It’s an election year. One presidential candidate is pledging that, if he’s elected, he’ll raise taxes on the wealthy. The other promises that taxes will never go up on his watch.
Sound familiar? The year was 1992. Democrat Bill Clinton wanted to lift the top marginal tax rate from 31 percent to 39.6 percent. Republican George H.W. Bush was doubling down on the vow that had helped him win four years earlier: “Read my lips: no new taxes.”
That election happened 20 years ago, but it has lessons for today’s investors, says accounting professor John Robinson at the McCombs School of Business. In a 2005 research paper, he and two other scholars examined whether the tax rhetoric of political candidates during the 1992 campaign affected the prices of bonds and stocks. Their conclusion: It did.
“It’s a lot like what’s going on right now,” says Robinson. “Clinton said he was definitely going to raise taxes, particularly upon the wealthy. Today, Obama is saying we’ve got to have a millionaire’s tax, while the Republican nominee is saying, ‘No, we’re not going to have any tax increases.’”
The original idea for the research, he recalls, came back in 1992, when a colleague was teaching a class on tax planning. A student suggested that the election could be influencing the price of securities.
Robinson and his colleagues were intrigued by the notion. They had already been researching the ways in which taxes were capitalized into security prices. But what economic yardstick could measure an election cycle against a market cycle? “It took us awhile to find an angle,” he says.
The first key was to choose a type of security that was especially sensitive to taxation. He found it in municipal bonds.
Unlike most bonds, the interest on municipals is tax-exempt. If investors expected taxes to rise, Robinson theorized, they’d pay more for tax-exempt bonds and less for taxable ones. Their tax-exempt returns would drop as municipal bond prices went up, while taxable returns would move the opposite way. The gap between the returns, called a yield spread, should get larger, as more investors expected Clinton to win.
Yield spreads were easy enough to track. Robinson compared an index of AAA-rated municipal bonds against an index of U.S. Treasury bonds. The two groups were identical in ratings and maturity dates, but the Treasuries were taxable.
That left a second challenge: How to measure the expectations of investors, without getting inside their heads. He found the answer in an unusual kind of investment market: the Iowa Political Stock Market.
Launched in 1988 at the University of Iowa’s Tippie College of Business, the IPSM (now the Iowa Electronic Markets) was designed to test how efficiently markets collect and analyze information. Could a futures market, in which investors bought and sold shares in presidential candidates, predict election results better than public opinion polls?
The shares were strictly penny stocks: They paid $1 if the candidate won and nothing if the candidate lost. Over the 82 days leading up to the election, share prices would fluctuate between $0 and $1, depending on how likely investors thought their candidates were to win.
“The prices of these shares vary according to the probability that a candidate will get a elected,” says Robinson. “The nice thing about this effect is that it’s observable each day. The electronic market was basically our control for the probability of election.”
Besides its convenience, the IPSM was uncannily accurate. In 1988, its forecast of Bush’s victory margin over Michael Dukakis had been closer to the election results than any of six national polls.
When Robinson compared the daily rise and fall of candidate shares with the ebbs and flows of yield spreads, he found close parallels. A 100 percent probability of Clinton’s election, he found, would shrink the gap between taxable and tax-exempt returns by 3.12 percent for 5-year bonds and 3.50 percent for 10-year bonds. Investors, in other words, factored in a future tax increase of at least 3 percent.
“What we found,” says Robinson, “was that municipal bond prices went up as Clinton’s probability of being elected went up. On securities that were fully taxable, the prices went down. The market basically impounded the expectation of a tax increase.”
The effect was smaller on 30-year bonds: only 1.73 percent. That means, says Robinson, that investors didn’t expect higher tax rates to last. The election of Bush’s son in 2000 proved them right.
To double-check his conclusions, Robinson found a similar, though weaker, effect among stocks. Equity investors wanted their overall returns to be 1.46 percent higher if Clinton won — presumably, to compensate them for higher taxes on dividends. As with taxable bonds, they pushed stock prices down to get those higher returns.
What might Robinson’s research mean for 2012? Could a canny — and fearless — investor make money off this year’s election, betting on how bonds and stocks might respond to either candidate’s tax rhetoric?
Possibly, says Robinson. “It depends on your personal beliefs and how risk-tolerant you are. If you believe strongly that Obama will be re-elected and that Democrats will retain control of the Senate, and you’re willing to take some risks based on that belief, then you should see some reactions in securities prices.”
He’s more interested in investigating how securities react to other kinds of governmental actions. “Lots of analogies should be reflected in securities prices,” he says. “There are a couple of electronic markets now, and they have bets on all sorts of things.”
As an example, he points to the ups and downs of the Patient Protection and Affordable Care Act and its effects on insurance companies. “The odds are that the health care industry will be severely regulated,” he says. “Depending on how you feel about what’s called Obamacare, we could see if it actually is reflected in securities prices for the insurance industry, and to some extent, the health care industry. If Obamacare is repealed, will insurance companies suffer, or will they have higher returns?”
You can bet Robinson will be watching.