Can equal weight solve our concentration crisis? Not so fast…

The startling conclusion is despite the Russell 1000 nominally providing exposure to its namesake number of stocks, the index affords an effective diversification of only 59 stocks.  This figure represents a historic low and a decrease to only 29.2% of the effective N of 202 stocks achieved in 2014.  Not only does market cap weighting induce substantial single stock risk, but the diversification provided by this foundational asset class has evaporated by 70% over the past decade—hence the concentration crisis.

If weighting by market cap is pushing portfolios to their breaking point, surely weighting companies equally can achieve the diversification for which investors are clamoring?  For all the same reasons the market is so concentrated, the equal weight methodology produces quite radical portfolio constructions, outcomes perhaps even less desirable than the concentration itself—a classic case of the cure being worse than the disease.

Simply put, this is not your grandfather’s equal weight market.  What is often perceived as a simple alternative is no longer a substitute benchmark, but instead an aggressively active strategy.  Specifically, equal weight suffers from significant operational costs, underperformance, questionable assumptions, and skewed risk bets.

As market cap and equal-weighted portfolios have diverged in structure, tracking error has soared to 8.05% on an annualised basis. This value is the highest on record outside periods of market stress, even though volatility is only at the 21st percentile measured on a 20-year range.  To illustrate just how extreme this tracking error is, the 60 largest active mutual funds in the US average 5.50% annualised tracking error.  Yes that’s correct, equal weight is far more active than the leading active funds owing to its onerous reallocation schema.

As a card-carrying active strategy, equal weighting exhibits the familiar encumbrances of high turnover and tepid performance.  The need to countermand all share-price movement at each rebalance means the Russell 1000 Equal Weight Index has averaged 71.0% two-way turnover since 2000.  Moreover, this turnover is historically inconsistent ranging from a low of 44% in 2012 to a high of 132% at the height of the dot.com bubble; this imprecision is a resonating theme of equal weighting.

Notes:  Decomposition of benchmark, equal-weight and multifactor returns around June 30th 2014, the peak of equal weight returns. Source: FTSE Russell Data, June 2024.

Yet it is the performance drag that most indicts the equal-weight framework.  When returns have been so inequitably distributed, owning companies in equal measure has been a perilous approach—the mega-caps did not achieve stratospheric concentration by performing poorly.  

Indeed, equal performance was maximised when the degree of market concentration was minimised.  The halcyon days for equal weighting were a decade ago in 2014, the absolute peak notched on June 30th, and since then the strategy has underperformed relentlessly in nearly every market condition.  

The above figure illustrates this stark bifurcation in performance juxtaposed against changes in top 10 index concentration.  Whereas equal-weight outperformed by 405 basis points annualised from 2005 to mid-2014, it underperformed by nearly identical measure (408 basis points) over the subsequent ten years.  In fact, for every 1-point increase to top 10 index concentration from 2015 levels, the Russell 1000 Equal weight index lost 2.17 points of relative performance to its market-weighted counterpart

Why does this schism in equal weighted returns emerge starting in 2014?  While cap weighting assumes markets are efficient, with asset prices accurately reflecting all information, equal weight takes the opposite approach—it assumes we cannot know anything about the market.  

When concentration rests at manageable levels, this “know nothing” assumption is still very large, but implementable, nonetheless.  However, as the market cap of the largest company expands to 7,658 times the average size of the smallest 10 stocks in the Russell 1000, weighting these companies equally has long since passed credulity.

This size spread between largest and smallest companies is not only emblematic of the concentration dilemma, but indicative of why equal weighting fails in this market regime. In 2005, this size gap was a 224-fold multiple, increasing nine times to a 2,018 multiple by 2015, before expanding a further 3.8 times to present levels.  This scale factor increase of 34 times means a more calibrated means of achieving portfolio breath is necessary—the simple assertion that all companies are the same cannot span the gap.

In periods of hyper-concentration, equal weighting radically departs from market fundamentals, and indeed a return to these fundamental characteristics can foster the more balanced portfolio investors desire.  By targeting independent drivers of historical outperformance, a multifactor model can achieve a more informed diversification along the lines of a structured risk profile.

To illustrate the merits of this approach, the Russell 1000 Comprehensive Factor Index applied a fixed and equal-strength tilt to each of the factors of Value, Quality, Low Vol, Momentum and Small Size.  Redistributing weight according to risk premia as opposed to raw agnosticism succeeds in increasing portfolio effective N to 385, a 554% improvement to market cap diversification.

On the performance front, a complete factor suite not only matches equal-weight’s best years of performance from 2005 to 2014, but it outperforms the latter by a factor of 1.17 over the ensuing 10 years in uncorrelated fashion.  Hence, the multifactor model is able to outperform the benchmark by an annualised 99 basis points over the complete history, compared to equal-weight’s annualised underperformance of 10 basis points. 

In examining the key risk bets of equal-weight and multifactor, the distinctions become clear.  More than performance, expenses or naïve diversification, it is the convoluted and unstable factor exposures that impugn equal-weight strategies.  For instance, while a moderate skew towards value and away from momentum would be expected when holding companies equally, the significant underweights to quality and low volatility may come as an unwelcome surprise—herein lies the underperformance!

In a concentrated market where cap weighting is increasing strained, equal weighting would seem an obvious candidate for a more balanced portfolio.  Yet in neutralising the concentration, it produces a wildly unbalanced series of risk bets to the fundamental drivers of portfolio performance.  In targeting equal exposure to these crucial risk premia, a multifactor methodology can be a restorative balance to US equities when more traditional measures fall short.