Before we get into some of my thoughts on things like natural language generation, application interoperability, and Taylor Swift (naturally), I’d like to highlight three outstanding (and long) features and an opinion piece that we published this week.

First, Max Bowie spoke with a swath of tech companies to learn about how they go about putting their software through rigorous testing environments. As mission-critical elements of trading firms’ architectures adopt new technologies such as cloud, suppliers must test their products rigorously to minimize service issues. As Max writes, enterprise data services, in particular, must be bulletproof, as they interface with many other systems. It’s an interesting read about a subject that isn’t much discussed—in the end, everyone just needs the product to work, but to get to that point, a lot of back-end work must first be done.

Second, Reb Natale spoke with several quants and physicists about agent-based modeling. AB modeling has taken root in the fields of theoretical physics, military operations, public health, and with ride-sharing apps like Uber. Finance is joining the fray. This technique breaks away from stochastic, or mathematical, varieties of modeling and forecasting, and is being used to conduct stress-testing and risk mitigation, as it seeks to simulate the rational and irrational behaviors of agents in their environments. However, as Reb writes, the method is not without criticisms. It’s a fascinating subject.

Third, Joanna Wright has been our European market data regulation guru, and for the December issue of WatersTechnology magazine, she looks at how market participants say Esma’s latest efforts to address the cost and complexity of market data fees don’t address the root of the problem. Even though Jo is focused on European market data issues, it’s a story that anyone from any jurisdiction can understand. 

Finally, while I wrote about the proposed S&P-IHS Markit deal last Sunday, Max provided his thoughts on the subject on Monday, and he’s far more intelligent than I am.

Out of Many, One (or Maybe Two)

Application interoperability can take on many forms. There are the containerization/operating system tactics undertaken by the likes of OpenFin, Cosaic, and Glue42. Then there are companies that, through acquisitions, try to stitch together full front-to-back workflow offerings, such as the Intercontinental Exchange, SS&C Technologies, and even State Street. You also have more closed off systems that allow limited interop with other systems—looking at you, Bloomberg.

Now, there are zealots in this field who might argue some of these examples are not instances of application interoperability. The fact is, though, that every bank, broker, asset manager, hedge fund, exchange, and vendor have to put interoperability on the table to some extent, thanks to—and here we go again—cloud adoption, API adoption, and open source adoption as they look to utilize emerging technologies including—but not limited to—automation tools, big data analytics, machine learning, and natural language processing.

At the start of this year, I wrote a column titled, “For the App Interoperability Movement, 2020 will be a Big Year.” Sure enough, even during a pandemic that split up IT teams across the globe, interoperability was still an important driver inside capital markets firms of all stripes this year.

I once again bring up interoperability because of a story that we published about the Royal Bank of Canada on Friday. Max Bowie reported that the Canadian bank is in the process of building a new broker and wealth management custody platform, which will begin beta testing in the first half of 2021, and which will be live by the end of next year or early 2022.

What I found to be most interesting was the sheer number of vendors that RBC is using to build out this solution, which will run in parallel with its RBC Black platform. For the to-be-released offering, RBC is incorporating tools from Salesforce, Deloitte, Mulesoft (owned by Salesforce), Addepar, Docupace, and Beta Systems by Dain Bosworth (which, to be fair, is owned by RBC). RBC Black was built using applications from CircleBlack, Riskalyze, MoneyGuidePro, and Redtail. That’s 8-10 vendors, depending on how you want to look at it.

Now, this project is mainly geared toward wealth management clients and not the wholesale capital markets, but the bank’s capital markets group has embraced interoperability initiatives. As you can read here, RBC is working closely with OpenFin on containerizing its internal order management system. These efforts are leading the institution to find new and interesting analytics vendors to bring into the fold.

As interoperability initiatives progress, taking a “best of breed” strategy will become more viable. The problem, though, is that at the same time, the industry is seeing massive consolidation, and that is only going to continue.

So this situation creates something of an existential question—which, if we’re being honest, has always been the existential question: Do you want à la carte, best-of-breed solutions, or do you want full-service, front-to-back offerings? If I’m a technology manager looking to plan ahead, I don’t know what the right answer is. I think deep down in their hearts, trading firms want best of breed, but need a one-stop-shop, front-to-back solution.

I’d truly love to read your thoughts on the subject: anthony.malakian@infopro-digital.com.

From NLP to NLG

I recently wrote about how natural language processing is taking on greater importance in the world of capital markets research, which is something that I’ve written about previously, as well. This year has been the year of NLP. There are many who hope that soon it will be the year of NLG, or natural language generation.

Jo Gallagher recently wrote a deep-dive feature (about 3,500 words) on how NLG is slowly seeping into the capital markets and what the challenges are. Banks like RBC and Morgan Stanley, as well as data giants like Bloomberg and Refinitiv, are experimenting with and rolling out NLG tools. While there’s a ton of promise in this subset of AI, it’s not a plug-and-play tech in the capital markets because of how highly regulated financial services firms are.

There are some very good use-cases in the story, but sources were also careful to note that this is not something that is going to be set loose at a bank or asset manager, especially when NLG is combined with machine learning.

For example, in June, OpenAI released the latest version of the Generative Pre-trained Transformer, or GPT-3. While the technology is impressive, Paul Tepper, an executive director in Morgan Stanley’s technology division, says that heavily-regulated financial institutions need to be mindful of the risks associated with advanced pre-trained language models.

“I would be pretty shocked if any large company let one of these things loose on their customers, or in any other specific case, because you really don’t know what it’s going to do, like it is a probabilistic model. And it is basically just generating the next word based on things it has seen already, so it is quite unpredictable. We are not using anything like that,” Tepper says.

There are a lot of snake oil salesmen when it comes to AI in the capital markets. You’ll often read press releases where a firm is touting that they’re using AI, but it’s glorified data science. NLG will be the next form of that type of overstated marketing. We’ll try and weed out the fakes from the real innovative tools, but be on guard for overblown NLG—as our sources note, it’s quite challenging to release regulated tools in this space.

The Cultural Zeitgeist

Before I leave, I wanted to write about something different—music. When talking about today’s pop stars, it’s easy to be dismissive. Much of today’s music is auto-tuned to the nth-degree, making humans sound like robots. Additionally, so many pop stars follow a cookie-cutter regiment set by executives at Universal, Sony, and Warner music groups. And the type rock music that defined the 1970s, 80s, and 90s hasn’t been as popular in the 21^st century.

As a result, it’s easy for us older folk to stop paying attention to new music and dismiss today’s talent.

Many of my friends are musicians who have been in bands at one point of their lives; most of my friends play a musical instrument. When I bring up Taylor Swift as being one of the greatest musicians of all time, her name elicits sneers and heavy eye-rolls.

Now, I’m not here to provide album reviews, but let me just say that one of the most impressive things I’ve ever seen in the world of music is Swift producing not one but two absolutely-sensational albums within less than five months, and doing that during a pandemic that stunted the release of new music (as well as new TV shows and movies).

And just as impressive, she really leaned into the overall feel of 2020; a rollicking dance-party album (i.e. “1989”) would not have fit the moment. Instead, she reached back to her country/folk roots, showing dexterity in her sound and timelessness in her lyrics.

I’m now 41 years old. I do not ever want to be the old guy who only listens to music from my teenage and college years. Yes, I’m frustrated that bands like Nine Inch Nails, Nirvana, Soundgarden, and Metallica are no longer in the mainstream, and thus it’s difficult to find the new rock and metal bands of today. But I’m also impressed and moved by artists releasing new music, whether that’s Swift, Billie Eilish, Lady Gaga, BlackPink, Ariana Grande, or any number of other artists whose names I don’t remember, but whose songs I enjoy. (Right here is where Reb is saying, “WHAT ABOUT DRAKE?!” Drake sucks. I don’t mind writing that.)

This is all to say that we should keep up with new music, movies, TV shows, styles of art, books, political commentary, and sports. Sure, you’re likely to hate much of what’s produced, but the other option is to become calcified and dull. Why make the choice to be mentally dead? Expand your horizons, and start by listening to Taylor Swift’s new albums, “folklore” and “evermore.”

The image at the top of the page is Jean Baptiste Camille Corot’s “Pond at Ville-d’Avray” courtesy of The Cleveland Museum of Art.