Monday, March 5, 2012

Can mathematics improve bond market efficiency? Introducing the Magenta Line

Greetings reader. You have stumbled on the obscure blog of yours truly, Peter Cotton, Chief Scientist at Benchmark Solutions. My company uses applied mathematics to provide transparency in the bond markets, and this contains some occasional musings. I frequently find myself losing track of papers, links and so forth, and this makes it slightly easier to find them again.  

This first post is intended to explain why I get out of bed. The topic is the inefficiency of corporate bond markets and what a little mathematics and technology can do about it. The market possesses a "low level of automation", as one author put it delicately, and has for over a century. It is almost as inefficient as the municipal market, the only real point of comparison. In both markets we find some delightfully understated descriptions of the problem, by the way, such as that offered by Harris and Piwowar (2004):
"Our results show that municipal bond trades are significantly more expensive than equivalent sized equity trades"
Their finding is consistent with Ang et al, who are rather more forthright in their proposal to fix the municipal bond markets. Transaction costs in municipal markets comprise an annual $30 billion wealth transfer, according to the authors.The corporate bond markets are not much better. Very roughly speaking the bid-offer is on the order of a point, whereas the bid-offer on treasuries is well under a basis point. Equity markets and foreign exchange provide equally dramatic, if not more dramatic points of comparison for as former SEC Chairman Arthur Levitt put it:

 "The sad truth is that investors in the corporate-bond market do not enjoy the same access to information as a car buyer or a home buyer or, I dare say, a fruit buyer."  (WSJ 9/10/98)
Studies, such as Bao, Pan and Wang consider the relative liquidity of corporate bonds over time. The long view is provided by Biasis and Green who summarize the lack of progress. Their longitudinal survey of the corporate bond markets over the last century demands reflection:
Our finding that transaction costs in municipal bond market were lower in 1927 than at the beginning of the twenty first century is not unlike the finding, by Rajan and Zingales (2003), that equity markets were more developed in 1913 than in the major part of the post-World War II period. Both suggest that the interaction between economic agents does not always naturally lead to a trend of gradual improvement in financial architectures. 
So much for the invisible hand. In the absence of trade reporting participants have little idea where the market should be, and even post-trade transparency has its limitations. Hotchkiss Goldstein and Sirri consider the partial introduction of reporting for corporates (BBB rated bonds) and use this as a controlled experiment to test for decrease in transaction costs. 
Except for very large trades, spreads on bonds whose prices become more transparent decline relative to bonds that experience no transparency change. However, we find no effects of transparency for very infrequently traded bonds. The observed decrease in transactions costs is consistent with investors' ability to negotiate better terms of trade with dealers once the investors have access to broader bond pricing data
Perhaps it goes without saying that in a market with "low levels of automation" the revelation of traded prices for closely related securities does not necessarily translate into transparency for most participants. Trade reporting for other bonds might even increase the information asymmetry for those that rarely trade, such as this bond. Some participants have access to analytics that others do not possess.

Post-trade transparency provided by FINRA TRACE

A point of speculation is the impact on transaction costs of pre-trade, as distinct from post-trade transparency. The distinction is apparent, I hope, in the next diagram where the trades for a bond dry up but the so-called Magenta Line anticipates their reappearance.

Pre-trade transparency provided by Benchmark Solutions

In the findings of Hotchkiss et al are extrapolated, we may expect a reduction in trading costs for illiquid bonds such as the next one. Notice how the dealer trades in Jan and Feb are predicted with uncanny accuracy, using information from other bonds, rates and credit default swaps. Participants are certainly in a position to negotiate better terms of trade.

They should also be immunized against some really cynical trading. Notice how the customers are buying a floater at par, perhaps encouraged by the "floaters trade at par" fallacy.

Customers buy a floater at par
A few more observations about liquidity might be in order. Liquidity can be a function of age. New issues often dry up, as with this example:


Liquidity in a new issue drys up

In other bonds, liquidity can dry up and then re-emerge

A dry spell

Notice how once again, the Magenta Line anticipates the level at which trading recommences.

I might include a few more examples here at some point, but I'll wrap up by saying that the distinction between post-trade reporting and pre-execution intelligence is not lost on Medy Agami, industry analyst who was kind enough to write this:
From our perspective, this innovation by Benchmark Solutions has brought today's US corporate bond market to a pre-trade transparency era. 
Thanks in part to Celent's coverage, the news appears to be getting out.