"Kee" Points with Jim Kee, PhD.

The 2016 US Presidential election was a true surprise. Even PollyVote, which correctly predicted that Clinton would win the popular vote (who didn’t?) missed the Electoral College outcome, and that is instructive. Since the election, the media have been asking, “How can the polls have gotten it so wrong?” PollyVote was created partly because of polling’s short-comings. Rather than relying on one method’s accuracy (like polling), Pollyvote combines the information from multiple and independent methods, like econometric (statistical) models that incorporate demographics and geography, and so-called prediction markets that utilize betting outcomes. In all, Pollyvote combines over 20 independent sources, and not a single one predicted Trump to win a majority of electoral votes. A few of PollyVote’s sub-component models, like Alan Abramowitz’s time-for-change model (looks at the state of the economy, incumbent popularity, etc.) and Helmut Nortoth’s electoral-cycle model, did predict Trump to win the popular vote, but not a majority of electoral votes. Interestingly, prediction markets (Iowa Electronic Markets), which have been the “new thing” in election forecasting and typically have been the most accurate, produced the worst forecasts in this election, while econometric models, which had been the least accurate over the past six elections, did the best. That’s why combining forecasts from different methods generally leads to better predictions over time than relying on any single method. I think the 0 for 20 outcome of all of the different models reflects just how far out in left field this outcome was.


As for the economic and investment implications, stronger growth should result if Trump sticks to relatively straightforward tax/regulatory simplification and reform, and softens the trade stance relative to the campaign rhetoric. I expect business investment spending to be a much stronger driver going forward than it has been. On the demand (for investment funds) side, that’s largely because Obama was viewed as more anti-business than either Trump or Clinton, and that tends to keep cash on the sidelines. On the supply (of investment funds) side, some regulatory clarity and simplification on Dodd-Frank banking legislation is likely going forward. It is fairly widely acknowledged that the evolution of this particular legislation has thwarted credit creation, or the banking system’s ability to serve as a financial intermediary matching capital with opportunity. In fact, both candidates expressed the need for reform or fixes to major legislation like Dodd Frank and the Affordable Care Act.


As for the Federal Reserve, I would say that lately Fed Chair Janet Yellen has had to face somewhat of a conundrum with regard to raising rates. If the fed acts to push short-rates up, while global capital flows act to keep longer-term rates down (by bidding up the price of longer-term bonds and lowering their yield), then the result is a flattening yield curve. The yield curve is just a chart plotting short-term interest rates against longer-term interest rates. Longer-term rates are usually higher than short-term term rates, if for no other reason than that it is riskier to loan money longer-term than shorter-term, so long-term lenders require higher rates to compensate them for this increased risk. That’s what it means to have a normal, “positively sloped” yield curve. Now, banks tend to borrow short-term funds and make long-term loans, and they make money off of the “spread” or difference between the two, which is why analysts often refer to them as “spread businesses.” Steeper yield curves thus tend to help banks, and flatter yield curves tend to hurt banks because the spread between what banks pay for funds (short rates) and what they earn on them (long rates) is smaller. Hence Yellen’s conundrum because raising rates and possibly flattening the yield curve could hurt banks. But with a stronger growth outlook going forward, and less risk of disinflation, I think there is a greater tendency for longer rates to rise, and indeed we’ve seen that since the election. That makes it easier for the fed to raise short-term rates, so my conclusion is that the election has actually increased the odds of the fed raising rates. That’s not the only way to reach that conclusion, by the way. You could also make the case (and many have) that higher deficits under Trump will lead to higher rates. But after over 20 years of looking at interest rates and deficits, I’ve concluded that there is no reliable relationship between the interest rates and budget deficits/debt in the US data.