Even casual Fed watchers must by now have taken note of the Fed’s announcement that it planned to begin tapering its monthly bond purchases around mid-November. Now, here we are.
So far, the equity markets seem to be taking the news in stride – maybe because the current slate of Fed officials have been more transparent and communicated better than some of their predecessors. However, with headline inflation now topping 6%, the highest in 30 years, the U.S. economy may be running hotter than Fed officials anticipated. And inflation may be looking less transitory than they hoped. For example, recent wage hikes will likely prove difficult to roll back, once entrenched.
If inflation pressures the Fed to hit the brakes harder than the market expects, what happens if we experience another “Taper Tantrum?” That term was coined after the Fed began talking about withdrawing stimulus after the third round of Quantitative Easing (QE3) in 2013, causing a sharp market correction.
We’ve subsequently experienced two other ‘mini’ Taper Tantrums. The three time periods we’ll focus on were the sell-offs that started in May 2013 when Ben Bernanke commented about ending monetary easing; in January 2016 after the Fed hiked interest rates just 0.25bps; and under Fed Chairman Jerome Powell in October 2018. The sharp market correction during that third Taper Tantrum essentially pressured the Fed to capitulate and discontinue winding down its balance sheet, contributing to a quick partial market rebound.
The Refinitiv StarMine quant research team has been tracking the live monthly performance of its quant models since each commercial release. Here, we drew upon that archive of old performance reports to see which models, or factors, worked best during these historical Taper Tantrum periods.
Exhibit 1: May through June 2013 Taper Tantrum
Source: Refinitiv Datastream
Exhibit 2: “What Worked” from the June 2013 StarMine Monthly Performance Report
Source: StarMine Monthly Performance Report
The charts above stack-rank quantitative model and factor performance based on their Information Coefficients (I.C.s) for the month. That measure, often favored by “quants,” is a rank-correlation of a stock’s model rank vs. its price performance. Bars to the right of the vertical line at 0.0, with positive I.C.s, added value while those to the left with negative I.C.s show factors with negative correlation to price changes. Four major market regions are shown here: U.S., Developed Europe, Developed Asia ex-Japan and Japan.
During the month of June 2013, the StarMine models and other factors that performed best in the U.S. were the StarMine Analyst Revisions Model (ARM), Long-Term Growth, StarMine Smart Holdings (which examines the characteristics currently in favor with institutional investors, based on their recent purchasing behavior), and StarMine Earnings Quality (EQ).
In Developed Europe, Smart Holdings, Earnings Quality and ARM again were among the top four performing factors, joined by StarMine Price Momentum (PriceMo).
Turning to Developed Asia ex-Japan, we find Dividend Yield, and again, PriceMo, ARM and Smart Holdings.
Even in Japan, where factor performance frequently deviates from the rest of Developed Asia (which is why we break it out as its own region) EQ, ARM and Smart Holdings were also among the top four performers, along with Market Capitalization.
Exhibit 3: January 2016 ‘Mini’ Taper Tantrum
Source: Refinitiv Datastream
Exhibit 4: “What Worked” from the January 2016 StarMine Monthly Performance Report
Source: StarMine Monthly Performance Report
During January 2016, the factors that worked best in the U.S. were the StarMine Smart Holdings and Short Interest models, Dividend Yield and Trailing 12-Month (T12M) Returns. Like the 2013 version, this ‘mini’ Taper Tantrum quickly turned into a global phenomenon with markets selling off across regions. Here again we find Smart Holdings, ARM and EQ all showing robust performance across all regions, except for EQ in Japan.
Exhibit 5: October through December 2018 ‘Mini’ Taper Tantrum
Source: Refinitiv Datastream
Exhibit 6: “What Worked” from the December 2018 StarMine Monthly Performance Report
Source: StarMine Monthly Performance Report
During the month of December 2018, when the market was in sharp decline, the factors that performed best in the U.S. were the StarMine Short Interest, PriceMo and EQ models, along with T12M Returns.
During this Taper Tantrum, momentum dominated in Developed Europe, although Smart Holdings ARM and EQ added incremental value. In Developed Asia ex-Japan, PriceMo, T12M Returns, ARM and Dividend Yield were the top performers. PriceMo, Dividend Yield, Market Capitalization, and T12M Returns were the top four factors in Japan, but Smart Holdings added value across both Asia-Pacific regions.
Conclusion
Across regions and years, market participants favored, and rewarded, high quality companies, ignored by the short sellers, and experiencing positive sell-side analyst estimate revisions. Also rewarded were companies with similar fundamental characteristics as those recently added to the portfolios of institutional investors in aggregate (as disclosed in their most recent SEC Form 13-F filings).
Part of this seems intuitive; the trader’s adage “flight to quality” is one behavior our StarMine Quant Research team validated during other market sell-offs at the beginning of historical bear markets, which we published in a 2018 research note: https://www.refinitiv.com/perspectives/future-of-investing-trading/survive-market-downturn/
What did not “work” here were valuation factors. Value stocks underperformed momentum stocks and factors during each of the three Taper Tantrum periods examined.
History doesn’t always repeat, but the consistency of the outperformance of these factors might be worth bearing in mind if the market again gets the jitters as the Fed begins winding down its bond purchasing program.
