With the increase in algorithmic and high-frequency trading, there has also been an increase in machine readable news services which attempt to gauge market sentiment and assist in measuring the impact news has on the marketplace in real-time.
These technology services can provide traders with the ability to construct trading models off of unstructured data such as news and social media streams. And, there are trading strategies that have been developed that rely solely on social media feeds as the basis for analytic trading decisions.
Of course, there are questions surrounding these services such as are they reading the “sentiment” correctly; are all the relevant news streams being captured; and even if they are reading the “sentiment” correctly, is the associated market move predictable?
As some companies will explain, even though they may be analyzing the news feeds and determining the appropriate sentiment, the market does not always move as they would expect. Recently, great strides have been made in this area in terms of better refinement of discerning public emotion to news and resultant market moves and blending market sentiment with more traditional quantitative models.
The capabilities of the portfolio managers and traders to understand and work with these feeds can also not be underestimated, and some have been able to use this information in creating sizable returns.
However, whereas this development was predicated by the increase in algorithmic trading, this technology can have valuable uses in other areas as well. Understanding these more qualitative views can also be of use to risk management, particularly in the areas of market and credit risk. For instance, in one presentation, an example was shown of what analysts would traditionally use as comparable companies, but when a machine readable news feed was brought into the analysis, it demonstrated that the market viewed correlations and comparables much differently with the incorporation of market sentiment.
Market sentiment as related to macroeconomic news, corporate reports, countries, products, social trends, associations can be beneficial to the risk manager in refining correlations, volatilities and market shifts. This information can be used in the construction of models; early identification of potentially large market moves; and in the understanding that what may be first perceived as an outlier, may be a precursor.
Machine readable news data helps bring together the fundamentals with the quantitative data and gives an insight to market emotion and overall public perception in a measurable format. There can be benefits to trading strategies, asset allocation, risk appetite, credit and market risk management, and even to individual companies. But, the accuracy in behavioral analysis and what the perceived sentiment may predict in terms of market moves and trends is still being refined. Personal expertise and judgment is still required in the investment and risk management process.