About three weeks ago, I made an election bet on Trump to win the popular vote. Here is my post-2016 post on the prior big election bet. Last week I YOLO'd the account on the bet. This is not a huge sum of money, but it's a high convinction bet and has me thinking a lot about the implications.
My default assumption is that a Trump landslide (it will be a huge landslide if he wins the popular vote) will turbocharge the stock market rally and set off a surge in the U.S. dollar. My view is the "smart money" isn't betting on a landslide. If they shift that way, we might see this rally unfold into the election. Normally, there is a dip ahead of the election that will bottom before the election actually takes place.
Why a landslide? From 5 days ago, look at the Biden and Clinton leads on this date versus Harris: Biden +9.7, Clinton +4.6, Harris +2.0 percent. Trump is also running about 4 percentage points ahead of Harris in almost all the swing states, such that he leads in most of them. In 2016 and 2020, Trump trailed in the swing states one month out. This, if history plays out as it did the past two cycles, and Trump further gains, he will win easily. I project 312 Electoral College votes. You can see my map at RealClearPolling. This is based on Trump adding a bit to his swing state totals. He actually doesn't need to gain as much as in 2016 or 2020, so if pollsters are doing a better job this time around, I think he wins.
A note on polling, which I made back in 2016. The polls have never been that bad as a whole. What happened in 2016 was "experts" interpreted poll numbers and drew very incorrect conclusions.
To wit, Republicans almost always rally late in elections because for whatever reason, more people who lean Republican will say they're independent than vice versa. That's the latest number pointing to a big Trump win that I saw yesterday. Pollsters can only go by what people tell them, and if people are shy-Republicans, the polls will always skew Democrat. Historically, Democrats have easy wins when running +8 in voter identificaiton polls. Republicans win when it's only Democrat +3. To repeat, Republicans win when the pollsters report Democrats are +3 in party identification. The current polls have Republicans at +1.
Finally, an anecdote. There are people called "low propensity voters" who don't hardly or ever show up to vote. These people do not care most of the time or all of the time. But sometimes they get motivated to vote. In Pennsylvania as of last week, there were already more than 300 votes from people who hadn't voted since 2004. It's a small number, but statistically significant. My hunch is people who haven't voted in 20 years are getting off their butts to vote Trump. Additionally, early voting by Democrats is down around the country from 2020. My hunch is the very high turnout number of 66 percent that year will head back towards the more normal level of around 60 percent.
In conclusion, all polling data points to a large Trump victory. It looks large enough that a "steal," if you believe that is possible, will have to be at a scale so large as to be blatantly obvious. Personally, I don't see anything like what happened in 2020 repeating. Going through the swing state vote data, the only state that seemed to buck the Trump rally trend was Georgia, which has been at the epicenter of claims of election fraud. As for polling, if pollsters have completely adjusted to 2016 and 2020 and are overstating Trump's support in response, then of course my interpretation will be proved wrong. If pollsters have improved a bit, but Trump still gains 1 to 2 percentage points instead of the usual 4 to 6 percentage points, then he might eke out a popular vote victory while winning an electoral college landslide. If he actually surged as much as he did in 2016 and 2020, he should win the popular vote by an easy margin.








































