US Elections: Divergence in the poll analysis - BBH

Research Team at BBH, suggests that there is a divergence in the poll analysis in the US that has become pronounced since the FBI's announcement on October 27 about the new development about Clinton's emails. 

Key Quotes

“Some, including Fivethirtyeight.com, and RealClearPolitics, see the race as having tightened considerably.  The former has Clinton as favored with a 65.5% chance of winning, after peaking near 88% a couple of days before the third debate.  The latter has Clinton ahead by 1.8 percentage points in a survey of national polls.  The Financial Times caught this spirit in the headline of its weekend edition: "US election knife-edge as rivals weigh early-voting data."  While late in the game, we suspect these outlets will adjust in light of the weekend FBI news.

Their calculations of how the Electoral College system breaks down has Trump less than 30 Electoral College votes shy.  That means that one large state, like Florida (which fivethirtyeight.com has leaned toward Trump and RealClearPolitics have favoring Clinton by 1.2 percentage points) can give deliver Trump a victory.

There is another set of respected poll work that sees it quite differently.  These include the New York Times Upshot, Huffington Pollster, and the Princeton Election Consortium.  They, like the peso and Fed funds futures, have been unflappable.  Upshot puts the odds at 84% Clinton wins.  She peaked at 93%.  Huffington puts it at 98.4, and Princeton puts the odds at 99%. 

Despite the divergence of the analysis, none show Trump as the favorite, and most show some stabilization in the last couple of days.  We are inclined to see the recent market moves as a fundamentally-spurred technical correction.  We suspect that that correction is over or nearly so.  When the US political uncertainty is lifted, the economic divergence can reassert itself as the main driver.  Although the job growth reported before the weekend was nothing to get excited about, the uptick in hourly earnings growth to its fastest pace since 2009 (2.8%) is a wholly desirable development.  It makes it easier for the Fed to remove some accommodation, but a 25 bp rate hike to 50-75 bp is hardly a tightening of policy. 

All bets are off if Trump wins.  This is not said out of partisanship, but a function of induction (how investors have responded to such prospects) and deduction (investors don't like the unknown or surprises; a Trump victory is both on steroids).  Given the seemingly limited path Trump has to the White House, two states on the East Coast (Florida and North Carolina) will likely be indicative of the election outcome.  Trump must win both or defeat is all but certain. These may be known in the Asian morning on November 9.  Trump must go on to win nearly every other swing state, including Pennsylvania.”