The ‘Short’ Story on Short Interest

In last week's edition of Factor Spotlight, we introduced our new partnership with S3 Partners along with the availability of S3's market-leading Short Interest, Crowding, and Securities Finance datasets on the Omega Point Platform. This week, we leverage this highly actionable dataset to glean more insight into the short interest landscape and highlight methods to help improve market analysis, idea generation, and portfolio construction workflows.

Much attention has been paid, and rightfully so, to the retail squeeze events of late driven heavily by “meme stocks.” Still, it is vital to take a step back and understand the broader picture of short interest in the global markets. As investors research alpha ideas, evaluate potential hedges, and gauge market sentiment, datasets that better explain short interest, crowding, and financing are becoming indispensable in today’s environment.

Recent Market Trends

For example, when we isolate S3's Utilization, which measures the ratio of a stock's shares on loan relative to the total shares available for lending, we see a dramatic increase in this metric over time across all global equities.

The chart below identifies the 99th, 95th, 90th, and 80th percentiles of the S3 Utilization levels across the MSCI ACWI (proxied using the iShares ACWI ETF) since October 2020. The 99th percentile stock in the index had utilization of 49.5% on October 1st, meaning slightly less than half of all shares available for lending were borrowed. In February 2021, that shot up to almost 73%, and by June 1st, it perched at 94.3%!

The impact of this dramatic utilization increase resonates far beyond the distribution tail. Over the past eight months, each percentile has nearly quadrupled across this period in terms of utilization, with the effect of squeezing availability and making names increasingly difficult to borrow.

The Volatility Behind Short Interest

The volatility inherent in these high short interest stocks has been a cause for concern among investors. As noted in last week’s Factor Spotlight, these names tend to be highly idiosyncratic in nature. To explore the recent behavior of these crowded stocks, we returned to the ACWI to construct three global portfolios:

1. S3 Top SI % Float - Equal-weighted portfolio of stocks with S3’s Short Interest as a % of Float greater than 10%
2. S3 Top Utilization - Equal-weighted portfolio of stocks with S3’s Utilization greater than 70%
3. S3 Top Finance Rates - Equal-weighted portfolio of stocks with S3’s Last Rate greater than 5%

We rebalanced each portfolio weekly to pick up on intra-month changes in their respective measures and screened out stocks with market caps < $500m & ADV < $5m.

When we look at the alpha performance in the Barra Global Equity Long Term model relative to the broader index, we see that these portfolios exhibit a great deal of idiosyncratic volatility throughout the entirety of the eight-month timeframe.

In particular, the Short Interest as a % of Float and Utilization portfolios saw dramatic spikes in Nov 2020 and Jan 2021 during the retail crowding short squeeze events. The stocks with high utilization levels had relative alpha outperformance of nearly 11% compared to the ACWI over the last week and a half of January.

Further Applications of Short Interest Data

Investors can leverage this dataset in a variety of different ways, including:

1. Existing Portfolio Analysis: How susceptible are the names in my portfolio to increased pressures from short interest crowding?
2. Idea Generation and Hedge Construction: Screen out hard-to-borrow stocks when identifying single name shorts or constructing a hedge basket to reduce the unwanted, idiosyncratic volatility and avoid squeeze vulnerability
3. Market Analysis: What segments of the market are becoming more crowded? Are there particular fundamental characteristics, sectors, industries, or names towards which the market is gravitating?

To help peel back a layer from #3, we can look at key active style exposures of our S3 ACWI portfolios by leveraging the Barra Global Equity Long Term model. This will help us understand the characteristics of the top-listed stocks through the lenses of Short Interest as % of Float, Utilization, and Finance Rates.

The following table shows the portfolio's active exposures as of June 1st, 2021.

Using the table above, we can infer that crowded stocks are underexposed to Investment Quality and overexposed to Earnings Variability. These are names with variability in historical and forecasted sales, earnings, and cash-flows that also demonstrate a low growth rate of Total Assets. These stocks also have a high degree of Residual Volatility relative to the broader ACWI index and are low Momentum, meaning they’ve underperformed over the trailing twelve months.

Interestingly, the growth/value tilt represented by the Growth and Book-to-Price factors also stands out. On aggregate, investors are crowding growth names over value names from a shorting perspective.

As the analysis shows above, crowding in institutional short books has only become more prevalent over time and will likely continue to drive idiosyncratic volatility relative to the broad market. Combining short interest data alongside factor models can help provide a lens into the attributes of heavily shorted stocks so that investors can better prepare for the market shifts that tend to come hand in hand with crowding events.

US & Global Market Summary

US Market: 06/21/21 - 06/25/21

• The US market bounced back after last week, with both the S&P 500 and Nasdaq hitting new record highs. The S&P 500 posted its best weekly performance since February as investors continued to shrug off inflation fears.
• Numerous comments out of the Fed, including from Jerome Powell, helped ease investor qualms about rate tightening in the near term.
• Additional optimism abounded around the growth implications of the ~$1T bipartisan infrastructure deal that President Biden agreed to on Thursday, although it is nowhere close to a guarantee.
• The PCE prices index increased by +0.4% in May - the third consecutive increase. And over the past year, consumer prices have climbed by 3.9%, reflecting the biggest gain since 2008.
• Consumer spending in May was flat, representing a significant slowdown after gaining +0.9% in April and seeing +5% surge in March.
• The banks enjoyed a boost on Friday after the Fed announced that the biggest US lenders could easily withstand a severe recession, based on the results of the Dodd-Frank Act Stress Test.

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

• Market Sensitivity was again the biggest winner, climbing higher into Overbought territory after a +0.46 standard deviation move.
• Growth rose by +0.24 standard deviations and crossed into Overbought space, now sitting +1.08 SD above the mean.
• Profitability seems to be bouncing off of a trough of -1.78 SD below the mean on 6/21, ending the week at -1.55 SD below the mean.
• Volatility slowly climbed higher into Extremely Overbought space, now at +2.4 SD above the mean.
• Earnings Yield appears to have reached some resistance after free falling into Oversold territory, as it now sits at -1.85 SD below the mean.
• Value plunged into Extremely Oversold territory, now at -2.14 SD below the mean.
• Momentum was the week’s biggest loser again, falling by -0.32 standard deviations as it descended into negative normalized space.
• US Total Risk (using the Russell 3000 as proxy) decreased by 18 basis points.

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

• Global Growth shot up by +0.28 standard deviations, and now sits close to its historical mean.
• Earnings Yield rose +0.22 SD higher after bouncing off a recent bottom of -1.71 SD below the mean on 6/4. It remains an Oversold factor at -1.24 SD below the mean.
• Volatility continued to climb higher, albeit at a slower pace, and is an Extremely Overbought factor at +2.19 SD above the mean.
• Exchange Rate Sensitivity continued to decline away from Overbought space and now sits at +0.63 SD above the mean.
• Profitability continued to plummet, falling another -0.24 standard deviations deeper into Extremely Oversold territory at -2.36 SD below the mean.
• Value was the biggest loser worldwide, falling by -0.33 standard deviations and earning an Oversold designation at -1.21 SD below the mean.
• Global Total Risk (using the ACWI as proxy) decreased by 18 basis points.

Regards,
Kevin