In a famous experiment in 1906, members of a crowd at an agricultural fairground guessed the weight of an ox about to be slaughtered. Some participants were experts in livestock, others were not, and anyone at the fair could participate. The average guess of the crowd (of 800) turned out to be a very accurate one, considerably more accurate than that of the experts as individuals.
Dr. Yaniv Altshuler, described on the MIT Media Lab site as an “expert in collective intelligence methods,” working with Prof. Alex Pentland and the eToro investment network, developed a much more recent experiment to test the ‘wisdom of crowds.’
There are conflicting intuitions (or conflicting clichés) here. On the one hand, “group think” is a bad thing. It means no one thinks for him/herself, everyone drowns judgment in a collectivity, which amplifies bias. In financial terms, the critics of groupthink can point to bubbles and greater-fool speculation.
On the other hand, the “wisdom of crowds” is a good thing. After all, consider the weight of that unfortunate ox.
The problem, for Altshuler and Pentland, then, was: can individual traders harness the wisdom of crowds and avoid the pitfalls of group think?
Coupons and Advice
Altschuler distributed trading coupons with a face value of $20 to each of 500 active financial traders within eToro. He also provided those traders with the name of a particular “guru” they might want to copy. This was a small part of the whole; there were more than 2 million eToro members.
Given the nature of the eToro system (a trading platform equipped for Facebook-style networking) any of these individuals could have copied any of those gurus anyway. But the guidance (coupled with the coupons) presumably got their attention.
How did Altshuler decide who to recommend to whom? The statement from the MIT Media Lab is a bit evasive on this. The matching was the result of “an innovative algorithm designed to optimize information flow within the network.”
At any rate, it seems to have worked. What he calls a “tuning” of social trading networks increased gains both against online traders without the social networking and against online traders with untuned networking (i.e. those who select their favorite gurus without the benefit of coupons or algorithm.)
Pentland describes the research this way, “[W]e are helping to optimize the information they draw from the network, providing higher likelihood of increased gains, and helping to avoid dangerous market bubbles.”
Ahead of Schedule
With these thoughts in mind, I spoke recently to the CEO of the eToro Group, Yoni Assia. He described eToro as the largest social investment network in the world, and explained its mechanics thus: “If traders in the network find, follow, and copy another trader (a guru, in the language of the site) who is creating alpha, they are essentially copying his alpha.
“The most copied gurus on the site have about 4,000 copiers (users linking their trading account to a Guru’s trading action), which is roughly the number of clients a middle-to-large brokerage firm might have. This is a very profitable community. Within very broad limits, the larger the network gets the more profitable it gets.”
I made what seemed the obvious observation that at some point in all this following and copying, traders will be stepping on each others’ toes. Whether you regard crowds as wise or panic driven, they certainly can gobble up arbitrage opportunities. In reply, Assia asked me to “imagine all the limitless possibilities a person possesses to create his personal people-based portfolio. A portfolio consisting of numerous Gurus and Traders a user copies and by being able apply a wide variety of parameters to slice and dice the trader rankings every user can create a different set of traders that indirectly invest his funds. We can easily compare this to different trading algorithms that are defined by the risk appetite, investment strategy and portfolio breakdown of every trader chosen to be added to people-based portfolio. Adding to this is a $4 trillion dollar daily market around the globe, which would require substantial trade volume on eToro to affect any of the markets.”
When I asked him about the future prospects of eToro, he told me that when they got started, in 2007, he used to say that in three years they’d surpass Lehman Brothers. Of course that particular measuring stick disappeared somewhat more quickly (or, as Assia puts it, “that happened well ahead of schedule,”) and now he tells people “that in three years we’ll reach the size of Goldman Sachs.”