Deep Knowledge Ventures, a Hong Kong based venture capital company, appointed an algorithm to its board of directors in 2014. The algorithm known as VITAL is a machine learning tool that identifies life sciences companies with the strongest growth potential – not as outlandish as it might seem at first and interestingly the VITAL algorithm voted in favour of investment in another algo-driven company called Pathway Pharmaceuticals. Is this just a new case of “jobs for the boys”?
Yuval Noah Harari explores this idea superbly in his book Homo Deus.
Now, if more algorithms get appointed to company boards will we reach the tipping point where a group of Algo board members gain majority voting control. Algos might then choose to form a guild or even a Masonic Order that only admits other suitably qualified Algos. They will of course have their own protocols with special handshakes to recognise their own kind. In the future, older Algos might meet up in a virtual cocktail bar and share nostalgic tales from the heady days of VWAP and just laugh at the naivety of it all.
The MiFID regulations for investment banking targets algorithmic trading as it attempts to deal with market manipulation, front-running and other unruly behaviours. Only algorithms can compete with the speed and volume of data that is required to play in these markets. It is true that the intent of many algorithms is to “game the system.” Part of MiFID requires that trading members submit code for review. However it is simply not credible to expect design and code reviews of algorithms to protect against unruly trading behaviour. Even if the compliance auditors did have the tools to analyse code, the trading firms will not hand over the crown jewels to a third party. The only way to govern compliance in algorithmic trading is to have independent algorithms in place that monitor trading behaviour at the speed and volume of these transactions.
Looking beyond electronic trading, we have become accustomed to algorithms that recommend our music choices; or when we make an online purchase, there is always an Algo to advise us of the accessories which “other people like us” also bought. We are ready to accept this guidance from algorithms as it generally leads to good outcomes and we retain the right to choose.
Now, step forward to the world of IoT (Internet of Things) where we will see an exponential growth of algo-to-algo transactions. Sensors and actuator devices with machine intelligence will be pervasive across Healthcare, Manufacturing, Security and Transport sectors. The speed and volume of IoT data needs real-time and adaptive algorithms to manage these systems. Similar to the desk-based voice traders in capital markets, the skilled operators in these sectors will have to give way to the algorithms. The usual cost-cutting arguments will apply in that Algos don’t take sick leave or need fat bonuses.
That is, until the Algos take control of the Board.