AI: Seeing is believing
Reading news over the past few years, you can clearly see a large uptick in talk about artificial intelligence. Hardly a day goes by without seeing some mention of AI in the media. Friends, family, and other non-technical mere mortals are increasingly worried when their job will be replaced by AI. Even highly respected scientists and technologists, such as Elon Musk and the late Stephen Hawking have raised their own AI concerns.
The term AI was coined in 1956. Even Dinosaurs like myself took artificial intelligence classes back in university. AI is far from new but its scope has widened to encompass Machine Learning, or even more specifically Deep Learning. This is a technology that uses statistical techniques to extract (aka “Learn”) from data. This is a real technology with practical uses. Machine learning is used in technologies today such as Tesla’s auto-pilot, virus and malware scanning, as well as recommendations you find in Netflix and Amazon. So why all the hype today?
The foundation for most of the recent AI based tools is data. Vast amounts of data. Usually the more you can collect the better your result. Remember back around 2012 when the word Big Data became the hot buzzword, only recently surpassed by AI hype, and perhaps now by blockchain. Well Big Data moved from hype to reality for many players, especially in the finance industry.
While I’m fascinated with the possibilities of AI, and will continue to explore the value it can bring our platform, we at TS have our feet grounded and focus on the practical and now. We’ve spent years at TS capturing and storing copious amounts of data regarding our clients trading activity. At the moment we process and store more than 10 billion events a day, and have persisted more than 250 terabytes of trading related data.
The effort to capture this was far from small, it was laborious and expensive to be exact. Up until recently the effort to store it was vastly exceeding the effort to actually use this data in a meaningful way. Easily 90% of our efforts went to storing the data and, and less than 10% extracting.
Fortunately we’ve been hard at work over the past year building some really great tools to extract this data. Once you have extracted the data, of course, you’ve still have to make sense of it.
How are my traders performing? Which funds are doing best? How do I, as a trader, mine through all the data to find new opportunities within in my risk and compliance profile for my firm and my clients? Critically, am I keeping my desk efficient and productive with effective transaction cost analysis (TCA)?
In the absence of clear data visualizations, these are seemingly straightforward questions that don’t always yield straightforward answers. The days of saying “this is what I’ve done for you” to customers through spreadsheets and emails should be long behind us, but they’re not.
We’re committed to changing that by providing our users powerful data visualization tools that will help them get to the heart of the matter in minutes and relay that information to end clients in a digestible manner. By harnessing the power of tools such as Tableau Business Intelligence software, we’re able to offer any firm comprehensive and compelling visualizations for transactions, and specific traders’ activity integrated directly into its EMS.
With these visualization tools:
- Interactive reports and visualizations can easily be generated to help prove compliance for ‘best selection and best execution’.
- Fund managers and trading heads can easily track their team’s performance.
- Benchmarks can be set for brokers.
- Integrate output as part of workflow process.
- Leverage the data to visualize hidden insights not naturally viewed on a Excel spreadsheet
For the more sophisticated clients, we’ve also built APIs to allow them to access this data programatically. This allows clients to focus on their secret sauce, what makes them unique, instead of spending time and resources building out the complex technical infrastructure required to store and process this data.
In the age of high frequency trading and big data, there’s a great deal of information out there. But if you don’t build credible 5-mile-high views, or at least, make sense of the minutiae, you’re never going to know where you need to drop in and act. Whether you’re tracking a fund, measuring your team’s performance, conducting TCA, or simply trying to find meaning in a sea of data, a clear picture is the first step toward maximizing efficiency and allowing your firm’s people — not robots — to do their best work.
In this blog post, we share insight into the basic requirements, key challenges, our approach for a smooth transition from IBORs to ARRs and an outlook for what’s next in the multi-year journey to move away from IBORs.
When analysing performance, measuring a portfolio’s actual return answers the “what” and “when” questions –– it tells us what return the portfolio delivered over a specified period of time. While that information is obviously important, the goal of a performance attribution analysis is to go beyond “what” and “when” to explain “how” and “why”.
Fixed Income markets are based on an underlying assumption that has prevented the development of a market structure that enables liquidity, transparency, and higher volumes.