Looking through offering materials for a given security is fairly complex work, not just in reading between the lines in so-called plain language documents – an endeavour that tests patience – but also in modelling the risks against what is already known. Offering documents, frankly, are blank slates – amenable to optimistic and pessimistic projections alike.
Analytical abilities count, suggests a new working paper from the U.S. National Bureau of Economic Research (subscription required). In “Trading Complex Assets,” Bruce I. Carlin and Shimon Kogan. But complexity can tax them.
What is complexity. Here it’s the difference between a government bond and corporate bonds with embedded American options, credit default swaps, collateralized debt obligations, and other expressions of financial engineering.
“Even though complexity may increase uncertainty, this is not its most salient feature,” say Carlin and Kogan. “Complexity makes it difficult for market participants to forecast the essential inputs required to value the asset in the first place.”
How to test that? Like all empirically minded professors, they set up a classroom experiment – the proverbial “laboratory setting.”
“Participants were asked to evaluate the price of certain assets and were then given the opportunity to make trades based on their information. They each participated in fifteen distinct periods, each of which was composed of two stages. In the first stage, each participant was given information regarding several portfolios composed of four assets and was asked to submit their best estimate of the value of a particular asset included in these portfolios. Following that, in the second stage, participants were randomly paired and were given the opportunity to trade the asset through a well-defined bargaining process.”
That’s much better than simply watching James Cramer every night. In stage one, we have the IPO. In stage two, we have real, live price discovery – of a sort.
“Our results show that complexity affected both the liquidity and price volatility of the assets traded in our experimental setting. The frequency of transactions was significantly lower and the payoff asymmetry was significantly higher when the computation required was more complex.”
That seems a lesson frenetic CDO traders should have learned, especially when their colleagues across the prop trading desk were shorting what they were buying. And now, an argument for plain language:
“Importantly, though, these findings impacted the trade surplus generated in each round: making the required computation more simple increased the trade efficiency by 11% (from 73% to 81%).
This was more pronounced when there were more bidding rounds allowed. Whereas efficiency rose from 73% to 84% for the simple treatment when the number of rounds increased from one to three, efficiency remained unimproved in the complex treatment (72% versus 73%). This implies that while on average some participants enjoyed an advantage over their counterparts with complexity (higher payoff asymmetry), the aggregate surplus tended to be lower.”
It’s an interesting experiment, confirming, unfortunately, what many already expected. “Asset complexity may have asset pricing implications and may drive how assets are managed and traded. For example, during the recent financial crisis, it is an understatement to say that many financial models failed. One of the driving forces was the inability of key market participants to value assets that were serially securitized. Following this, the challenge in valuing toxic assets and managing credit default swap obligations worsened the ability of the market to right itself and avoid illiquidity spirals.”
As the regulators say, a little sunlight is a good thing – not just for investors, but for orderly markets.