Ahead of today's presidential and congressional elections, Goldman provides some brief thoughts on various election-night (and beyond) events. From a viewer's guide to the poll-closing times to a discussion of the apparent 'closeness' of the race and post-election market performance, they note that equity performance post 'tight' races has been better than in elections where the winner is more clear-cut. This election has a twist though in that it will be immediately followed by debate on the fiscal cliff, and thus resolution of the election will reduce, but not eliminate policy uncertainty.
Via Goldman Sachs: Q&A On Election Night
- Given that pre-election polling indicates that, at least at a national level, this election is closer than most others over the last 50 years, there is a clear risk that the results may not be known by the end of November 6.
- Equity markets in presidential election years do not follow a consistent pattern like markets in midterm election years. However, median equity performance in the weeks following very competitive elections slightly outpaces performance following less competitive contests. That said, this election is unusual in that it will be immediately followed by debate on the fiscal cliff, and thus resolution of the election will reduce, but not eliminate policy uncertainty.
Q: How does this election compare to other close elections?
A: At the presidential level, it is closer than most contests over the last 60 years. Exhibit 1 lists several previous elections in which public opinion polling on the eve of the election suggested a very close race. Even compared to those, polling at the national level suggests a closer race than most in recent history.
Exhibit 1: Pre-Election Polling in Competitive Presidential Elections
Source: Gallup. Realclearpolitics.com. GS Global ECS Research.
Q: Is there a risk that the election result will be delayed past November 6?
A: There are two foreseeable risks: a recount in a key state, or a delayed result due to absentee ballots or other voting issues. Colorado, Florida, and Pennsylvania conduct automatic recounts if the results are within 0.5 percentage points; Ohio's threshold is 0.25%. The average of recent polls in Colorado shows the President with a 0.6pp lead, just outside this threshold. In Virginia, President Obama has traded leads with Gov. Romney over the last couple of weeks; the President currently leads the average of recent polls in the state by 0.3pp (in Virginia, like other states that do not have automatic recount rules, candidates can still request a recount in a close election). While it is not inconceivable that at least one state's results could be subject to recount, it should be noted that even in the event that this occurs, it would take an incredibly close election to trigger a recount process similar to what happened in Florida in the 2000 election; President Bush's final margin was only 0.009% of the combined vote of the top two candidates.
Some states may also have difficulty dealing quickly with absentee and provisional ballots. Election laws in Ohio require that provisional ballots--votes cast by those whose eligibility to vote is in question, or who requested but did not return an absentee ballot--should not be counted until 10 days after election day, when absentee ballots must have been received, to avoid potential duplicate votes. As of November 2, there were around 200,000 outstanding absentee ballots, or roughly 3.5% of the number of votes cast for the two major party candidates in 2008.
Q: Do markets behave differently following close elections?
A: Equities react slightly more positively after competitive elections. Market trends are much less clear-cut around presidential elections than they are around midterm elections, when equities have fairly consistently gained following the election and into the early part of the third year in the presidential cycle.
By the end of the week following the election (i.e., November 16 this year) the median gain in the S&P 500 has been 1.9%, compared with 0.3% following less competitive races. The reaction in Treasuries is inconsistent; the yield on the 10-year Treasury has risen following competitive elections and declined slightly following less competitive contests, but is characterized by large moves in either direction in a few election years, and very small moves in most others.
While there are undoubtedly parallels with some of these prior periods, as we have pointed out recently, there are also major differences: the fiscal cliff looming just after the election at year end, the greater fiscal imbalance that lawmakers are likely to address next year, and the increased political focus on monetary policy ahead of the end of Chairman Bernanke's current term in January 2014.
Exhibit 2: Equity and Interest Rate Reaction Following Competitive Presidential Elections
Source: Federal Reserve. Standard and Poors.
Q: What will the early indicators be on election night?
A: Indiana for the Senate, Ohio for the White House, and Virginia for both. Exhibit 3 lists each state by poll closing time, along with the number of electoral votes at stake and the party leading (and margin of lead) in the average of recent polling in the presidential and Senate races. Taking the poll averages at face value, current poll results imply that the President would win 303 electoral votes and Gov. Romney would win 235. In the Senate, taking the current polls at face value implies that some seats will change hands but that the net result will be 23 Democrats elected and 10 Republicans, leaving the partisan composition of the Senate unchanged (there are 23 Democrats and Democratic-caucusing Independent seats and 10 Republican seats up for election this year).
In the race for the White House, Virginia will be the first closely contested state to close its polls, at 7:00 pm ET. The state is essentially tied in recent polling, implying that if the President wins here, Gov. Romney would need to win in either Ohio or in a combination of other states (e.g., Colorado, Iowa, Nevada, New Hampshire and/or Wisconsin) where the President's polling advantage has recently been stronger than in Virginia. Ohio closes its polls half an hour later, at 7:30 pm, and while other electoral combinations are possible, there is a good chance that the winner in Ohio will also win the election.
Regarding the Senate, the earliest indicators will be in Indiana and Virginia. Democrats, and independents who caucus with them, currently hold 53 seats, while Republicans hold 47. If the Indiana seat, currently held by a Republican, is won by the Democratic candidate Rep. Joe Donnelly over Republican candidate Richard Mourdock, it will be an early indication that Republicans will have a difficult time gaining the majority in the Senate. Likewise, if Republican former Senator and Governor George Allen is able to win against Democratic former Governor Tim Kaine, it would indicate that Republican gains might be larger than generally expected.