Any trading technology firm that is co-headquartered in New York City and Austin, Texas, gets my attention, even when it’s been flying under the radar for the past few years. Now, Broadway Technology is taking investment from its clients – including Goldman Sachs – to accelerate its rate of growth. IntelligentTradingTechnology.com caught up with the company’s CTO Joshua Walsky to find out more.
Q: For starters, when and how did Broadway Technology get started, and what does it do?
A: Broadway was founded in 2003 by Tyler Moeller, our CEO, and myself. We both are engineers by training and have advanced degrees in computer science from MIT and Cornell, respectively. Our backgrounds span from high-frequency automated trading to enterprise software.
Prior to Broadway, Tyler and I had been building and trading automated trading models as part of a successful proprietary trading group called KATS. We enjoyed the trading aspect and market dynamics, but being engineers, Tyler and I always wanted to get back to the software and system building aspect.
In 2003, we had an opportunity to split the software arm of our operation from the trading arm. This opportunity came out of a real need for innovative and reliable trading technology as there wasn’t anything like that in the market at all, much less anything that could solve a problem on the scale we needed. With a working system that we successfully traded on for several years, our real-world trading experience, and enterprise software backgrounds, we seized that opportunity and started Broadway.
Broadway has now become the emerging leader in high-performance financial trading solutions for top-tier global banks and hedge funds. We lead the market with innovative turn-key software solutions and our patented enterprise platform, the TOC, an event driven, distributed data-oriented architecture that underpins all Broadway’s solutions. Our software manages hundreds of billions of dollars in transactions daily worldwide for premier financial institutions that include the biggest names in the business, such as Goldman Sachs.
Q: You just took some investment from your customers, including from Goldman Sachs. The company has growing revenues and is profitable. So why is it good to have customers have a stake in the company, and what will you be doing with the funding?
A: Broadway has been profitable every year since its launch in 2003, with revenues having grown more than 50% for each of the past four years and tracking for a fifth consecutive year. We have been taking our time building this company and building it right. Just like building great software, building a great company takes time.
We have been very purposeful and methodological in our approach to building Broadway. We wanted to prove our business model, prove our software development model and platform, prove there was a market, prove that we could attract and hire the top minds in the software industry, prove that we could build a great company with a great culture, and prove that we could do it year over year through times of economic growth and downturns.
With the groundwork and strong foundation in place, we were ready to substantially accelerate our growth to take advantage of an incredible demand for our software. This was the ideal time for Broadway to take on capital to accelerate our long-term global expansion plans and move into a broad range of asset classes since our trading platform is asset-class agnostic.
Q: Can you describe some of the systems that you have built for customers – what range and types of applications? Has that led to any areas where you have deep competencies?
A: One of the strengths of Broadway’s system is its flexibility. At its core, Broadway’s TOC platform is a real-time integration platform capable of connecting disparate systems in an extremely low-latency manner. We use that platform to deliver distributed trading systems that span the globe, run 24-hours a day, scale to multiple server farms, to thousands of servers, and thousands of client machines and users.
That sort of scale is interesting, but the scale I am most proud of is what I call “functional scale.” Functional scale isn’t something you hear many people talk about because it is hard to get your hands around, yet it is one of the most critical aspects of building long-lasting systems that produce real value year after year. Functional scale is the capability of a system to successfully scale to multiple different kinds of applications and uses, especially innovative uses that haven’t yet been imagined. Broadway’s systems achieve tremendous functional scale.
Our customers have been able to build incredibly fully-featured complex and complete integrated trading systems – systems they could not have created before Broadway’s software existed. On Broadway’s platform, our customers have independently built or we have worked with customers to build: automated trading, low-latency arbitrage trading, smart execution algorithms, automatic pricing, market making, e-commerce distribution, automated hedging, position tracking, risk management, trader and counter-party limits checking, credit checking, trade capture and deal processing systems and all the user interfaces required to manage those systems. These capabilities are all integrated together within a single system, and that list of capabilities is constantly growing.
We have the unique advantage of designing the software from day one to be asset-class agnostic, so we’re able to move into other asset classes and develop expertise quickly and seamlessly. We did it when we moved from fixed-income to FX in less than six months. We are doing it again as we enter credit and swaps. To date, our success has been in fixed-income and FX trading front office, the manual side, the fully automated side and everything in between. We have a very deep understanding of those assets but, more so, we have a deep understanding of the trading front office.
Q: What would you say are the advantages of a company engaging Broadway, versus buying a packaged system, or developing in house?
A: One of the reasons for Broadway’s success is that we offer customers “build and buy” solutions. Out of the box we provide a full-service trading system, and some customers stop there and use just that. Others use it as a launching pad to build their own custom advanced capabilities that easily integrate with their existing systems. This allows our clients to target their valuable technical resources at initiatives that maximise their specific market opportunities.
The greatest compliment I ever got was from a traditional trader who used Broadway’s screens every day yet claimed that he wasn’t doing any algorithmic trading. Under the hood he was actually running one of the industry’s premier algorithmic order routing and spreading tools, yet to him it was a natural screen-based interface. We are allowing traditional traders to compete with the black boxes out there in an intuitive manner.
Broadway provides best of breed liquidity aggregation, order routing matrix pricing and e-commerce solutions. This is all available through screens and APIs. We provide software that historically was only available at large institutions with multimillion dollar in-house IT departments.
An organisation that has its own IT or development staff can really open up the power of Broadway’s system. Every aspect of Broadway’s system is API- accessible. Broadway’s customers are able to build incredibly feature-rich, well integrated and powerful trading programs endlessly extending what Broadway provides out of the box. Broadway allows customers to focus their effort on their value-add, their secret sauce, those things that differentiate them. That’s their pricing models, proprietary trading algorithms, risk models, hedging schemes, etc. Broadway takes care of the rest. We do the work day-in and day-out to keep improving those infrastructure and connectivity parts, and keep them up to date with modern changes, ECN API changes, etc.
Q: You have patents on some technologies? What do they cover?
A: We have patents in a number of areas. Some are in general distributed computing; these are not finance-specific and apply to Broadway’s advances in large-scale distributed computing technology. Others are in trading technology and e-commerce as well as algorithmic trading patents.
Additionally, we have new and innovative ideas that we’re currently filing patents on. The foundation of Broadway’s software is an extensible and growing platform. It is built to extend to new capabilities. That growth and forward-looking nature permeates everything we do. Most importantly, it inspires a culture of innovation. As an organisation we nurture that culture of innovation – the capability to innovate is one of the things I love about software. Our employees are constantly innovating on many levels within the distributed computing world and finance world.
Q: Are there any particular underlying technologies that you base your applications on, that you would tend to adopt in a green field development?
A: Broadway’s TOC platform does have an underlying messaging layer and an underlying database layer. Both of these are vendor agnostic and we support multiple underlying database and messaging layers. We also support both Windows and Linux.
As far as how this would change in a green field environment, I respond “not at all.” Great software, fast software, functional software takes years and serious work to build. That software must evolve and mature, and be able to incorporate new technologies as they are invented.
Broadway focuses on building software that can take advantage of new technologies (RDMA, new hardware, real-time platforms, etc) without needing to throw out the existing successful and functional systems. Instead our software allows for a natural evolution into new technologies while still leveraging existing software. This is the exact opposite of disposable technology, which is thrown out and rebuilt on a three year cycle. At Broadway, we build evolving technology that gets better and better over time.
There is a common myth that there is an advantage to “green field” that existing systems don’t have. After all, the grass is always greener in the green field development, right? A well architected system should be able to take advantage of these new technologies without requiring a green field rewrite. In fact, a lot of today’s green field development is destined to become legacy in three years because it is unable to evolve.
The real challenge that Broadway solves is allowing software to evolve to green field without being destined for legacy. That takes serious engineering and simultaneous focus on both the short term and long term. That is something we think about constantly.
Q: And are there any technology developments that you are tracking that you feel will impact future development of electronic trading systems? What might your applications be making use of in a couple of years?
A: One of the developments that we are focusing on is the substantial increase in cores available on affordable machines running standard platforms with broad application support. 24 core machines are now commonplace and economical options even for smaller firms. Bigger machines are also available. These machines fit within firms’ hardware budgets, system administration proficiencies, and have substantial support for a variety of applications and can be readily used in heterogeneous environments. As a result, we are seeing more and more components running in concert on a single machine in shops of all sizes.
In addition, we are seeing rapid evolution of how and where different components are run. Customers are changing which processes and functions run on which hosts and in which layout. They move processes to colo, to in-house server rooms, from New York to London, or even across hosts within a facility. Sometimes they make changes like this on a day-to-day basis. This means having the flexibility to support all situations, making it seamless yet transparent and taking advantage of the best and most efficient communication mechanism for that specific configuration. You want to do all this without having to rewrite or even reconfigure software.
The TOC makes integration and management of large systems easy and seamless. We facilitate big integration of hundreds or thousands of distributed components. Traditionally that meant many machines connected by a network and that meant network-level latency. Now machines are getting big enough where these components communicate at the intra-machine level: loopback, named pipes, shared memory, etc.
For Broadway, this means a couple of things. First it means a focus on more intra-machine communication technologies. Secondly, it is becoming a reality to start integrating different desks or different business functions within an organisation (i.e. actually run a Corporate desk, US Treasury desk, and FX desk on a single real-time integrated platform, or even more forward-looking actually running a bank’s front and middle office together as an integrated real-time system). This is going to open up a new class of complexity for managing disparate information in a coherent and high-performance/low-latency manner.