Is there a point at which information processing is so rapid and complete that capital markets could as a result become less (rather than more) efficient? It would seem, on the contrary, that the more quickly, the more comprehensively, investors can act upon new information, the more efficient the market pricing mechanism becomes. But this perspective has never been tested in an environment in which massive amounts of global data can be factored into execution virtually instantly. For reasons discussed below, Google gives us food for thought.
Much attention has been paid lately to “big data” – industry jargon for the processing of large-scale analytic information – a concept that has taken on new meaning as public data, accessible online, has grown with the Internet to enormous depth and breadth. As the big data segment, still nascent, promises vast potential with new tools and systems to facilitate pattern recognition and other intelligence in consumer behavior, economic forecasting, and capital markets, a new competitive criterion may emerge in the private sector: the ability to absorb and utilize the most available data the fastest. Taken to an extreme conclusion, a concentration of “brainpower,” lacking a better term, may emerge, that at least in the case of capital markets would not be dissimilar to a concentration of capital.
Enter into the picture, Google. A job posting that was recently picked up and commented about by Business Insider, and pursued further by Valleywag, solicits experienced bond pros to help staff up a new trading platform at the world’s dominant online search organization. In itself, this would not be sensational, because a bond trader or two can be easily explained by a cash-rich enterprise seeking to extract yield in the money markets. Recent insights into Google’s thinking, however, such as its decision against starting a hedge fund and its belief that it could probably “predict the stock market,” give some of us an excuse to pause and ponder further. Any sophisticated trading unit housed inside Google – the global leader in public data processing, if you would look at Google that way – would have access to information and tools that few if any market participants could rival… a notion that was apparently considered by the team.
The example of Google is used here for convenience of illustration; we could imagine any entity (existing or hypothetical) that may rise to big data prominence. The consequences of such a group’s financial markets activities, based on expertly processed information in an environment equipped to handle massive data, should probably not be dismissed without question. And the question that I struggle with is not as much one of unfair profit opportunity, because legality should address this issue and hard work and a proprietary system is fair game in a competitive capital market. Rather, I wonder about systemic risk.
We have already seen what may be the result of fund flows that are pushed and prodded by concentrated pools of capital – in the formation of bubbles, the popping of bubbles, and an increase in market volatility. In theory, an efficient market exists most perfectly when there is a fragmented base of participants with diverse perspectives. By extension, as such diversity diminishes, efficiency becomes less perfect. In the extreme scenario – a market of one (or, more plausibly, a completely unified market) – the theoretical inefficiency is total. Now, is it possible that what is true for capital concentration is also true for information concentration? If a narrow group should come to dominate information processing in the manner described, would this not similarly result in a reduction of market diversity? Would such an environment not encourage those less “informed” to emulate the trading of investors “in the know,” resulting in even more concentrated centers of influence?
The questions are conceptual, and lead to other conceptual questions. But with the “big data” movement well underway, the currently conceptual might take on more practical undertones soon enough. If not Google, then some other entity or investor class could lead us there. At a sufficiently high level, capital and information become synonymous. As we begin to test such limits, we might ponder these issues further in our market studies.
