Listen to this article
In his first few days in office, Donald Trump has begun to fulfil his mandate, doling out executive orders left, right, and centre. The first batch arrived on the week of January 20th, but it was not until the weekend of January 28th that things really started to heat up. That weekend saw the world respond to his immigration ban announced on Friday, and observers started to worry as to how markets would react. The world awaited a Monday that would probably see major volatility – what investors and traders fear most.
As expected, the VIX, a commonly used indicator of the volatility of the S&P 500 index options, jumped.
Markets fell across the board, and the sugar rush since Trump’s election took another hit. But were markets already developing a sense of fear before then? If fear in the markets eventually translates to volatility, then surely it needs to be gauged. This seems a difficult thing to measure accurately – but it can be done.
What Does the Crowd Think?
The concept of the crowd and their sentiment (what direction they lean in) is something that historians and writers like Tolstoy, to analysts and pollsters like YouGov have tried to figure out. Some try to take sample sizes, others try to see what key ‘influencers’ are saying, and others yet try to look for tell-tale signals.
The beauty of the modern world is that people can now share their thoughts from anywhere in the world on any topic they choose, and put it into one place online. What people are saying about certain topics can be crowd-sourced and passively analysed to gauge their overall sentiment. This type of data is called Sentiment Data.
In the world of finance, there are over a million articles and posts produced every day from all over the globe. They can then be distilled through a natural language processing (NLP) algorithm. NLP can vary from using deep-learning to running a stream of words through a dictionary, giving it a score indicating a level of overall sentiment.
Mission: Difficult, but Not Impossible
Anyone who works with data can confirm that turning the qualitative into the quantitative is hard but also extremely valuable. Language is difficult to transfer, translate, manipulate and process, in a way that numbers are not. For instance, for one million articles and posts taking an average 3 minutes to read, one would need approximately 2,080 analysts to work over 24 hours and the mother of all committees to agree on what the data says. It may sound like a typical Monday morning meeting, but that is not what happens.
Most Investors and Asset Managers Don’t Even Try. Why Not?
What the crowd thinks is vital. Traders and investors are constantly trying to predict what others will do, and to use that knowledge to determine what they should do themselves. Usually, this is done using their knowledge of prices alongside any indicators developed by analysing those prices (such as price volatility indices). But this has proven time and time again to simply not work well enough for investors to make good returns.
New and different strategies for reacting to prices are constantly being introduced. However, while prices will remain critical, understanding how markets respond when they experience certain emotions – especially optimism and fear – is invaluable.
Although markets will always be susceptible to shocks, understanding the sentiment in a market helps to understand its momentum and the destination that that momentum is building towards.
Why Sentiment Data Is Relevant Today
The processing power, the algorithms, and the ability to draw from immense data pools (Amareos, for example, use some 50,000 data sources) are what now makes this dream a reality.
What is also important is the depth to which sentiment data can be explored. Instead of just suggesting that people feel positive or negative in certain market circumstances, the data can now determine and express more specific emotions.
The level of fear in the market can be seen to relate directly to the volatility it experiences, but what about optimism? Or trust, anger, anticipation, and the other primary emotions people experience? And can these be built into something beyond a mere index?
Evaluating conversations around commodities, currencies, and equities on a daily basis produces a level of granularity not seen before. Today, we can answer the question of someone’s sentiment regarding NASDAQ: AAPL. Are they relatively angry, optimistic, or even experiencing peak anticipation?
Granularity also comes from identifying which source the sentiment is coming from. For instance, from the constant stream of news, Amareos can distinguish between social media and mainstream sources.
Before the Italian referendum, a look at mainstream news would have suggested an element of optimism for Renzi’s pro-change cause. The majority of anti-Renzi voices were heard mainly in social media, though, and paying due attention to this made despite his eventual demolition in the polls unsurprising.
The Big Picture, the Big Crowd
Of course, it is critical to read – hence why sources like Bloomberg, the Economist and The Market Mogul are so vital. One must be able to read and understand the reasons why the world may be experiencing this or that sentiment, from a variety of perspectives and trusted sources. Hand-in-hand with information on what every individual making up a market feels, that knowledge is a powerful tool.
Embarking on a new journey, the most important part is knowing where you’re starting from. Making a trade or giving a prediction is as strong a move as the ground from which you step off.