispriced—driven by ‘animal spirits’ rather than rational investors assessing the underlying value of the business. The simple answer may be to construct an index of shares that does not use the market price of those shares as one of its inputs.
The alternatives to price-based approaches seek to find more stable measures of the ‘true value’ of a company and use these value-weights as the basis for index construction.
GDP-weighted indexes offer this type of value-weighted index. The approach can be implemented in a variety of ways, but the central thesis is that the capital allocation decision is linked to the size of a region’s or country’s economy, not the value of its stock market. While the relative GDP weights of countries are highly stable, and this approach has merit, it is a blunt instrument that fails to take account of the increasing globalization of markets and fails to recognize that domestic stock markets are decreasingly representative of the national economy.
A more sophisticated approach to value-weighting is to construct an index in which each company is sized according to its economic or fundamental value rather than a value driven by its share price. Companies could be sized in the index according to any number of different variables representing the economic scale or value of the business, such as profits, earnings, dividends, cash flow, employment, etc. What would be missing from the index could be thought of as ‘market sentiment’—i.e., one unit of cash flow would have the same value whichever company it belonged to (not reflective of the market’s forward looking positive or negative view of that company) and the company’s weight in the index would increase as its cash flow (if that were the only measure used) grows relative to the market as a whole.
This approach is appealing. First, it is reasonably straightforward to construct an index based on a small number of parameters, and this index will be highly investable (since large companies as measured by economic value generally exhibit large market capitalizations). Second, the investor sees that the investment is linked to the underlying growth of the business in which they are invested and is not driven by the vagaries of market sentiment. Over time, such a value-weighted index combined with sensible rebalancing might be expected to deliver returns that are superior to the capitalization-weighted index because it will avoid the value-destroying effects of a high exposure to asset price bubbles.
Taking the technology/media/telecommunications bubble, for example, simulated track records for global value-weighted indices show that these indexes would have had a significantly lower weight in tech stocks (in the region of a 10% to 20% underweight relative to market cap-weighted indexes) prior to the bubble’s burst. The performance of value-weighted indexes would therefore have lagged the market cap-weighted index on the way up, but significantly outperformed later on.
It is worth noting that, while we might expect value-weighted indexes to better control an investor’s exposure to asset price bubbles (relative to market cap-weighted indices), such indexes should not be expected to mitigate an investor’s exposure to all market crashes. As investors are now acutely aware, the equity market can suffer a broad-based sell-off in the absence of a bubble in any given part of the market.
Style impact—value bias
The value-weighted index approach can be seen as an effective equity market beta capture mechanism, but with an inherent value (using ‘value’ in the traditional sense now) bias relative to a market cap-weighted index. However, the exposure to the value style varies over time as a function of how far the market price has moved away from intrinsic value (as measured by the value weight) and has over some periods resulted is being overweight to stocks and sectors that may be regarded as growth stocks.
The value bias relative to a market cap-weighted index is not of itself a negative. Fama and French have demonstrated that value investing as a style was expected to deliver excess returns over long periods. In addition, there is a significant body of academic evidence in favour of the existence of a ‘value premium’ over the long term. But the value-bias (relative to a market cap-weighted approach) does mean that the value-weighted approach may produce a materially different pattern of returns from the market cap-weighted index (i.e., have a significant tracking error against standard index benchmarks) over a particular period without any guarantee of an equivalent stream of excess returns. However, many investors should find this caveat to be acceptable; what we are trying to do with the value-weighted index is, over the long term, capture the equity market beta (return) in a way that mitigates the well-known flaws of the market cap-weighted index.
Conclusion
While market cap-weighted indexes have material flaws, there are ways in which these flaws can be addressed. A well-diversified, actively managed equity portfolio is one approach. However, not all investors have the desire or capacity to put such a program in place. Alternative indexation approaches, including value-weighted indexes, are cost effective and simple ways investors can capture the equity market return while mitigating some of the flaws of market cap-weighted indexes.
David Zanutto is a principal at Mercer’s Calgary office. david.zanutto@mercer.com