Political hot potatoes wherever you look

So, what makes 2024 so special that we identified elections as a key issue as early as November 2023? It’s only partly because U.S. presidential elections are on the list. These always tend to be big news events that everyone has opinions on. But to be frank, market interest prior to 2016 tended to quite limited and concentrated during the final stretch of the campaign and the immediate aftermath. 2016 was the year of both the UK's Brexit referendum and Donald Trump's surprising, and somewhat accidental, Electoral College victory.^[6] Ever since 2016, investor interest has generally been quite high in electoral events, especially in mature democracies.

 

To better understand the sustained, historically unusual level of market attention towards electoral politics, and to evaluate how long it might persist, one effective approach is to examine an electoral event that garnered much less excitement from both pundits and market analysts—the U.K. general election of July 2024. Overall, this was a throwback to a pre-2016 era in British politics, with the 1997 Labour landslide being a particularly apt comparison. An unpopular, long-serving Conservative government was widely predicted to lose, and it did. The UK's first-past-the-post electoral system, which accentuates swings in the House of Commons, gave the new government an oversized majority, even though various polls during both campaigns suggested that voters' expectations of the new team were somewhat low. Labour's landslide followed years of moderation after being cast into the political wilderness. But despite having plenty of time to prepare, the new ministry (as premierships are known in the UK) was soon embroiled in scandals, facing accusations of sleaze in the government.^[7]

 

And markets? Well, long before the voting actually started, equity investors had signaled that they quite liked the prospects of more boring, predictable politics. During and after the turmoil of the Brexit years, UK equities became increasingly undervalued compared to their peers. While there is healthy debate regarding the extent to which the underperformance of the London stock market is due to political factors, it is also influenced by structural factors, such as the sector composition of the main London indices.^[8] In any case, market reactions were slow and moderate. Instead, attention of politically minded market pundits quickly turned across the Channel to France. President Emmanuel Macron’s unexpected call for legislative elections came on the heels of steep losses for his allies in elections to the European Parliament^[9] News of the snap election for the National Assembly (NA), the dominant chamber of parliament, triggered a significant sell-off in stocks and bonds. This reaction was especially pronounced between the low-turnout first round and the more decisive second round.^[10] Subsequently, equity investors quickly realized that not much had changed, while bond investors remained concerned about the longer-term outlook for French public finances.^[11]

 

All of which is quite typical for how markets “think” about politics. An analogy would be artificial intelligence (AI) approaches that also rely on pattern recognition. Give market participants a representative sample of patterns for a given set of problems that are essential enough to make money, and they are sometimes remarkably good at predictive tasks. Over time, markets and machine learning techniques are becoming increasingly reliable at predicting future events from past experience, provided that the future is sufficiently similar to the recent past. But deviate from these conditions and conventional market wisdom can be as inaccurate or nonsensical as the AI hallucinations of even a state-of-the-art large language model (LLM). This might happen, for example, because there simply is no opportunity or motivation to learn from prediction errors.

 

The key is having both adequate data and proper feedback loops for market participants to recognize and assess patterns. As Friedrich August von Hayek observed, markets of various sorts can be a “marvel”. How else to describe their uncanny collective ability to absorb new information and often tacit knowledge spread across millions of individuals and aggregate it into price signals to guide decision making.^[12] There is little reason to doubt on either a priori or empirical grounds that this market magic can work on political forecasts just as it does with company events. From 1884 to 1940, for example, U.S. political betting markets were deep and liquid; and even in the absence of reliable opinion polling, they did a remarkable job in 14 out of 15 presidential elections, despite the oddities of the U.S. electoral college voting system.^[13] The same applies to a similar prediction problem from around the same time in another emerging market: how well stock markets, in particular, priced the risk of political violence in Tsarist Russia.^[14]

 

Nowadays, legal restrictions mean that especially in the U.S. political betting markets tend to be neither deep, nor liquid.^[15] As in mainstream financial markets, betting markets usually have plenty of other topics for their participants to consider. Moreover, the infrequency of elections and the inherent unpredictability of their outcomes pose significant challenges to developing accurate forecasting skills. Mastery in this area requires not only extensive experience across numerous election cycles but also a deep understanding of the unique factors influencing each specific electoral event. For example, election polling in India is notoriously unreliable, due to the heterogeneity of India’s mammoth electorate, political biases of the news outlets paying to conduct polls, as well as logistical problems and legal restrictions making accurate polling almost impossible. In France, turnout is known at times to go up dramatically between the first round and the second round, resulting in very different electorates.

 

Despite this tendency, market commentators routinely highlight early polling data to affirm existing investor beliefs, only to be taken aback by unexpected results on election day. This year, the elections in France and India proved no exception, offering ample cases of this phenomenon, similar to experiences seen in countries like South Africa. These instances illustrate the challenges in accurately interpreting market implications, even after election results are announced but while coalition formations are still pending.^[16]

 

Before 2016, investors were accustomed to predictable market patterns, guided by their experiences prior to the unforeseen market shifts of that year. During this period, markets often maintained a clear distinction between politics in emerging and developed markets. A seminal paper from the late 1990s highlighted systematic differences in how political events affected asset prices in these two different regimes.^[17] To simplify somewhat, mature democracies tend to be in fairly wealthy countries with a range of other institutions such as independent courts, functioning legal systems, a vibrant press and plenty of private enterprises. Whatever policies a new government decides to implement might well get overturned by the next one, but only within the framework of existing, established institutions and regulations. Under such circumstances, it makes sense for investors to price stocks mostly based on firm-specific information. In such countries, movements in individual stock prices tend to be less correlated with the overall market, and tend to decrease, as markets mature further (as was the case in the U.S. throughout much of the 20^th century). By contrast, in countries where laws protecting investors’ property rights and legal system are weak, and governments are relatively more prone to corruption and political rent-seeking, a lot more is at stake depending on who wins – especially if a slide into authoritarianism or a hybrid regime cannot be ruled out.

 

In summary, recent years have witnessed the U.K. and several other established democracies behaving more like emerging markets. With politics becoming more boring once again, we would expect to see a reversion to pre-2016 norms, in which it may pay for investors to once again pay less attention to politics. The 2024 U.S. presidential elections promise to be another key test case to this thesis, of course. We will leave that subject to another day.