CEO CHAT: Jeff Stein, Xinthesys
Buy versus build? It’s the age-old technology question on Wall Street.
In a modest commission environment where both the buy- and sell-side are making due with less and technology needs only leaping in bonds, should firms build their own technology or employ specialty vendors to help them? While it seems a question more suitable for the accountants and financial officers, Xinthesys Founder & Chief Executive Officer Jeff Stein discussed the benefits of buying versus building software, data workflow platforms and how to manage consolidated audit trail data.
TRADERS MAGAZINE: What some of the benefits are for organizations to consider a data workflow partner vs. building a solution in-house?Jeff Stein: Over the years the financial services industry has had a love-hate relationship with the concept of build vs. buy. Years ago, many believed that having the best software development shop in-house gave you a competitive advantage. In recent years, many have successfully deployed very sophisticated software platforms by external vendors particularly in commoditized functions. This buy vs. build pendulum fluctuates every two or three years and particularly given the steady decline in trading revenues over the years, many banks and brokers are reevaluating where and how they spend their tech budget.
Should you be spending dollars in areas that provide no inherent competitive advantage or that are not revenue generating? There is a trend among financial institutions to look for ways to reduce spend in non-revenue generating areas to better allocate resources for revenue-generating areas like trading infrastructure, algorithms, pre-trade analytics, new other product offerings.
When it comes to regulatory reporting such as CAT, why build a platform that will generate zero in revenue for your firm? On the contrary many firms are looking to partner with the right firm to reduce risk for post trade workflows like CAT.
Partnering with the right firm for regulatory reporting, at half the cost and half the risk, just makes fiscal sense. Regardless how large you are and how much tech budget you have.
TM: What key characteristics should CTO’s consider when looking for regulatory data platforms?
Stein: That’s a tough question because in our experience it’s not only the CTO who is involved in a decision for a platform. These days, COO’s, CRO’s, CCO’s and Heads of the Trading are equal stakeholders in these decisions and it’s important to recognize different perspectives and joint responsibility for success in these types of endeavors.
The upcoming CAT implementation has caused many C-suite executives to rethink how platforms are implemented. The following elements are top of mind:
• Delivery risk
• Ease of the implementation
• Impact to the organization
• Access to data
Delivery risk is an important factor as many of us have seen large projects fail, use significant amount of resources and if implemented, not hit the mark of the initial desired value. Which is why in our view smaller is better. And as we know from software implementations, agile surpasses long delivery timelines. This has been proven successful in many disciplines.
So, when looking at implementation some of the key things to identify are:
• How easily the platform integrates into the existing environment?
• How quickly can it demonstrate value and proceed on the right track?
• How easy and quickly can changes be made? (i.e. Can your requests be measured in days, weeks or months?)
• Are you dependent on a vendor to make changes or can you exercise self-help and make changes yourself?
In our view, easier implementations are made up of systems that leverage existing firm’s assets and systems with short delivery timelines that are measured in days not months.
Assessing the impact on the organization on a program like CAT is complex as it cuts across many disciplines: the front office, operations, compliance and technology. Each of those areas will be impacted. Therefore, for firms that are already wrestling with existing programs in regulatory or trade processing, these parts of the organization will be stretched even farther. From our perspective, selecting a vendor that has best practices in this space, can significantly reduce or even eliminate the impact on the organization.
All that said, it’s always about the data. How easily you can access the data to answer questions or respond to queries by internal and external stakeholders should be a primary factor in your decision.
Without getting too technical, approaches generally fall into two categories, a “layered approach” and an “integrated approach.”
In an “integrated approach” a system integrates and is fed by an existing data warehouse for trades to report. For many firms can be quite challenging. The reason for this is that most data warehouses in banks are specialized and undergoing continual change. Moreover, most have been engineered and re-engineered over decades. Keeping this integration working requires constant care and feeding. Most solutions fall into this category.
In a “layered approach,” a system integrates directly into the trading streams and feeds a data warehouse. This approach is not dependent on the firm’s data warehouse for trading data but rather the source for it. It may be necessary to pull static data such as codes, accounts, classifications but this minor dependency too can be eliminated in some systems so as not to create a dependency on day-to-day basis. This is the approach we have taken with ADEPT and why the integration and implementation are far easier.
It is from this perspective that your CTO’s and technology architects can guide you on the merits and difficulties with either of these two approaches at your firm.
But the bottom line is this: You must be able to respond to regulator queries within a 24 hour period. Therefore, any solution you chose should put the data access in your hands and not strictly your vendors.
TM: How does Xinthesys’ solution differ from other big data systems?
Stein: Our solution which is named ADEPT and stands for Agile Data Engineering and Processing Technology is both a software platform and tools to empower data scientists, analysts and technologists to quickly create and the deploy data engineering pipelines.
We have and/are using this platform to create data workflows for the CAT, legacy OATS end-of-life, Rule 605/606 and Bluesheets reporting with many others coming over the coming months. Our vision is to allow anyone that may or may not be a developer but that understands data to create any data workflow imaginable and to run those workflows at scale in a public or private cloud platform or on self-hosted infrastructure.
We believe that we have a unique approach and platform that will bring ”Agile” data engineering to the post trade environment.
With ADEPT, it’s all about empowering users not overpowering them or holding them hostage. Our vision provides those tools in that technology that will enable any firm to reduce the risk for data intensive operational needs.
TM: What is Xinthesys doing to attract customers as the deadline for CAT nears?
Stein: For starters, we are offering a no obligation informational and assessment session to any firm that is interested in discussing some of the best practices in operational efficiency around trade reporting. Specifically, around some of the challenges that the CAT brings and managing the OATS end-of-life. These sessions are designed to understand better the operational complexity of an organization and what challenges or potential ‘gotcha’s’ that will be factors in becoming compliant for the CAT.
Secondly, we are excited to introduce our new program called CAT Lite. It is a trimmed down version of our larger offering for CAT. By design, it is a low-cost, super easy implementation that enables a firm to set up, test and report on a daily basis to the CAT within as little as two weeks and it is available today.
CAT Lite is targeted for those broker-dealers, small or large that find themselves either in a risk position, uncertain about the direction they should take or would like to take an evolutionary path towards managing through the CAT deadlines. It is a solution which bypasses the need for a large upfront expense, a complex implementation and tying up precious firm resources. It leverages your existing trading infrastructure, connectivity and data.
There are also incentive programs available if you are on some of the larger FIX networks, such as AUTEX Trade Route (ATR), by Refinitiv.
TM: Aside from CAT, who are you targeting? Is Xinthesys focused only on financial services?
Stein: Yes. For the foreseeable future our focus is on our core competency, the financial services industry. More specifically around pre-and post-trade. That said, we have been approached by organizations that service other industries such as research, government, healthcare and insurance that are interested in taking our technology and applying it to those industries. We will gladly license the technology but leave it for others to develop those markets.
Later this year we will be releasing a new product that we’re super excited about. We believe it will be a game changer in the quantitative/portfolio analytics and financial engineering disciplines. It is called ADEPT Models™ and is targeted to Buy and Sell side firms engaged in portfolio analytics and trading signal generating activities. We will definitely be sure to keep Traders updated when this becomes generally available.