In the aftermath of the financial crisis, we've all discussed the effects of balance sheet decline and lack of fixed income liquidity ad nauseam. This same period also has brought us an explosion in fixed income electronification. To date, electronification in fixed income has largely been about fixed income electronic trading platforms. This “gold rush” of fixed income electronic trading platforms also has resulted in a flood of new information in a market that had already been subjected to an excess of information. In effect, a by-product of greatly reduced fixed income liquidity is a burdensome overload of information.
More interesting is the response of many participants to this overload. In the face of decreased fixed income liquidity, they’ve proactively sought to limit the inflow of information by reducing the number of counterparties with whom they communicate, especially in the pre-trade part of the trade life cycle. Essentially, diminished liquidity has produced the unintended consequence of market participants decreasing the flow of information at hand, which exacerbates the problem of diminished prospects for finding liquidity.
Despite these trends, a fundamental undercurrent of change is afoot in the way we consume information. The most nimble market participants are already starting to take advantage of the development of new technology solutions that address these information challenges and will provide a long term competitive advantage versus peers. In Part I, we will explore the largest and longest running information challenge and two new ways to solve this old problem.
The majority of fixed income information is sent out in unstructured runs messages. Most participants are overwhelmed by the resulting information overload. You probably don't need to poke around too far to find a buyside firm that is decreasing their dealer coverage due to runs overload. It is probably the number one issue we hear from prospects. The real problem is that runs information is immense and does not interact with other data. Moreover, reading every bid and offer in the market is neither scalable nor conducive to generating meaningfully more liquidity.
Based on what we process at FixtHub, there are nearly 1 Million bids and offers in fixed income runs -- every single day. This doesn't include the thousand or so bid list items and each corresponding bonds' price talk. Once you add quotes, market color and other information, the amount of unstructured data that is available is truly staggering. This information has been historically consumed and analyzed by people.
While some market participants seem fine with the onerous daily task of proactively searching large volumes of incoming runs messages, we continue to hear that:
- Searching for information is overwhelming
- Axes of interest are often missed
- In order to manage better the consumption of information they receive, market participants are reducing amount of information they take in
While the solution to the problem of diminished liquidity obviously isn't to decrease liquidity further via information reduction, the real issue is how to increase the efficiency with which available information is processed and consumed. We see two choices to materially improve the issue of voluminous and unstructured data:
There are several message parsing solutions in the market. Generally, they do a very good job at structuring the data, with accuracy rates above 90%. In short, these solutions take your unstructured runs messages and put them into a more consumable format. Having purchased and built parsers, we understand the issues and complexities incredibly well.
Generally, these messages may be made available via a feed or a software tool. If via a feed, then you will be required to spend money to make the feed valuable by integrating this information with other data. Reference data and associated project costs can be expensive in fixed income, so this is usually only an option for the largest buyside/sellside firms with large technology budgets and the long timelines that go along with it.
If a software tool, then some may offer out of the box functionality via fixed income reference data, easy integrations and other value added functionality. A software tool with a parser will decrease your time and investment to derive value.
In August 2015, Project Neptune launched as a utility to boost market liquidity. This initiative structures dealers' axe data into a more easily consumable format for the buyside. The data is available via user interface or FIX feed into an EMS/OMS. According to recent reports, Neptune has been adding dealers, which is a good thing. The best way to take advantage of this information is to leverage its API into an internally built or 3rd-party platform. Look for Neptune to continue adding content from dealers and expanding asset classes.
The need to process unstructured information has always been a problem in fixed income. When liquidity was ample, most trading desks (buyside and sellside) were ok with missing some trades. Now that liquidity is diminished, firms are more likely to feel the need to capitalize on every opportunity. Structuring data in a more consumable format allows you to extract valuable information more readily and thereby leads to more trades. There are many firms already able to access better liquidity via message parsing and Project Neptune feeds – expect this trend to continue.
Part II explores the current fixed income electronic trading landscape, the macro trends and some opportunities.