When it comes to predicting elections, Moody's Analytics comes about as close to the seers of the crystal ball as we can get. Accurately predicting every election since 1980, with a perfect scorecard in 2012 (including all electoral votes), the Moody's model is pretty hard to beat.
The Moody's model relies heavily on economic data, including real income, real housing, and gasoline prices, as well as political data on a state-by-state basis.
As of August 27, Moody's has the Democrats winning 326 to 212 in the Electoral College. But this year is also the first year with a caveat attached to each updated projection:
"The model results are less valid if either political party nominates a non-establishment candidate. Elections since 1980 have been between candidates who are generally thought to be largely in the mainstream of American politics. Some of the current presidential candidates are more on the fringes of the political spectrum." - Moody's Analytics
For the Republicans to win, they must sweep these battleground states, winning a bare minimum of the three large states to win the Electoral College.
If Sanders or Trump are thrown in, the balance in the "leaning" states is disrupted, possibly putting into play all of them. This would represent the largest collection of battleground states in a modern election if that played out.
One of the biggest factors in the Moody's model is how long the current gas prices will stay low. The lower they remain, the better off the Democratic Party fares.
This is an odd reversal in prediction models, since cheap gas was more of a Republican standard as late as 2012.
The Moody's team has also warned that the recent volatility in world equity markets could impact future updates to their model, as well as the implementation of the now certain Iran nuclear deal.