PARTICIPANTS

  • Ross Allen, Managing Director, IHS Markit
  • Andy Ji, PhD, Head of Research, BlueFire AI
  • Tim Riordan, Portfolio Manager, Aware Super
  • Desmond Tjiang, Managing Director and Senior Portfolio Manager, Conning

DISCUSSION HIGHLIGHTS

Key data challenges

Take-aways

  • Assessing the relevance of ever-growing data-sets is weighing on investors – in an audience poll, around two-thirds of asset owners and managers who responded said their biggest challenge in using data for investment purposes is “making sense of it amid the overload”
  • Adding to the deluge of information flows today is the increased focus among investors on non-financial data, such as environmental, social and governance (ESG) and sustainable finance-related factors
  • Further, higher allocations to private market assets amid the search for yield is fuelling the rapidly expanding volume of data – while at the same time creating an extra burden given this data is opaque and not easily available\

Insights

Desmond Tjiang, Conning
“The biggest challenge for us is always what kind of data to use, how to use it properly and whether we are using too much or not enough when making our decisions.”

Ross Allen, IHS Markit
“There is an increasingly important and clear link between real-world economy data and financial markets.”

Andy Ji, BlueFire AI
“’Early’ is a subjective term. We look at it as a six- to 12-month period in line with how investors look at tactical portfolios. We aim to systemise human decision-making processes, so we try to ingest data-sets in the same way as an investor does to frame the early warning signal.”

A more dynamic approach to data

Take-aways

  • Accelerated by Covid-19, investors are finding they need to connect global data across numerous industries to generate insights into risks and opportunities as part of a highly integrated view of the world
  • Investors are combining old models in a bid to derive practical insights. For example, most data is backward-looking with limited history, so today’s unprecedented indicators potentially make it dangerous to stick to old models without adapting to the wider environment; it would have meant missing market highs during much of the pandemic
  • Early warning signals provide indicators of credit risk, investment opportunities and risks based on ‘trigger events’ that create inevitable knock-on effects, such as the blockage of the Suez Canal in March (idiosyncratic impact on specific vessels and commodities) and the Australia-China trade tensions (implications for commodity prices)

Insights

Tim Riordan, Aware Super
“The faster adoption of technology has meant we now see everything move faster, from news flow to data collection and availability. This is reflected in financial markets based on how business models evolve. We use a mosaic theory approach by stringing together multiple pieces of data to ‘triangulate’ the information and give us confidence and conviction around the view that we have.”

Desmond Tjiang, Conning
“The quality of data is more important to us than quantity, so we screen out a lot of information and only use data which we see as having high statistically-significant predicative powers. For example, bullish investor sentiment and positions in equities might in the past have indicated a warning about the market nearing a peak. So, we need to understand current market cycles and be flexible.”


Ross Allen, IHS Markit
“Integrating data in a single place can help to make it more relatable and see it in the right context. By bringing data together, investors can derive insights from real economy data and understand how it flows into companies from an investment perspective, or as part of broader risk and opportunity assessment.”

Taking charge of data decisions

Take-aways

  • While there were differences in opinion among poll respondents about who should drive the analysis and application of data internally, there was consensus that data scientists are not best-placed to perform this role
  • Instead, data decision-making should either be left to the research team (40%), chief investment officer (33%) or portfolio managers (26%)

Insights

Desmond Tjiang, Conning
“It is important to combine the fundamental and quantitative investment teams. This means we use our expertise to understand the data and then use it.”

Andy Ji, BlueFire AI
“There is no defined pathway in terms of how to use data and in who uses it. We have noticed some human resistance to changes in routine.”

Tim Riordan, Aware Super
“We take a layered approach to assess the relevance of a single piece of information to the individual in the investment team – that is, the new piece of data is compared with our existing understanding & position to see how the data evolves our view.”

 

 

Applying new thinking

Take-aways

  • Despite the desire for greater exposure to alternative assets such as private debt, real estate and infrastructure, poll respondents mainly highlighted private equity (44%) as the asset class for which data is most important to support investment decisions. Public securities, meanwhile, accounted for 39% of responses
  • An example of how real-time data has influenced allocation decisions during Covid is housing; people spending more time at home than in the office put in context the shift in property prices in non-metro areas during the pandemic, and has supported investment views on the real estate sector

Insights

Desmond Tjiang, Conning
“Our tactical asset allocation uses a lot of traditional data, so I would like to incorporate alternative data into our existing model. This will save a lot of time and give us a lot of new insights.”

Ross Allen, IHS Markit
“Focused thinking about enterprise data architecture is required. This is needed to ensure both internal and external data across the firm are contained in one place, to eliminate data silos to provide the foundation for early warning signals.”

Andy Ji, BlueFire AI
“The financial markets are subject to various shifts that make Big Data unusable in terms of a consistent structure in data. This requires investors to look at data in bite-sized chunks.”