Internal Data: Tapping its Potential for Competitive Edge

This article is part of Gary’s ‘Outside-In Thinking’ series of articles on how the asset servicing industry can design solutions and technology that brings clients closer to achieving their objectives. You can also view the article and comment here on LinkedIn. Read more about Gary and see further articles here. 

Asset managers continually use data to try to gain competitive edge in their investment and trading decisions. They turn to various – and sometimes exotic – data sources for this purpose, for example, using satellite data feeds of Walmart car parks (to help discern retail trends) and of oil-laden ships in the Straits of Hormuz (to gleam insights into oil flows).

But while having timely access to data sources and standardized internal metrics is helpful, it’s rarely the solution alone. These data streams are expensive and have limited time-value. Gaining the kind of sustained edge asset managers are seeking requires exclusive access to a unique data source.

But there is one available data source that is inexpensive, unique and has continuous potential value – an asset managers’ internal data.  

Yet, despite it likely being one of their richest sources of data available, in my experience few optimize or extract value from it. It’s the greatest untapped source of competitive edge.

The best fund managers have traditionally combined different types of competitive edge – primarily informational, analytical and emotional edge. But these are becoming harder to exploit.

Firstly it’s almost impossible to sustain informational edge – access to information that few others possess – in today’s era of freely available online data. The same goes for analytical edge – gaining differentiated insights through the quality of your analysis – when many rival firms’ teams are armed with equally powerful analytical and processing technology.

Emotional edge, however, will always exist. Intuitively, someone hardwired for loss aversion will seek to avoid crystallizing a loss – despite mounting evidence in many cases. As psychologist Daniel Kahneman’s work establishes, our human nature and biases mean we are ‘hardwired’ to make bad decisions even when presented with untainted data¹. We can all potentially let emotion cloud our judgement and become desperate when things go badly. Or come to believe – however fleetingly but expensively – in illogical ideas.

The value of mastering emotion can be traced back to ancient philosophies. Every great investor down the ages has their own take on it; from Warren Buffett: “If you cannot control your emotions, you cannot control your money” to Peter Lynch’s maxim: “When you sell in desperation, you always sell cheap”.

My favorite is that of Sir Isaac Newton, who after amassing and losing a fortune in the South Sea Bubble of 1720 declared: “I can calculate the motion of heavenly bodies, but not the madness of people”.

That sums up the idea of gaining an edge through mastering your emotions. But how can this be done?

Asset managers can use internal data to extract intelligence to help portfolio managers remain rational. Signals based on hard data can dispassionately alert them to the kinds of scenarios I’ve described.

Consider your firm’s unique internal dataset – which may range from information lying dormant in spreadsheets and notes to insights in the minds of your team. Imagine it’s placed into an interactive and continually-updated repository where it’s combined with other information to inform learning and improvements to your processes.

Codifying human activities allows asset managers to gain insights into the strengths and weaknesses of their personnel. For example, analysis of a portfolio managers’ trading history can reveal their behavioral characteristics, such as whether they perform better in volatile or trending conditions. Then, when you identify patterns of behavior that led to positive performance outcomes in the past, you could nudge the manager to size-up their position.

Conversely, if they are entering a negative pattern, you could nudge them to reassess their investment case and cut a potentially losing bet. You could pair historical holdings and transaction history with behavioral science potentially to identify the psychological factors that drive (or undermine) investment performance. 

Equally, asset owners can use similar processes to inform their asset allocation decisions or asset manager selection. It means that they could potentially base these decisions on metrics indicating skill rather than past performance – which, if our industry continues to undergo rapid change, will likely prove less important than it is today.

In these scenarios, every stakeholder in the investment value chain can have access to a unique array of datasets and can perform similar analysis. Any unique internal data source could lead to similarly unique insights and help its owner outperform its competition or optimize their decisions. That's exciting and powerful. It’s the premise of Northern Trust’s Investment Data Science capabilities, which pairs our core data and hallmark service with some of the industry’s leading technologies.

Every asset manager looking for an edge needs to move beyond using the same datasets. Their internal data is the most powerful source they can access – and which from my experience remains underutilized. But many institutions seeking to optimize it face an expensive data architecture problem.

Data is often siloed across multiple legacy platforms, perhaps resulting from acquisitions, based on incompatible assumptions and different types of outdated code. It can require an army of data scientists to extract, clean, and validate it before it is useful – and while this may be an option for the largest firms, viable solutions have not been available for less-resourced ones until now.

At Northern Trust, we faced the same problem internally. We wanted to modernize our technology stack to make it more usable and enable us to extract value more quickly. In my next article, I’ll outline more on this challenge and also how our solution supports tools built on internal data – which I believe are the only ones likely to sustainably produce competitive edge long into the future.

¹ Summarized in several works by Daniel Kahneman including Thinking, Fast and Slow, 2011