We shouldn’t expect a revolution in FI market structure, but the conversation at Fixed Income Leaders Summit continues to evolve. I loved Brad Levy’s observation that many solutions are in the late innings of delivering bond market efficiency, but that we are in the early innings of those solutions converging. Convergence and combination carried through the panels and sidebars. Here are my takeaways…
- There is a reason it is not called the IG e-trading Leaders Summit. FI is not just about liquid corps.
- Munis got more airtime this year, though swaps got even less than in the past.
- There is skepticism around a single viable solution across “generic” (govy, IG credit) and more nuanced (EM, HY, CLO, muni) assets. Blindly applying the same ideas/tech for CLOs and IG corporates is just as naïve as expecting an equities dark pool to work out of the box for HY.
- Protocols/concepts like Request for Market and algo trading often bleed from 1 market to the next, but success is not a given and adoption is slow. RFQ remains the main protocol for a wide swathe of govt market.
- Margin optimization through repo/derivatives clearing will further tie solutions across markets. Listed markets exchanges were key sponsors this year, and with new products, they are looking to disrupt not only each other.
- Data + data = more data. Data + analytics + good tech to clean/splice it = something actionable.
- The right combination of data, analytics and tech enables automation (and competitive advantage), even in traditionally difficult markets. One manager trades 97% of their orders across 1MM CUSIPs as low- or no-touch; that market is currently just 15-20% e.
- The definition of “actionable” varies by asset class…could be intelligently informing an algo, or directing a phone inquiry to a single dealer.
- “Portfolio trading is old news.” Traders need pre-trade analytics and broader elements not available on individual venues (TRACE, trade history) to help optimize PT for situational needs (spread, duration, cost).
- Venues, exchanges and dealers are promoting traditional and “alternative” analytics across data types. Things like ETF detail, peer analysis, curve relative value, and “nowcasting” can help to enable more targeted inquiries.
- Most managers are working off the same set (and broadening amount) of core FI data, so analytics and interpretation is what generates alpha. Many (myself included) would argue that data is more commoditized than it should be, with vendor/buyside abuse and oversharing.
- No one system can do it all…neither horizontally across all markets, nor vertically through the lifecycle of an order. Clients are increasingly demanding robust collaboration across the ecosystem, from ideation to settlement.
- OMSs need to be more open with EMSs. Buyside is seeking to combine best-of-breed solutions, and very few consider their strong OMS to be a strong EMS.
- Dealers need to upgrade communication. Direct client connectivity is still lacking, and there is no standard for counterparty reference data.
- Vendors/venues need to get comfortable with buyside accessing capabilities via something other than their own screens. “Double dipping” (charging for data, then also for access to that data via API) is especially unpopular.
- Buyside-to-buyside. For all-to-all to continue growth, the buy side needs to behave better and be realistic on execution levels.
Lessons/advice from panelists
- There is no one-size-fits all for trading tech. Large and small clients have different needs, and needs differ by asset classes
- Start small with a POC and work up from there. Accept good, don’t demand immediate perfection.
- Having a tech/market structure expert on the trading desk both facilitates implementation and reduces the desk’s fear of the unknown
- Pick a vendor that is nimble and futureproofs you. Market dynamics and desk needs can change.
- Get assurance upfront that your vendor will collaborate with the other tech you have/want
- Build/buy is a spectrum, and most asset managers cite resource/budget limitations as they reconsider building FI capabilities inhouse. Buying does still require some tech work, but less of it, and resourcing varies widely depending on the type of vendor.
So here we are…
- Coalition Greenwich’s recent EMS study got a lot of airtime, where only 13% of the asset managers surveyed said they used an EMS for FI, yet responses from every manager pointed directly to how much they needed one.
- A fixed income EMS aggregates market data, routes orders across destinations, and connects to upstream/downstream systems.
- A great FI EMS incorporates proprietary and third-party analytics, respects FI markets’ microstructures and product nuances, and uses APIs to truly connect (as opposed to just calling up other screens), listens and acts. We just happen to be one of those.
