We believe that we’re today at the start of the 4th Industrial Revolution where Artificial Intelligence (AI)/Machine Learning (ML) and robotic process automation (RPA) are becoming instrumental components of an investment professional’s workflow.
The benefits are clear: more efficient processes, better capital allocation and better outcomes for clients. We would warn though that these technologies are no panacea but rather very efficient tools to significantly improve decision-making.
The issues
As we are now discussing the topic at the beginning of 2021, we can confidently say that front-office fixed income professionals, and therefore all other business stakeholders, have barely seen any positive impact from the innovative technology available to them. Credit analysts are still crunching the numbers in spreadsheets and still face challenges with data aggregation and the subsequent timely communication of risk recommendations or trading ideas to portfolio managers. Portfolio managers are overwhelmed by the number of reports generated by their analytical team and find it hard to monitor multiple portfolios and manage client relationships at the same time. Finally, traders often face difficulties with trade executions because of the lack of information on liquidity and prices.
It is in that context that we set up Intellibonds back in 2018 with a skillset in both fixed income investment management and technology to create the first AI-human collaborative platform offering virtual AI-assistants to the fixed income front office.
As the fixed income industry is playing a catch-up with other more ‘tech-savvy’ asset classes, we can comfortably say that the secret sauce is not the technology per se but rather bringing together structured data, expertise in both fixed income and technology, and finally having the right vision. All in all, we would describe the three necessary ingredients as:
- Access to and making sense of structured and consolidated or unstructured and fragmented data.
Data is the fuel and comes in different grades from different sources. This is a significant obstacle to the adequate use of data as it means that you need to subscribe to many different data vendors and standardise the format. This is not only complex but also very expensive (and sometimes impossible to acquire given the competitive dynamics within the industry). While this is somewhat true for equities, it is the number one issue for fixed income given the complexity of the asset class. There are more than 150,000 securities in the fixed income universe versus c. 3,000 for equities (MSCI World and EM). Also, debt comes in different shapes or forms: seniority, maturity, collateral, currencies, embedded optionality, etc.
Therefore, turning multisource, fragmented and unstructured data into risk recommendations or trading ideas takes time and requires a significant level of sophisticated labour or automation. This process can result in either higher compensation or technology costs, which only a few of the largest institutional investors can afford. At the same time, the competitive dynamics in the industry (for example the difficulty to buy data from a competitor) puts any established player at a disadvantage compared to independent start-ups like Intellibonds.
- Successfully combining technology with fixed income expertise
In-house technical expertise is often lacking and hired technical skillset often doesn’t understand the specifics of the fixed income business. Ideally fixed income professionals will obtain their second training in programming and vice versa. The usual pattern seen is that the quantitative research/development team hires new data scientists, or Python/R retrained quants. Depending on the company, this model can have some success, which will usually depend on whether the fixed income team embraces the new technology. The best way to make this happen is to look for the early adopters within the fixed income team, as they would likely be open to learning and trying the new techniques, integrating them in their investment process. Unless the application of technology is supported by the leaders (see next topic) that strategy would be protracted at best. With time the whole franchise would become uncompetitive as competitors get ahead on the adoption curve. Possible solutions to this would be to either bypass the internal difficulties and take on an external solution or engage in open innovation. Both of these solutions would still require the commitment of the leadership team in order to be successful.
- Commitment of the executive leadership team
Often a lack of strategic direction and commitment to innovation at the executive level means that many great ideas get lost or end up being pushed aside and never materialise due to internal politics. A lack of receptive executives often leads to the top talent leaving organisations and choosing an entrepreneurial journey. But there are some examples where out-of-the-box thinking in the asset management industry has been thriving, for instance, through the creation of intrapreneurial departments that nurture internal innovation or open innovation platforms via venture arms. The most common examples would be Fidelity International and Alliance Bernstein with the list of those managers getting longer every day.
Fertile ground for fintech start-ups
In conclusion, what is the secret sauce? It is not about having the best technology, the best programmers or having cash to invest. In our experience, the structural ability to source and make data insightful as well as the ability to integrate technology and investment processes are the key. This is a fertile ground for fintech start-ups.
The digital transformation of fixed income asset management is becoming a priority given the industry and market headwinds (low rates, margins pressures, intense competition). We further believe that transformation will lead to AI/Human collaborative cloud-based platforms that would allow for real-time data aggregation, processing and sharing plus AI-augmented workflow and seamless silo-free collaboration between fixed income front office teams. The elimination of inefficiencies could bring a new type of investment products: AI active, a product that will combine active alpha with a passive cost base.
