I recently came across a very interesting report from the World Economic Forum about the future of financial services. It provides a very insightful overview of how technological advancement might impact different aspects of the financial industry, from insurance, to payments to investment management. I particularly appreciated that they provide different potential scenarios regarding how innovation might shape future developments. This is definitely a highly recommended read (the full report can be found here)!!
One analysis that caught my attention explores "How will smarter and faster machines transform capital markets?". They develop three potential scenarios regarding how a shift towards big data, "smart machines" and a focus on "real life events" might impact dispersion of opinions.
One scenario predicts that as we move towards smarter machines the focus of quantitative investment might shift from speed towards the development of models that focus on "real-life" events. The availability and usage of such vast and diversified sources of information might increase dispersion of opinions among informed traders, with the consequence of increasing intra-day volatility and widening bid-ask spreads.
The second and diametrically opposite scenario, instead, predicts that as programs, guided by machine learning, become increasingly more precise and widen the breadth of their analysis, opinions might converge towards a single view of the market. This would have the consequence of decreasing trading volume and virtually eliminating all possible arbitrage opportunities.
Finally, the advent of smarter machines which trade on the bases of automated data feeds, might raise discontent among "traditional traders" who might claim an "unfair advantage"; forcing the regulator to intervene. This might cause some parts of market-making activities to revert to manual trading with the consequence of reducing market liquidity. Additionally, market-makers would not be able to react as fast to fact-based price arbitrage, which might lead them to increase their spreads with unfavorable consequences in price formation for both buyers and sellers.
It is very difficult to predict at this point which scenario will prevail. For certain the market has been moving towards incorporating more and more "real life" and "unconventional" information into decision making and data providers have been adapting fast. For instance, Ravenpack now offers a live feed that provides sentiment indicators based on thousands of online sources.
My personal opinion is that we are still far from the development of a supercomputer able to extract unequivocal information from the millions of twits, blogs, google trends and news articles in real time. Most of all we are very far from the development of thousands of such machines who all think exactly in the same way. Additionally I am not sure that this kind of information will lead to more informative prices (closer to fundamentals). Are these "real life" sources at all informative? Maybe the future will still lie on the last hire of the average fund having to sit in the parking lot of some company to see at what time the manger leaves his office...
I guess we will just have to wait and see, but definitely this is food for thought!