Friday 20 April 2007

Trading On The French Presidential Election

Sunday sees the first round of voting in the 2007 French Presidential Election and it seems a suitable prompt to produce a post about trading political markets. Political markets are easy to trade on, they aggregate information, and they produce real-time probabilistic predictions on future events.
There has been a considerable expansion in the number of trading firms offering political/prediction markets in recent years and the consequent increase in the range of markets has made the sub-sector worthy of the analytical attention of all professional traders. Liquidity is still thin on the ground though in the non-mainstream territories which is represented by the generally prohibitive margins in such countries. We will use the race to the Élysées as an example to demonstrate some of the factors that traders need to be aware of with respect to political markets.
1. Opinion Polls - Opinion Polls (OPs) take on the form of an index market where the percentages for the candidates aim to replicate the true percentages out in the country. Polls are different than a true market in several ways. Firstly, they are not representative of money flow or inside information. Secondly, the percentages have a built in standard deviation that makes such polls of minimal value even when all other factors are taken into account. Thirdly, people may be prompted to answering a certain way by the deliberate or accidental manner of phrasing of the question. Fourthly and most importantly, the politics of the polling firm will impinge on output on a variety of levels.
2. Trends, Breakpoints and Uproar - By plotting OPs (with error bars) against time, an observer is able to detect any trends that the OPs are exhibiting. All other things considered from a market perspective, you would generally be looking to be supportive of a candidate on an upward trend. Breakpoints occur when a key piece of electoral development occurs eg a head-to-head televised debate or a particularly poor PR occurrence. Sometimes a breakpoint may be so severe that uproar results and a candidate is blown out of the water - remember Howard Dean's 2004 US presidential campaign.
3. Many voters are not willing to disclose their election preferences to pollsters. Either such voters have not made up their minds or they may be ashamed at the admission of their choice. In the 2002 election in France, Le Pen surprised all pollsters by exploiting such guilt-ridden shame to reach the second round of voting ahead of the Socialist candidate.
4. Elections that are divided into two rounds like the French are very different to model than one-off races. The modelling of the events is considerably more complex in a two round race both due to the structural nature of the event and the impact of voters whose original choice is eliminated in the first wave of voting. The very nature of two round markets encourages voters to undertake a first round protest vote before supporting their mortgage and their class in the second round.
5. The key people in all elections (from both a political and a market analytical viewpoint) are the swing voters and the masses of the befuddled. Up to 40% of voters may exist in this state of limbo in the period leading up to polling day. The control of the mainstream media is critical in the final days as this is the window when the majority of these unfortunates develop a political view about the world in which they live. Countries like England, USA and Italy are particularly susceptible to late spectacular society targeting of "the don't know's" due both to the degree of control of media in such countries and the level of voter apathy.
6. One of the most important factors to assess in any election is the great intangible of national feelgood factor. The mathematics behind such fuzzy concepts are complex and cover a wide spectrum of societal and political forces as well as taking into account mass psychologies. Indeed, we use such an approach in international football matches in addition to political markets as there are clear parallels at play here.
7. Regional politics also impacts upon national elections. Some areas have moved en masse to the right in recent years eg the former Eastern Bloc while others have swerved to the left eg South America. Such continental momentum must be considered.
8. Increasingly, in a world of spectacular events being perpetrated by various groups of driven individuals, first world elections possess a structural volatility relating to the impact of a potential major atrocity in the election run-up. The Madrid bombings were so mishandled by the Partido Popular that they effectively handed the election to Zapatero in 2004.
9. Analysis of the market prices offered by the brokers is another key source of information. Many traders and market makers have clear holistic overviews of their sectors and market memory enables them to gain an edge through superior pattern recognition skills. This feeds through into price. Additionally, many of the firms that price up political markets are also involved in international financial markets and, in such cases, their territorial appraisal is enhanced further.
10. Underhand and, even, corrupt practices must also be taken into account. What sort of creative gerrymandering has been abusively legalised? How is the fear factor in the territory? If valid, what is the early reaction of external electoral observers? Does the power reside in the vote or in some external body eg the US Supreme Court? Are the computerised voting machines open source?
11. The final key variable that we shall discuss here is turnout. Increasing depoliticisation and apathy are the norm in developed economies and modelling the ability of the individual parties to get their people out is very key.
We are absolutely unwilling to disclose our proprietary analysis of the French election this weekend but we will point you in a couple of intriguing directions. The markets currently project a Sarkozy victory at a 65-70% probability which is, no doubt, based on his OPs lead. Voters who allowed the National Front candidate to reach the second round last time are going to be much warier this time with their first vote flippancy - we expect that this will favour Ségolène Royal and undermine the middle ground candidate, François Bayrou and the candidates on the left. Additionally, we expect Le Pen to steal some votes from Sarkozy as they battle for the right wing vote. Our estimation prior to the weekend media onslaught is for the two market leaders, Sarkozy and Royal, to progress to the second round but it should be noted that a leak from the Renseignements Généraux (the French home intelligence service) suggests Le Pen will again reach the run-off. The very leaking of this report makes such an outcome less likely.
Dietrological Trading Team divide our market activities between several different types of financial market. Our current market project portfolio involves Football Trading Pre-Match (45%), Football Trading In-Running (15%), Cricket World Cup (5%), International Financial Markets (25%), Political Markets (5%), Miscellaneous Markets (5%). Political Markets are still at the periphery of the marketplace but we expect that this sub-sector will markedly expand in the coming years particularly as all markets gravitate towards a global trading platform.