“America’s stockmarket has gained more under Democratic than Republican presidents” – “GOP Smacked” The Economist October 9, 2012
“Mitt Romney has forecast that the optimism resulting from his election to the presidency will give a lift to the economy… the claim that he will deserve credit for good things that happen after he wins…. In general, voters have tended to throw out the incumbent party after four-year periods when the Dow rose at a compound rate of less than 5 percent a year, and to keep the incumbents if the market gain exceeded that figure. If that pattern continues, Barack Obama will win re-election.”
— “Presidential Stock Markets” New York Times. October 26, 2012
“…Democratic presidents have been far better for the stock market, despite the GOP’s reputation as the party of business. A study by Barclays going back to 1929 shows that …the market rose 10.8% annually under Democrats and just 2.7% under Republicans. And Barack Obama has been well above average.”
** “Market-Election Gauge Bodes Ill for Stocks” Spencer Jakab, Wall Street Journal October 31, 2012
“Historically, a Democrat without business experience has been extraordinarily better for the economy and the stock market than a Republican who had a career in business. In the past 84 years, GDP has grown 7 percent per year under Democrats without business experience (FDR, JFK, LBJ, Clinton and Obama) and fallen by 0.2 percent per year under Republicans with business experience (Hoover and the two Bushes). The Dow has risen an average of 16.8 percent per year under Democrats without business experience and has fallen by 3.7 percent per year under Republicans with business experience. … There is a saying: “If you want to live like a Republican, vote Democratic.” Perhaps it should be amended to: “If you want to live like a successful Republican businessperson, vote for a Democrat without business experience.”
Robert McElvaine, Washington Post October 19, 2012
These quotations are a small sampling of the punditry in advance of next week’s Presidential election. One can draw many inferences from the intersection of economic and political data. How useful are such inferences as a guide to what’s ahead?
Not much. The simple inferences fail to acknowledge important cautions from the field of data analytics, such as these:
· Association is not causation. Does the party in power really produce the stock market returns? Or does the causality run in the opposite direction? Perhaps the recent stock market returns incline voters toward a certain party: elect Republicans if economic conditions have been buoyant and Democrats if not.
· Omitted variables. Perhaps other forces in the economic environment produce the stock market returns and the voting results. For instance, an ongoing war might favor re-election of the incumbent President; and war spending might stimulate the economy. What else might be going on to explain any particular election outcome and investor return?
· Regression toward the mean. Investor returns tend to oscillate around an average; when returns are high or low, they tend subsequently to move back toward the average. This tendency of regression to the mean is a regularity in the natural sciences and social sciences. Perhaps it has been the Republicans’ dumb luck to gain political control at the top of the economic cycle (or Democrats at the bottom of the cycle), only to see it regress back to the mean. To the extent that regression prevails, investor returns would have less to do with the politics and policies of the party in power. As Spencer Jakab wrote, “Partisan yardsticks of stock-market performance clearly leave a lot to be desired as good runs like Mr. Coolidge’s and Bill Clinton’s occurred during bubbles and bad ones like Mr. Hoover’s were in large part due to rotten timing.”
· Lag effects. Policies that Presidents implement probably have more impact in the long run than the short run. Lyndon Johnson handed to Richard Nixon the Vietnam War and rising inflation, the consequences of which took years to control. Obamacare is projected materially to control growth in health care spending by the Federal Government starting later in this decade and early in the 2020s. The rule-writing under the Dodd-Frank Act is far from complete; its impact on the financial services industry remains to be seen.
· The problem of few observations. Since 1789, U.S. citizens have elected 44 Presidents, each of whom lasted an average of 5 years. Social scientists get worried when the number of observations in a study grows “small,” less than a few hundred—this is because in small samples, a few truly “rogue” observations can skew the results. Then, you should consider that 31 of the 44 Presidents served less that two full terms: in just 13 cases are we able to observe the impact of a President’s policies applied over an extended time.
These and other considerations challenge the validity of inferences drawn from statistical analyses about the relationships between Presidents or political parties and returns to investors.
To be clear, I don’t question the suggestion that politics matters to investors. I just doubt the value of historical statistical analyses that “this kind of President” or “that party” will produce particular investment results.
So, what is an investor supposed to do? Politics matter, but the policies and their implementation matter more. In times like these, careful investors need to be policy analysts. This entails anticipating the political choices made by the President and Congress, and other arms of government—these choices become policies. Ultimately, the investor must assess the economic consequences of the policies. The immediate weeks and months after the election of November 6th will be a period of intense scrutiny and speculation about the forthcoming policies and implementation.
The conference that Darden will gather in two weeks will draw on the insights of an impressive line-up of speakers and topics to jump-start investors’ reassessment of the post-election outlook. More details are available at the conference web site. Join us!