I was invited to speak at the Princeton - UChicago quant conference on November 8 by Illinois Institute of Technology, Stuart School of Business. Chicago was a bit sunny when I arrived (a week later came the tornadoes). Confused between south and north, using Starbucks as reference points we moved around the city hopping business meetings and figuring out how to reach party ground in the evening. Moving from Barnes & Noble book readings to sitting through the Chicago music studios at Kildare, the city was bustling with activity. Two days were definitely less but the skyline, CBOT, CBOE, the L, and flying over Lake Michigan to cross over to Toronto gave little time to appreciate how cultures thronged across a water body.
All this co-existence of social, economic and natural systems somehow were on the same temporal scale as the quant strategies. Every natural system was temporal; then strategy and risk associated to these systems had to be temporal. This is how Citadel's Peng Zhao explained the quant strategies on a scale of holding period ranging from three months to high frequency like macro, statistical arbitrage, etc.
If the holding period could classify a quant strategy, extending quant strategies beyond three months could redefine the whole investing industry from active to passive and even redefine the indexing business. The smart betas, dividend indices, custom hybrids on one side and news based analytics on the other, the quants were transforming fundamentals, technicals, sentiment, economics, news into data and the analytics around it. No wonder three of the speakers at the conference talked about news analytics. Then of course there were options, efficient portfolios and a talk on greed, intuition and systems, and how markets could play the role of both a competitor and a friend.
The talk also explained how active vs. passive debate was redundant as both investing styles shared common temporal scale and active had no business to be, if it could not beat passive styles consistently. The very fact that active style has floundered was because of a lack of framework, lack of stock market science, which is why market participants see investing (trading) on a piece meal rather than on a continuum. We needed to find our context in a framework. Context is defining a relevant universe for a special stock market problem. This universe could not be the 100,000 traded underlying and or the one million top-traded derivative instruments. If computing was the answer, we would have found it by now. Computing could crunch big data faster; but markets needed data universality; common rules, which simplifies the problem rather than complicates it. If computing power had the answer, we should have found the framework by now.
Why was the context so essential to define any stock market problem? First; because it's easy to give the stock market problem a causal explanation. Second; the agents (variables) have increased; and many variables can claim to have a causal relationship to the behaviour. Third; despite the seemingly working causal relations active performance was dismal. Fourth; there was no objectivity regarding which cause was better than the other; was psychology more objective or was it sentiment data or was it fundamentals, statistics or technicals. We could just see causal patterns, but not measure them or understand what drives those causal reasoning. This is why the need for a framework which could encompass, explain, classify available investment approaches. Once the framework fell in place, quants could really take over and extend the work done by the UChicago laureates Fama and Hanson. The conference ended with a Sears (Willis) tower feast at the 66th floor.
The author is CMT, and Founder, Orpheus CAPITALS, a global alternative research firm
All this co-existence of social, economic and natural systems somehow were on the same temporal scale as the quant strategies. Every natural system was temporal; then strategy and risk associated to these systems had to be temporal. This is how Citadel's Peng Zhao explained the quant strategies on a scale of holding period ranging from three months to high frequency like macro, statistical arbitrage, etc.
If the holding period could classify a quant strategy, extending quant strategies beyond three months could redefine the whole investing industry from active to passive and even redefine the indexing business. The smart betas, dividend indices, custom hybrids on one side and news based analytics on the other, the quants were transforming fundamentals, technicals, sentiment, economics, news into data and the analytics around it. No wonder three of the speakers at the conference talked about news analytics. Then of course there were options, efficient portfolios and a talk on greed, intuition and systems, and how markets could play the role of both a competitor and a friend.
More From This Section
My talk on "extreme reversion" explained how there were many universal laws, but few or none were used to understand markets. Stock markets were natural systems expressing universal laws. It was not easy to build a framework using such laws, as failure or noise affected universality, too. Mean reversion was a universal law, which also suffered from failure and noise. The patterns and noise in mean reversion has been witnessed across subjects like behavioural finance, inter-market analysis, fundamentals and statistics, etc. However, little has been done to understand the failure of mean reversion. My talk focused on comprehending the failures of mean reversion and how it could be redefined into a better statistical and universal risk management framework, have cross sector application like investment management and predicting trends, etc.
The talk also explained how active vs. passive debate was redundant as both investing styles shared common temporal scale and active had no business to be, if it could not beat passive styles consistently. The very fact that active style has floundered was because of a lack of framework, lack of stock market science, which is why market participants see investing (trading) on a piece meal rather than on a continuum. We needed to find our context in a framework. Context is defining a relevant universe for a special stock market problem. This universe could not be the 100,000 traded underlying and or the one million top-traded derivative instruments. If computing was the answer, we would have found it by now. Computing could crunch big data faster; but markets needed data universality; common rules, which simplifies the problem rather than complicates it. If computing power had the answer, we should have found the framework by now.
Why was the context so essential to define any stock market problem? First; because it's easy to give the stock market problem a causal explanation. Second; the agents (variables) have increased; and many variables can claim to have a causal relationship to the behaviour. Third; despite the seemingly working causal relations active performance was dismal. Fourth; there was no objectivity regarding which cause was better than the other; was psychology more objective or was it sentiment data or was it fundamentals, statistics or technicals. We could just see causal patterns, but not measure them or understand what drives those causal reasoning. This is why the need for a framework which could encompass, explain, classify available investment approaches. Once the framework fell in place, quants could really take over and extend the work done by the UChicago laureates Fama and Hanson. The conference ended with a Sears (Willis) tower feast at the 66th floor.
The author is CMT, and Founder, Orpheus CAPITALS, a global alternative research firm