By Richard Phillips, Chief Investment Officer and Kevin Muir, CFA, Market Strategist, for EastWest Investment Management, republished here with permission
At first blush, weighting a stock index by market capitalization seems to make a lot of sense. After all, the goal of an index is to track the value of the stock market as an aggregate. What right does an index provider have to decide they know better than the invisible hand of the market when it comes to weighting the various stocks in the index? Passive investing is supposed to be just that – passive. And that’s exactly what market capitalization weighted indices do. They allocate resources based on the market’s judgement about each individual stock’s worth.
However, market capitalization based indices mean that as a stock’s value increases, it represents a larger portion of the index. Therefore, as Amazon increases in value, any new money allocated to the index is by its nature weighting Amazon more highly than before the rise. Does it make sense to buy more of a stock as it goes up in value? To some extent, grabbing hold of winning companies like Amazon seems like a good bet, but what happens when the market overpays for that growth? Surely there has to be a better way.
There are a whole myriad of different quantitative strategies that attempt to beat the simple market-cap based indices. Some indices weight the individual companies by a fundamental metric like revenues. Other strategies attempt to buy more of the less expensive companies by ranking them by price-to-book or some other filter. There are even those indices that use the number of employees to determine how to allocate to each company.
These strategies might work, but there is an even simpler one. How about not trying to predict which stocks will outperform and instead weight them equally? Surely something so simple can’t work. Or can it?
The Canadian Experience
Canadians have a unique perspective about the problems with market-cap based indices. During the Dotcom bubble, Nortel Networks rallied to the point where it represented 35% of the TSE 300 Index, which at the time was the primary Canadian benchmark. It was made even worse by the fact that its parent company BCE owned a big position in Nortel, and that it was trading as a Nortel-surrogate. When combined BCE and Nortel represented over 45% of the index.
Although the market-cap index did reflect the appreciation in the Canadian stock market as a whole, was it prudent for a fiduciary to invest in an index almost half of which rested on the fate of one company? For Canadian equity managers, it was an awful choice. Common sense would dictate that passively replicating the TSE 300 was an irresponsible decision yet as Nortel continued to soar, those managers who had done the “right thing” by underweighting Nortel were dramatically underperforming the index.
Herein lies the problem with market based indices. They are subject to the madness of crowds.
Have a look at this chart of the TSX 60 market cap index (orange line, here used as a proxy for the market-cap weighted TSE 300 index) versus the TSX 60 Equal Weighed index (gray line).
From late 1999 to the height of the Dotcom bubble in early 2000, the Nortel-heavy TSX 60 index rallied over 65% while the equal weighted index was up only 15%. Yet by the end of 2002 the TSX 60 index had given up all its gains and was now down 5% while the TSX 60 Equal-Weighted index had edged higher for a total return of 20%.
Obviously this made-in-Canada example is extreme. Rarely do individual companies represent such a large portion of an index.
The outstanding quant shop Research Affiliates has done an extensive study on market capitalized indices versus equal weighted throughout the world. For a detailed analysis, please see here.
We have summarized their findings into this chart:
As you will notice, for most countries, the equal-weighted indices outperformed their market-capitalization based counterparts.
What’s interesting is that when you dig into the data, this outperformance does not come at a cost of higher volatility.
In fact, on average, the equal-weighted indices outperformed by almost 100 basis points with 20 basis points less volatility.
The next question the sharp reader will ask is “if this strategy is so obviously better, why doesn’t everyone employ it?”
Equal-Weight Indices Drawbacks
There are two main reasons that stop equal-weight indices from being widely adopted for passive capital deployment.
The first is that for large capital allocators, equal-weighted indices have high friction costs due to liquidity constraints. Contrast that to a market capitalized index where the largest amount of capital is being directed towards companies most able to handle that inflow.
For example, Apple’s market capitalization is $741 billion while Torchmark Corp (also in the S&P 500) is only $8.2 billion. If a large pension fund or one of the big retail S&P 500 ETFs were to suddenly shift to an equal weighted index, that would mean these clients would be buying the same dollar value of Torchmark as of Apple. This is impractical for large pools of capital. The big money has no choice but to be invested in market capitalized indices. There is no other way to put that much capital to work without substantial liquidity penalties that would outweigh the benefits.
The second reason that investors and allocators, even those managing smaller pools of capital, don’t buy equal-weighted indices is tracking risk. Think back to the year 2000 TSX example. At one point the equal-weighted index had underperformed by more than 50%. It is difficult for investors to deal with that sort of underperformance. Although this is an extreme example, there have been (and will continue to be) long periods when equal-weighted returns less than market-cap weighted.
Look at the S&P 500 since the beginning of 2017. As the FANG stocks took off, the spread between the market-cap index and the equal-weight index exploded higher.
At the top of the market in late September, the S&P 500 market-cap index was up almost 35% while the equal-weight index was only at 26%. This 700 basis points came quickly and even now that the FANG stocks have sold off, the market-capitalization based index is still 6.5% ahead since the beginning of 2017. The volatility in the relative performance between the two indices creates an environment where investors or allocators doubt the strategy.
In fact, if we look backward over a longer timeframe, there were much worse periods when equal-weighted indices underperformed.
Let’s start with the S&P 500.
Although the spread between the two indices trends higher over time, looking at the dips yields some scary drawdowns.
Here are the periods when the S&P 500 equal-weighted index underperformed the market-capitalized index with the corresponding level of underperformance.
These periods of underperformance are measured in years, not months!
Let’s look at the same graphs for the TSX60 index.
There are plenty of periods when equal-weighted indices underperform their market-capitalization brethren and this “drawdown” of the relative performance of the equal-weighted index makes it difficult for investors to stick with the strategy.
It’s easy to look at the results and intuitively say investing in equal-weighted indices is a no-brainer. But the reality is that sticking with this strategy is much more difficult than it appears.
Although it is impossible for large smart long-term capital pools to invest in equal-weighted indices due to liquidity constraints, for most investors, this is a non-issue. The only question is whether the investor has the fortitude to withstand the long periods of underperformance. And make no mistake, there will be periods when this strategy lags. Yet over the long-run, equal-weighted indices potentially offer an attractive alternative to standard passive investing . Make sure you take the time to check out what’s available and think about equal-weight strategies for your portfolio.