Super-Human Credit Trading: Automating Workflows and Run Generation

Credit Trading, AI-Powered Trading Alerts, AI and Data Analytics, Credit Trading

At Deep Market Making (Deep MM), our heads are constantly buzzing with new ideas about how to utilize our technology.  One of our core goals is to free up time and mental energy for professionals across the credit universe, and we are actively working towards that purpose.  To that end, we are rolling out new features over the next several weeks (alerting, yield and spread-based pricing, integration of new macroeconomic data sources) that will represent massive steps forward in functionality for our users.  

As a sneak peak, our first test of the new model shows a ~15% improvement in our pricing data.  We couldn’t be more excited to unveil our new products to the world and see what you think!

In the interim, we thought it would be useful to show one of the many ways our users are finding Deep MM to be transformative in their daily workflows. 

Case Study: Sell-Side Runs Automation

Some Background

As a former sell-side market maker, I can personally attest to the extensive time commitment involved in sending out daily runs and market updates.  Historically, the process has been extremely manual (and sometimes stressful!), involving checking TRACE, street updates, market movements and any number of other variables.  The goal, of course, is to send out the most accurate market possible on a given corporate bond – and the downside of being wrong (or off-market) can be extremely costly.

In my personal experience (and heaven knows I wasn’t the best at sending out runs) – over time, this led to some predictable consequences in the way I would send out markets.

  • I sent out runs early – starting at 7 am, on the days I showed up on time – and late (after the close).

This gave me enough time to do the runs I was sending out properly (usually finishing up 1-1.5 hours later), while still giving me time to focus on my bigger goals during the trading day (the bigger “axes”, or situations I was working on, as well as managing my risk positions)

  • I sent out less runs and updates than I would have liked to

Clients love (and need) fresh market updates from traders.  It gives them current context on positions and helps with the execution uncertainty associated with executing on a “stale” market.  In an ideal world, a sell-side trader would be sending out updated runs as frequently as possible…it helps drive trading volume, market presence, and P&L.  However, any sell-side trader would tell you that a given trading day is a balancing act of prioritization – and sending runs can often fall to the low end of the “return on time” spectrum depending on the relative “busy-ness” of a given day

  • I generally sent out runs covering a relatively small slice of my trading universe

Sending out updated markets on bonds is generally a sliding scale of difficulty.  At the easiest (and least risky) end are the liquid “benchmark” securities for a given sector or credit.  The market for these bonds are typically well-defined and help frame context for other quotes in a given name (or sector).  At the opposite end, one would historically find the illiquid or “off-the-run” bonds – generally thinly traded and rarely quoted.  Perversely, one could argue that quoting these bonds provides the most value to the client base in valuing a given portfolio and being able to act on market inefficiencies – but again, sell-side trading is often an equation involving a “return on time” component, and the return on sending markets on “off-the-run” bonds can often be sporadic and unpredictable.  As previously stated, I wouldn’t categorize myself as a diligent runs sender, and thus I generally stuck to my benchmark runs – and in my experience, I was not alone in following that system.


How has this changed with Deep MM’s technology?

With our “Runs” function, sell-side dealers can send out accurate markets on hundreds of securities in minutes.  Our “percentiles” function allows users to set the bid/offer parameters on a broad (market or sector-wide) basis, as well as on an individual security level where needed.  One of the sellside dealers using our platform has automated the entire process, with a junior trader putting together runs on EVERY bond traded by the desk at once (over a thousand securities) – then sending the runs to the individual traders to check over and send out to the client base.  

In one fell swoop, this dealer (using Deep MM’s technology) has solved all of the mistakes I was making in sending out my runs.  They are able to send as many updates as desired during the trading day, covering their entire trading universe, and feel confident in the levels sent (due to the accuracy of our model).

As stated, one of our core goals at Deep MM is to free up time and mental energy for credit professionals to focus on the goals of their individual role – for sell-side traders, life is about trading volume, mind share and market share, and P&L.  Our “Runs” function saves innumerable hours for clients using our product and even makes the utility of the runs sent more useful for clients (by broadening the coverage universe of securities).  That helps traders focus on what they do best – manage core “axes” and risk, and drive P&L for their firm.

What’s next?

This case study is just one example of the impact Deep MM’s technology can have on the lives of credit professionals.  We are excited to tell (and show) you more about what is coming next, just click the schedule a demo link below and we will look forward to hearing from you shortly!

If you would like to learn more about AI in corporate credit, please email us at sales@deepmm.com and we can set up a time to talk.