October 23, 2020
The Quarterly Statement: Q3 2020
While Q2 was marked by a record drop in economic activity with real GDP declining at an annual rate of -31.4%, the recovery in Q3 appeared equally historic. Rebounds in industrial production, retail sales and housing starts have contributed to an estimated GDP growth rate for the quarter greater than 32%. As the effect of stimulus checks and enhanced unemployment benefits begins to fade, the initial V-shaped recovery is showing signs of slowing down however. With continued political gridlock on the next round of fiscal stimulus and the resurgence of COVID-19 cases, a deepened recession may be possible and a return to pre-COVID growth rate levels could take years.
Global stock markets posted positive returns in Q3, despite September being the first month of negative returns for most global indices since March lows. In the U.S., the S&P 500 reclaimed its previous all-time closing high in August, officially ending the shortest bear market in history. While September saw a retreat from recent records, that index still finished the quarter with a strong +8.9%. The most notable performers continued to be the mega cap technology companies, widening the performance of large cap growth versus value. Over the past 12 months, the Russell 1000 Growth has outperformed the Russell 1000 Value by +42.6%. This spread is the largest since inception of the indexes in 1979, far surpassing levels during the peak of the internet bubble. Large cap stocks continued their outperformance of small cap stocks with the Russell 2000 experiencing a total return of +4.9% this quarter. On the international front, developed markets continued to lag their domestic counterparts with a total return of +4.8% this quarter, though emerging market equities benefited from a weaker dollar and returned +9.6%.
In fixed income markets, the Federal Reserve continued to signal accommodative monetary policy and a commitment to low interest rates at the zero-bound through 2023. Chairman Powell also announced a major policy shift to the central bank’s inflation targeting framework, formally agreeing to an average inflation targeting of 2%, which may allow inflation to run hot before pulling the trigger on rate hikes.
Turning to Q4, the presidential election outcome will influence key policy issues such as corporate tax rates, individual capital gains tax rates, and trade policy. Over the long-term, market returns and political results do not appear to have a definitive relationship. Since 1872, there has been no meaningful difference in the performance of U.S. large-cap stocks over full presidential terms, on average. Is this time different? To further assess the investment implications of the election, we provide election themes commentary and investment implications below.
Presidential Election 2020
As the U.S. Presidential Election on November 3, 2020 approaches, the world at large remains transfixed by which candidate will be victorious and the aftereffects that will have on the nation. In this Quarterly Statement, we discuss three themes surrounding the election: the current odds of various outcomes, a timeline of important dates and potential policy change impacts on investments.
Historically speaking, U.S. elections tend to be short-lived events resulting in rather muted market moves. While market uncertainty was initially extremely elevated around this election and the potential “contested election” period, as the polls have shifted to suggesting a less close outcome, the market’s risk around the election has decreased, but still remains above historical averages as measured by S&P 500 Index implied volatility.
The uncertainty surrounding this election as measured by volatility has been elevated for the past several months. Exhibit 1 compares the CBOE Volatility Index (VIX) futures curves as of end Q3 for each of the past several election years. While 2008 has an inverted shape due to the market volatility around the Great Financial Crisis, the other election years of 2004, 2012 and 2016 all have similar shapes that are upward sloping and typical of the VIX futures curve.
The 2020 VIX curve is unique because both the before and after the election implications of the curve indicate elevated volatility. Pre-election, the 2020 “Election Volatility” Bump of 4 volatility points is more extreme than the previous four elections (2004, 2008, 2012, and 2016). Similarly, the 2020 “What’s Next Post-Election Volatility” Bump of 2.5 volatility points is also the most dramatic relative to the previous four elections, suggesting that the market’s perceived risks extend beyond the election. The November VIX futures price is the highest point on the curve as far out as May 2021. This divergence has not been the case in the previous four elections during which VIX futures data was available.
While there are some who suggest that this elevation is due to the increased likelihood of a contested election, other motivating factors may include uncertainty surrounding a fiscal stimulus deal, changes in the economy brought on by potential presidential and congressional party shifts and increased concerns of a possible coronavirus resurgence. The persistent uncertainty reflected in the VIX futures market curve suggests volatility will not dissipate on Election Day.
Odds of Winning
The odds of President Trump’s reelection have been on a downward trajectory over the past several months. As of early October, sites like PredicitIt.org implied a 68% chance of former Vice President Joe Biden winning the 2020 U.S. Presidential Election versus 37% for President Trump, as shown in Exhibit 2. Other sites, such as Nate Silver’s FiveThirtyEight.com, had President Trump’s odds of winning based on polling significantly lower at ~14% versus Biden’s 86%.
Current national polling averages have Mr. Biden commanding a double-digit lead over President Trump (~52% for Biden vs. ~42% for Trump as of early October), as seen in Exhibit 3. If the polls were to remain near current levels and Biden did not prevail, it would be the first time since 1936, when Gallup poll data began, that a candidate with a 52% polling number lost the election . Despite the strong polling results, an “October surprise” is always possible. Additionally, following the 2016 U.S. Presidential Election, polling data has seen increased skepticism, which is likely a contributor to the uncertainty in market volatility.
In a recent election webcast The Glenmede Trust Company’s CEO, Gordon Fowler, hosted Nate Silver, founder and editor of the award-winning website, FiveThirtyEight and Mr. Silver addressed the increased skepticism in polling. Specifically, he noted that the 2016 polling errors favored Republicans because pollsunder-estimated the number of non-college degree voters, which tend to vote Republican. Many pollsters have adjusted this “educational bias” in the data. Even with the adjustments, the polling data for the current election shows notable differences from that in 2016.
While a polling swing of 10% is not unprecedented, as seen in Exhibit 4, there is only one example in the past 84 years where that swing resulted in a surprise result. The 1948 U.S. Presidential Election was the last time that a candidate overcame such a polling deficit. Specifically, incumbent Democratic President Harry Truman trailed by 5% in the pre-election polls to Republican candidate Thomas Dewey (45% for Truman vs. 50% for Dewey)1. The popular vote results were 49.6% for Truman vs. 45.1% for Dewey, or an almost 10% swing from pre-election polling data. Admittedly, polling data was not as efficient in 1948 as it is today,but there is historical precedent for a 10% swing from pre-election polling data. There are also examples of polling errors underestimating the winning margins of a favored candidate: in 1936 then President Franklin Roosevelt was expected to win by 12% and won by 24%, in 1952 Dwight Eisenhower was expected to win by 2% and won by 11%, and in 1980 Ronald Reagan was expected to win by 3% and won by 10%.
Adding to the skepticism around data is that President Trump is polling at 42%, which is the same as 2016. However, FiveThirtyEight calculates Trump’s probability of winning this year is half of what it was in 2016 (14% now as opposed to 28% in 2016). The difference is in the opponent’s polling data. Both Clinton and Trump needed to win undecided voters in order to pass 50%, while this year, since Biden already has 52% of the electorate, Trump would have to get voters to change their decisions, and that’s assumed to be a harder task1.
The 2016 U.S. Presidential Election was seen as a “choice” election between between two nonincumbent candidates. The news cycle dictated much of the polling data that year. The 2020 U.S. Presidential Election has an incumbent president which typically is more influenced by approval ratings. President Trump’s current approval rating versus previous presidential approval ratings suggests a lower probability of re-election, as seen in Exhibit 5.
Sources: Glenmede Investment Management L.P., Wikipedia.org, Gallup
Dots represent each election since 1948 in which the sitting President was running for re-election. Figures along the x-axis show the President’s approval rating via Gallup just prior to the election. Figures along the y-axis show the President’s share of the vote cast for Republican and Democratic Party candidates only. The dotted line is a linear regression trend line based on the data shown. Past performance may not be indicative of future results.
* Gallup did not take presidential approval ratings for Truman, Johnson, Nixon and Ford after June. The final ratings for other presidents are in late October or early November, except for Eisenhower (August) and Carter (September). June data are based on an average of polls conducted during the month.
** Figures for Trump are projected based on a simple linear regression of his latest approval ratings and the historical relationship between approval ratings and the presidential incumbent’s share of the two-party vote.
While the probability of no clear winner on Election Night seems to be decreasing as the potential for Mr. Biden’s vote share increases, the probability is not zero. Nate Silver believes there is a 5% probability that the election is within 0.5%, which could lead to a contested election. This probability, albeit small, is a likely contributor to the increased uncertainty in the post-election volatility noted above1.
Despite the polling results referenced above, increased discussions of a potential contested election this year still abound. Comprehending the timeline of specific events associated with the election may be helpful in navigating post-election uncertainty.
As for the economic recovery and medical breakthroughs, both remain uncertain as they require more time to resolve. While there have been positive developments in medical treatment and a potential vaccine, we have yet to see a major breakthrough or specific timeline. The economic recovery appeared to be starting this quarter — as measured by the jobless rate, credit card data and restaurant reservations, to name a few — but the pandemic’s resurgence continues to create market uncertainty.
In Exhibit 6, the timeline incorporates the important deadlines for the election such as Election Day (November 3), “Safe Harbor” Day (December 8), Electorate Vote Cast Day (December 14) and Inauguration Day (January 20, 2021), all of which bear importance based on dates set by the Constitution or the 1887 Electoral Count Act. An expected increase in mail-in ballots due to the pandemic has contributed to the contentiousness seen this year as it relates to calls of potential voter fraud. While all states require mail-in ballots to be post-marked by no later than Election Day, many provide a post-election day
grace period for the ballots to be received. The timeline in Exhibit 6 includes the states that accept mail-in ballots after Election Day and are leans, tilts or toss-ups according to site 270towin.org. Additionally, the timeline includes the total Electoral Votes for each state and the percentage of mail-in votes in the most recent 2018 election. While the probability of a contested election is expected to be low, knowing the important dates could be helpful in understanding the headlines contributing to market volatility.
Since 1936, the S&P 500 Index has experienced an absolute 1-day close-to-close move of ~140bps following Election Day, roughly double of the average 1-day close-to-close move since 1936. The most significant Election Day moves were in 2008 with a ~5.2% drawdown and the 1948 “polling surprise” election with a 4.6% drawdown. The realized market volatility, as measured by the S&P 500 Index closing prices, tends to be greater the 30 days prior to the election versus the 30-days following, with the most notable outlier the 1948 election when market volatility was markedly greater post-election. That is, uncertainty tends to precede, not follow, an election; which is at odds with what the volatility markets reflect this year.
Exhibit 7 looks at the 4-year returns of the S&P 500 with several different political power combinations. Interestingly, the performances do not differ significantly over the entire 4 years. However, when we look at annualized returns of more recent periods (1956-2018), we identify some divergence between a single partycontrolled congress with a different party president, as seen in Exhibit 8.
Sources: Glenmede Investment Management L.P., Robert Shiller data
Data shown are S&P 500 Index total returns that are inflation-adjusted and include dividend income. The returns are based on monthly data gathered from Standard & Poor’s and Cowles & Associates. Returns shown are four-year averages of real total returns since 1872 to 2016. “Republican Presidents” refers to market performance during all Republican presidencies. “Republican Sweep” refers to the four years after an election where Republicans win the White House and both chambers of Congress. “Republican Divided” refers to Republicans win the White House but do not have control of both chambers of Congress. The same logic applies for the “Democratic” scenarios. The S&P 500 Index is a market capitalization weighted index of U.S. large-cap stocks. One cannot invest directly in an index. Past performance may not be indicative of future results.
The possibility of a “Blue Sweep” resulting in a Democratic White House and Congress, has increased with sites like PredictIt.org suggesting ~60% probability. Mr. Silver noted that the volatility of the outcome in the House of Representatives is the least volatile this election cycle with a 93% probability remaining Democrat. Given Mr. Biden’s polling data and the low probability of a party swap in the House, Mr. Silver believes there is a 67% probability of a “Blue Sweep”1
Exhibit 9 considers various potential impacts of a “Blue Sweep.” Given Mr. Biden’s proposals for green energy and infrastructure, materials, industrials and renewable energy sectors could be positively impacted over the next several years. Healthcare providers may potentially benefit from the expansion of the Affordable Care Act if coverage is expanded versus single payer proposals. Pharmaceutical and biotech companies could experience pressure with the proposed Medicare drug negotiations. In more recent commentary, Mr. Biden has expressed support of antitrust oversight of some large cap tech and Internet businesses, which would likely result in a negative impact for these companies. Additionally, the potential of a more sizable fiscal stimulus package could benefit the consumer discretionary sector. Overall, given the potential for expanded fiscal and monetary policy, inflationary hedges like gold may see increased demand.
Throughout many of the recent elections, investors have undoubtedly had concerns about potential policy changes and broader market performance based on a particular candidate’s campaign promises. Yet over the long-term, market returns and political results have not shown a clear relationship. Since 1872, there has been no meaningful difference in the performance of U.S. large-cap stocks over full presidential terms, on average.
With the odds of the upcoming election leaning towards a Democratic White House, we foresee the likelihood of policy shifts on the horizon that could have sector and tax implications. However, the degree as to how dramatic and at what pace potential changes will occur is uncertain. Given the uncertainty of what campaign rhetoric ultimately influences market returns, we believe maintaining a diversified portfolio with a long-term investment horizon versus a short-term reaction to a political news cycle remains a prudent investment decision.
1 Election webcast with Nate Silver, founder and editor of the award-winning website, FiveThirtyEight. (https://www.glenmede.com/investment-management/our-ideas/2020-election-probabilities-and-predictions-nate-silver)
All data is as of 9/30/2020 unless otherwise noted. Opinions represent those of Glenmede Investment Management, LP (GIM) as of the date of this report and are for general informational purposes only. This document is intended for sophisticated, institutional investors only and is not intended to predict or guarantee the future performance of any individual security, market sector or the markets generally. GIM’s opinions may change at any time without notice to you.
Any opinions, expectations or projections expressed herein are based on information available at the time of publication and may change thereafter, and actual future developments or outcomes (including performance) may differ materially from any opinions, expectations or projections expressed herein due to various risks and uncertainties. Information obtained from third parties, including any source identified herein, is assumed to be reliable, but accuracy cannot be assured. This paper represents the view of its authors as of the date it was produced, and may change without notice. There can be no assurance that the same factors would result in the same decisions being made in the future. In addition, the views are not intended as a recommendation of any security, sector or product. Returns reported represent past performance and are not indicative of future results. Actual performance may be lower or higher than the performance set forth above. For institutional adviser use only, not intended to be shared with retail clients.