Where is Diversification Hiding?

Alpha and diversification have been especially elusive since the start of the pandemic in early 2020. As we’ve discussed previously, common systematic factors grab a tighter hold of stocks during periods of market stress, and assets begin to get pushed in a much more correlated fashion. So, to supplement our research on the return of alpha opportunity in the market, we will explore diversification this week.

Diversification can take on a lot of different definitions. For example, in the world of equity portfolio management, diversification can revolve around sector, market cap, or style allocations, the number of names in a portfolio, the degree of concentration in the top X holdings, and more. In the context of risk, a helpful way to define and quantify diversification is through the Diversification Ratio.

Diversification Ratio

We can calculate the Diversification Ratio in several ways. For the analysis below, we divided the average predicted annualized volatility of a portfolio of stocks by the predicted annualized volatility of the equal-weighted portfolio.

The premise is simple: the higher the ratio, the greater the degree of diversification across a portfolio. If a portfolio’s stocks have very similar characteristics and tend to move in a highly correlated way, we would expect the portfolio’s total risk to be lower but not much lower than the total average risk of each stock. On the other hand, if a portfolio is invested in stocks across different sectors with varying sizes, fundamental characteristics, and more, that portfolio's total risk will likely be much lower than the average of its parts.

Where is the Variety in the US?

Now that we have a definition for diversification, let's look into what's been going on with diversification most recently. The chart below shows the average security-level predicted volatility (Total Risk), portfolio predicted volatility (Total Risk), and the Diversification Ratio of the Russell 1000 Index (Equal-Weighted) using the Wolfe QES US Broad Risk Model.

At the beginning of 2020, the Diversification Ratio was above 2, meaning the combined risk of the Index (~13.5%) was less than half of the average security risk (~28%). However, that changed in the blink of an eye in March 2020 when the equal-weighted index risk rose to ~42% and the average security risk rose to ~63%, cutting the Diversification Ratio down to 1.47.

The results here aren’t surprising. Under intense market stress, assets that might typically be uncorrelated often start moving in unison. The natural hedging in a diversified portfolio or Index gets lost when significant macroeconomic events take over. As a point of reference, we ran the same analysis during the Global Financial Crisis in 2008-2009. Although the disintegration of diversification happened a bit more gradually, the trend looks remarkably similar. It's worth noting that after the Global Financial Crisis, the Diversification Ratio didn't reach a level of 2 until mid-2011, just in time for the Debt Ceiling Crisis!

Intra-Sector Diversification

It's clear that there is still lower diversification in the large-cap US market than we saw pre-pandemic, but what can we expect from different market segments? We repeated the same Diversification Ratio analysis using sector groups along with Wolfe’s sector-specific risk models listed below.

We notice that before March 2020, Consumer Discretionary and Consumer Staples show by far the highest degree of diversification. Energy exhibited the least diversification, but we should note that the Energy sector is significantly smaller than the other sector groups in the Russell 1000. Having fewer names makes diversification a bit more challenging.

Taking a closer look at Consumers, the chart below highlights the distribution of Total Risk % across the Russell 1000 names in Consumer Discretionary and Consumer Staples. The red dots indicate the Total Risk % of the equal-weighted Index. Before March 2020, not only did most Consumer stocks sit within much tighter bounds of Total Risk (~20% to ~45%), but the Total Risk of the equal-weighted Index was significantly lower than that of its least volatile stock. When the Diversification Ratio drops, however, the Total Risk of the Index joins its underlying stocks. The stocks themselves also varied much more widely in terms of individual volatilities.

How can we apply the Diversification Ratio beyond using the measure as a gauge or point of understanding? First, the ratio can help managers determine market segments where they might be able to find diversification. Second, it can help quantify the effectiveness of a broad hedge (i.e., a high Diversification Ratio within a hedge basket or Index might indicate that the characteristics within the universe vary too widely and need refining). Finally, managers can also use the Diversification Ratio to quantify the relative payoff between exposure to individual stocks or a broader ETF.

US & Global Market Summary

US Market: 09/07/21 - 09/10/21

• U.S. equities notched their biggest weekly decline since mid-June as all major indices retreated during a volatile, shortened trading week.
• Investors assessed the latest signals on the economy after more strategists weighed in with cautious comments on the market. The strategists’ common themes included historical extreme valuations, a near non-stop 7-month rally, an economy that appears soft and the imminent tapering of Fed stimulus.
• On Thursday afternoon, President Biden announced that all large employers must require workers to either be vaccinated or submit to weekly testing, while vaccination would be mandatory for federal workers and contractors.
• The Labor Department reported that producer prices rose 0.7% in August, above consensus expectations.
• Gold posted its first weekly decline since early August after data on U.S. producer prices sparked debate over when the Federal Reserve will pull back on stimulus.
• Treasuries were lower, putting upward pressure on yields, and the U.S. dollar nudged upward, while gold slid, and crude oil prices rose.
• The Biden administration is weighing a new investigation into Chinese subsidies after the U.S. president urged China’s Xi Jinping to cooperate on a phone call.

‍Normalized Factor Returns: Axioma US Equity Risk Model (AXUS4-MH)

• Volatility’s stout showing last week in the wake of its extended slide was no fluke as it vaults to the top spot on this week’s leaderboard and now sits on the cusp of positive territory.
• Value slammed on the breaks after an entire month atop the US factor leaderboard, but still ends the week on a slight positive note as it rose +0.02 standard deviations.
• Market Sensitivity (Beta) shows spark for another week and creeps upward to sit at -0.33 SD below the mean.
• Momentum’s extraordinary 8-week upward run is finally in the history books as it slides -.08 standard deviations.
• Profitability plunged for the 2^nd straight week and now sits only 0.03 standard deviations from negative territory.
• Size fell for the 6th straight week and moves more firmly into negative waters.
• Growth’s 3-week biggest loser reign ended with the new ignominious crown now borne by Earnings Yield.
• US Total Risk (using the Russell 3000 as proxy) decreased by 10 basis points.

‍Normalized Factor Returns: Axioma Worldwide Equity Risk Model (AXWW4-MH)

• Volatility rockets skyward for the 3rd straight week and is now perched just 0.01 SD below the mean.
• Exchange Rate Sensitivity moves upward for the 2nd consecutive week as it gets closer to positive territory following 2+ months of weakness.
• Value’s reign atop the leaderboard came to an end but still gained +0.16 standard deviations to round out this week’s top 3.
• Momentum continues to show weakness as it moves closer to the negative column and ends the week sitting at +0.20 SD above the mean.
• Profitability did indeed plunge as we anticipated last week, emphatically erasing its recent extended tear.
• Earnings Yield ends as the week’s biggest loser once again and passes into negative territory to sit at -0.16 SD below the mean.
• Global Total Risk (using the ACWI as proxy) decreased by 9 basis points.

Regards,
Kevin