Political risk and markets
Politics moves markets. One need only look at the dramatic movement of global equities and bond markets in recent years, as politicians have fumbled their way through the global crisis, to understand this reality. Even without a crisis political decision-making, or the lack thereof, influences investment outcomes.
The political causes of these market movements are varied. They include elections and the degree to which the outcome was expected, changes in political leadership, dramatic shifts in public policy and the underlying ideology of the political party in office.
US elections and markets
Which political party is better for market returns? For many the answer is intuitive. Republicans are considered the party of free markets and therefore are more inclined to promote and support policies that benefit business and shareholders. However, this perception is not always supported by the data.
Since 1965 the S&P 500 has performed better with a Democrat, rather than a Republican, in the White House. The average annual compounded gain for the S&P 500 with dividends included, equaled 10.9% under a Democratic President, versus 7.3% with a Republican President. The compounded annual gain over this entire period was 9.1%. (Kass, D, 2012, Seeking Alpha, The Presidential Election 2012: The Economy And The Stock Market, 4 Sept 2012).
Yet other data points suggest the very nature of the question being asked is producing an answer that means very little. These theories suggest it is the electoral cycle itself rather than the political leanings of the victors that most influence the market.
The Presidential Election Cycle Theory argues stocks decline in the year immediately following a presidential election, then go up over the course of the next three years, no matter who wins the election. While there is evidence to support this view it is not 100% conclusive and relies on averages.
From 1937 onwards the Roosevelt, Truman and Eisenhower Administrations all adhered to the theory beginning with market declines in their first year followed by a rebound over the next three years of the term. However, more recently the elections of George H. W. Bush and Bill Clinton brought market increases in the first year. (Francis, D, 2012, US News, What the Presidential Election Means for the Stock Market, 19 April 2012). In essence, the averages work but there is a touch of luck involved in picking which Presidents will produce above and below average years.
More recent data – from Ned Davis Research – suggests that the analysis should consider the question from the perspective of incumbent versus challenger rather than political parties or stages in the electoral cycle. This analysis leads to a clear trend in favour of incumbents regardless of their political party of origin. On average markets have performed best in presidential election years when the incumbent has been returned to the White House. However, again this analysis is relying on averages allowing us to identify a trend – albeit a strong one – rather than an ironclad law.
Unlike equities the performance of bond markets seems to more clearly correlate with the election of Republicans to office.
According to a 2006 study long-term government bonds produced an average annual return of 4.14% when a Democrat was in the White House and 10.80% when a Republican was in control for the period 1961 - 2004. The average annual return for the full period was 7.77%. This gap has since closed, however, as long-term government bonds have performed very well since President Obama was elected in 2008. (Kenny, T, 2012, How do elections affect bond market returns, citing 'Tactical Asset Allocation and Presidential Elections' 2006).
This Republican skew is supported by data from the Barclays US bond index. Since 1929 the index has had an annual nominal gain of 1.9% under Republicans as opposed to an average loss of around 1% under Democrats. (The Economist, 2012, Voting with the wallet, 6 October 2012).
Corporate bonds appear more correlated to the electoral cycle than any particular party. A 2004 study showed that corporate bonds perform best relative to Treasuries in the months immediately surrounding the election. In the period from 1928-2000, corporate bonds outperformed in the months of August and September prior to the election, underperformed slightly in October and November, then outperformed once again in December and January. The two best months of relative performance were the January after the election, followed by the August before. The positive effect on corporate bonds gradually dissipates following the February after the election, and disappears entirely by April. (Kenny, T, 2012, How do elections affect bond market returns, citing 'Tactical Asset Allocation and Presidential Elections' 2006).
Of course these pieces of research examine what happens once the election is over and the result is known, which is all very well for academics but investors are keen for insight into what the likely election outcome will be so they can adjust their risk strategies accordingly. For this answer investors are perhaps best off looking at betting markets.
Anecdotally, betting markets have been reasonably prescient in recent electoral cycles. However, the relative newness of these exchanges after a long hiatus means any meaningful analysis is challenged by the size of the sample. Examining the performance of betting markets in their heyday, and before the rise of opinion polling, reveals that putting your money where your mouth is introduces a deeper level of analysis resulting in betting markets proving to be highly predictive.
In their paper ‘Historical Presidential Betting Markets’, Paul Rhode and Koleman Strumpf establish the case for the predictive power of betting markets. In particular, they found that in the fifteen elections for the period 1884 – 1940 the mid-October market favourite went on to win 73% of the time. The underdog won only once and in the three remaining cases pundits had essentially brought the market to a statistical dead heat. (Rhode, P, W & Strumpf, K, S, 2004, Journal of Economic Perspectives, Historical Presidential Betting Markets, vol. 18, no. 2.).
Recent research has suggested a correlation between political party ideology and market outcomes. Anderson, Malone and Marshall argue the electoral cycle influences asset returns with property performing better under Labor governments and stock prices more favourable to Conservative governments. In the case of bonds capital losses are discernible during Labor periods and capital gains are apparent under Conservative governments. (Anderson, H, Malone, C & Marshall, B, 2004, n.d., Investment returns under right and left wing governments in Australasia).
For investors the challenge is not just identifying the emergence of political risk but being able to determine with a degree of certainty the market impact of that risk, including any ripple effects.
Adding to the challenge is that the relationship between politics and markets is not always linear. Rather than political decisions having an immediate and verifiable impact on market performance it may be market events themselves that shape political outcomes that in turn result in a new outlook for investors.
Regardless of whether the relationship between politics and markets is causal, or otherwise, the fact that a relationship exists should be an important consideration in any investment strategy.