That Raucous Iowa Caucus

As his thumbs up seemingly indicate, Martin O'Malley is out. Photo by Steve Pope/Getty Images.

As always, we favor no candidate or party and assess politics solely to analyze how it may impact markets. We believe political bias is blinding and dangerous for investors.

And then there were 12. Former Governors Martin O'Malley (MD) and Mike Huckabee (AR) dropped out of the Presidential race after Monday's Iowa caucus^[i], which saw Democrat Hillary Clinton edge Bernie Sanders in a photo finish, while Ted Cruz trumped^[ii] pollsters and The Donald on the GOP side. Kentucky Senator Rand Paul, who finished fifth on the GOP side, suspended his campaign early Tuesday.^[iii] Pundits are slicing and dicing the caucus results, and if you're into that sort of thing, there is entertainment and wonkery galore. But for investors, there are really only two main takeaways at this juncture: Polls are terrible at predicting primary races, and it's too early to handicap who wins the Republican nomination. That means it's too early to assess how November's vote will impact markets.

Heading into Iowa's contest, polls looked like this:

Exhibit 1: Real Clear Politics Iowa Polling Average, 1/24 - 1/31

Source: Real Clear Politics, as of 2/2/2016.

Here is what the final tally^[iv] looked like:

Exhibit 2: Iowa Caucus Results!

Source: Real Clear Politics, as of 2/2/2016.

As usual, the results hinged on turnouts and the undecided. According to entrance polls, Trump won among voters who'd made up their mind a month ago or more, with 39% (compared to 26% for Cruz and 13% for Rubio). Cruz won among those who decided a week ago. Those who decided in the final days gravitated overwhelmingly toward Rubio. (For those who like the nitty gritty, FiveThirtyEight's Nate Silver has some detailed analysis on that poll.) Perhaps this is one indication outsider candidates will eventually fade, leaving a more traditional race between two establishment figures, but it is far too early to say.

On the Democratic side, the results actually aren't terribly out of whack with the polls. Of the eight polls in Real Clear Politics' average, two put Sanders in the lead. Of those putting Clinton ahead, most showed Sanders within the margin of error, which was fairly wide. The margin of error often gets short shrift in media coverage of polls, which tends to focus on the headline results, but it does matter. A three-point lead isn't necessarily a three-point lead.

Keep all of that in mind as you consider polls for next week's New Hampshire primary, which presently look like this:

Exhibit 3: Real Clear Politics New Hampshire Polling Average, 1/25 - 2/1

Source: Real Clear Politics, as of 2/2/2016.

Yes, Trump and Sanders have giant polling leads, and it would take a massive wave of late-deciders to upend them. Of more interest is that dead heat just under Trump. Voters could easily swing to any of those four-or, for that matter, Chris Christie. This ain't over by a mile, and it is impossible to know today which way the pendulum will go.

New Hampshire's results will matter, but not for the reason you've probably heard. The old adage that "Iowa picks corn, New Hampshire picks Presidents" made the rounds Tuesday morning, as it does every four years. But as the table at the end of this article shows, neither race is terribly predictive. In contests where there was no incumbent, Iowa picked the winner 6 of 13 times since 1976. Less than half! New Hampshire was only marginally better, picking the eventual nominee eight times, roughly a 60/40 right-to-wrong split. But New Hampshire often does narrow the field, just as Iowa did somewhat Monday night, and that narrowing matters a great deal.

One reason this Presidential race has so many abnormalities is because it has a boatload of candidates. At the peak, there were 22-5 Democrats and an astounding 17 Republicans. Now, after the latest exits, we have 2 Democrats and 10 Republicans. When 10 people split the vote, oddities will inevitably emerge. Donald Trump looks like the front-runner because he has about 36% of polled voters support in the national field. But 36% is well short of a majority. It just looks big when nearly a dozen people split the 64% of polled voters who didn't say "Trump."

As more candidates exit the race, their supporters will rally behind other contenders, ultimately shaping the race. For many, it's hard to imagine their supporters flocking to Trump. Those who back libertarian-leaning Rand Paul, for instance, probably won't vote for someone who supports eminent domain. Beyond that, though, questions abound. Will Huckabee's devotees prefer a Senator with a similar platform or a fellow former Governor with similar experience? What about Rick Santorum's small band of followers? Who was choice B for the 12 Iowans who wrote down Jim Gilmore's name? There is no way to handicap any of this. It's very much a wait-and-see game still. We know the race will morph, but how is anyone's guess.

If you're a political junkie, that means the fun is just beginning. If you don't love the theatrics and would prefer to get on with considering who will win in November and how markets might respond, sorry, but it's too early. There is still way too much noise. Markets move on probabilities, not possibilities, and there is just no scientifically valid way to assign probabilities at this point. Your geeky MarketMinder staff have been watching a new prediction market like hawks, and we've noted Marco Rubio's stock soared after Iowa, but as an indicator, it's unproven. We've kept an eye on the UK futures sites as well, but some big foibles last year^[v] make us skeptical.

All that said, we can highlight some broad themes, albeit with a big caveat. Since 1926, the S&P 500 has an above-average frequency of positive returns in election years. There is also the phenomenon we call the perverse inverse-markets' tendency to rally in election years when Republicans win but deliver below-average returns the next year, and disappoint in years when Democrats win but rally in year one.^[vi] Combine those factors with gridlock's inherent bullishness, and US politics do point positively for stocks this year. But this election's abundant wild cards do add a wrinkle, and if any of them-from a Trump nomination to a Clinton indictment after she has the nod locked up-could upend traditional analysis.

If 1968's wild contest is any guide, we probably do get that traditional race in the end, but we can't ignore the alternate possibilities. Again, wait and see.

Exhibit 4: Historical IA and NH Winners

Source: The Des Moines Register, New Hampshire Secretary of State, The New York Times and CNN as of 1/30/2016. Red/blue cells indicate evenutal nominee.