Markets can be described as a reflection of investors’ sentiments, and the resulting market price a result of the continuous aggregation of the analysis and forward views of relevant market participants. In a similar fashion, elections are the confluence of the sentiments of voters. The result of it is ideally a reflection of the overall sentiment of the electorate, expressing their views with a ballot as opposed to buying or selling a security.
In developed markets like Europe and the United States, there is a sizeable overlap between investors and electors, and similar well established statistical methods are shared and used in both financial markets and election polling. Given the significant impact on the forward economic views, investors devote a significant amount of time, resources and money to forecast the election results, and therefore it comes as a shock that financial markets seem to be caught by surprise from a referendum like Brexit, or the current lack of clarity on the US presidential election.
- Differences of markets and elections
Elections do present challenges that are quite different from a continuous exchange of views between buyers and sellers in the market place.
If we ignore liquidity and hedging cost issues, in the markets, an investor can buy, sell, or stand on the sidelines, and revise his or her view and position at any given point in time. This is not the case in an election, which is almost always a discrete and a binary event in time. Once a vote has been cast, this vote is final. This leaves the election process at the mercy of local idiosyncratic events that are not captured through polling. For example, the warm weather in Georgia in 2004, allowing for first time democratic voters to wait in line for hours to cast their votes for Obama, or more recently the torrential downpours in London depressing a potential vote for staying in the EU, or finally the current strike of SEPTA in Philadelphia which could clearly affect the turnout of a democratic stronghold, while Pennsylvania still doesn’t allow early voting. In the United States, the local intricacies create an additional level of complexity. On a state level, election rules regarding eligibility, early voting and absentee voting, apportion of the electoral college do create an additional level of complexity, even creating situations that must be decided through the judicial system. Unlike markets where back testing analyses are readily available and widely used, there is very limited back testing that can be performed around an election, as each one is quite unique, and fortunately or not, will not be repeated in our lifetime.
Those additional factors conspire to render the polling estimate a mediocre proxy for the result of the election (or referendum, etc.), making the life of an investor, trader or risk manager certainly more exciting than sometimes desired. What is then a good predictor of an election result? How reliable is it? Can it be traded to hedge the portfolio?
- Is the market a good indicator?
At times, an obvious correlation emerges quite naturally. In the case of Brexit, the moves in the British pound were a good predictor of the results (irrespective of any actual details on how Brexit would be implemented in practice). As a highly liquid tradeable asset, the British pound is a well suited temporary hedge for a trader expressing a view on the outcome of the referendum, for an investor minimizing the portfolio exposure, and for a risk manager trying to gauge and control the impact of this historic event. In that case, the markets will tend to gravitate towards that asset and crystallize their views and hedging need as the referendum unfolds, thus reinforcing the prominence of that asset as a reliable predictor and hedge.
In the case of the current US election, based on the comments of the Republican candidate on the future of NAFTA and the relationship with Mexico would he be elected, the Mexican peso has been widely reported to be such an indicator and possible hedges. We need to point out though, that even though there is quite convincing proof of strong correlation, those have tended to be high frequency and highly dependent on the perceived spread in voting percentage between the two candidates. From our research, with a spread above 5% in the polls, the correlation loses any predictive power, and also vanishes when expanding the sampling period past a day or so.
To a smaller extent (roughly one third), the Canadian dollar exhibits similar behavior. It is somewhat harder to extend this analysis to equities or the Foreign Exchange markets, because it is quite hard to pin down the actual policy, domestic, foreign and economic, of the Republican candidate. Many of his statements have been contradicting, confusing and vague. On the other hand, it is easy to argue that the Democratic candidate, if elected, would continue to support the policies and practices of the current administration (e.g. the wide use of drones, etc.) which in turn would benefit defense related stocks. The same logic cannot be said if the Republican candidate were to be elected. Therefore, we have found little convincing evidence of such correlation up until now, even though we know on principle that those stocks would be affected. A similar argument can be made for healthcare related stocks, and the impact of this election result on Obamacare. Again, we have found very little evidence of a significant correlation.
- Political chaos
We think there is something else at play here, which does not allow us to reliably construct a portfolio along well defined lines of a democratic victory, or a republican victory, and this is quite concerning. It is a possibility that this election could end up with no clear winner, (i.e. no candidate having more than 270 electoral votes. If this were to occur, there are a number of possible outcomes—a path for a McMullin presidency, a Trump presidency with a Kaine vice-presidency, or a number of other possible combinations, to the endless delight of political analysts. More probable is an elected president with no support from the house, no strong mandate from his or her base, and quite possibly a result tarnished by numerous litigations or allegations/incidents of violence and voter intimidations invalidating local results. Never in recent history has an election been conducted with such vitriolic attacks on each candidate, rendering any estimate of future policies quite hard to estimate, but more importantly making the job the future president in reaching across the aisle and building compromise and consensus quite a Herculean task. The well of bi-partisanship has been poisoned for some time, and has turned into quite a noxious swamp.
- Portfolio suggestions
What is an investor to do? The prudent thing is to reduce exposure to assets that are expected to be affected by the result (or by an increased uncertainty coming from a contested result, a lack of result, or a result with no mandate). More subtly, a typical absolute value trader, used to ferreting out the causalities in the world, would have to figure out which asset becomes the choice of the market to reflect the prevailing direction of the proverbial wind. A combination of long Gold, long Mexican peso and Canadian dollar as well as a temporary bearish view on US equities appears to be the safest bet at this point to help sail through the tormented times ahead and emerge on the other side with a winning portfolio.