The analytics arms race
As the multi-asset investing market grows, the race is on to provide the superior analytics needed to understand portfolio performance, says Sebastian Ceria
At a glance
• Multi-asset approaches have grown but there are increasing correlations between asset classes.
• Without analytics it is almost impossible for asset managers to understand the complexities of most multi-asset portfolios.
• Complex analytics demands ever more computing power.
• Superior analytics is key to differentiation and superior competitive performance.
Multi-asset investing has grown remarkably over the past decade. It has democratised access to investable assets that were once available to only a select few. But at the same time, thanks to the proliferation of multi-asset investing itself, the independent asset classes that once provided portfolio diversification are becoming increasingly interconnected.
More connections mean increased interaction, which leads to increased correlations among those asset classes. And the only way for asset managers to understand those complex dynamics – and to capitalise on them – is with mathematical models and heavy-duty analytics. Multi-asset investing is triggering an analytics arms race, the scope of which few grasp, and even fewer are equipped to meet.
A host of new, accessible and liquid instruments, such as exchange-traded funds (ETFs), including smart beta, can be easily combined into multi-asset portfolios. And that has blown away the barriers that once kept asset classes – and asset managers – distinct. Investors have the flexibility to invest in liquid instruments in equities, fixed income, futures or commodities as they choose – and to mix those with other assets, such as derivatives, in order to hedge unwanted risks. Considering that until recently almost everyone once spoke in terms of nothing but portfolios of individual asset classes, this transformation is remarkable.
A recent survey of firms, with a total of $30.6trn (€27.3trn) under management, found that one-third had launched a multi-asset investment product in the last year, and over 66% had done so in the last five years. And there is no sign of the pace slowing. Multi-asset solutions are forecast to increase by nearly 33%, from $2.8trn in assets under management in 2015 to $3.7trn by 2020.
Which brings us back to analytics. Without analytics, it is almost impossible for asset managers to understand the complexities of most multi-asset portfolios. Unlike most single-asset portfolios, which are built from the bottom up and based on detailed knowledge of each asset, multi-asset portfolios are typically built from the top down and are designed to meet certain defined-factor exposures and characteristics.
To accomplish this, multi-asset portfolios are essentially packages of different kinds of assets, including ‘pre-packaged’ liquid instruments, such as ETFs, and even complex instruments, such as derivatives and convertible bonds. These multi-asset portfolios can also be designed to deploy quantitative strategies, which adds another level of complexity, because quantitative strategies are by definition transparent, rules-based, repeatable and scalable – and require a different set of tools to analyse.
Ultimately, the complex interaction of all these characteristics is behind the need for a new breed of powerful analytics, because in one critical sense multi-asset investing is no different from any other form of investing – it is all about the trade-off between risk and return. Like other investments, multi-asset investing requires looking at the risk profiles of portfolios. And for that analytics are essential – analytics that enable asset managers not only to understand the risks in multi-asset portfolios, but to build better multi-asset portfolios.
For asset managers, the response to this situation seems self-evident: bring on the analytics. But it is not that easy, because today’s investment landscape is all about ‘more’. More asset classes. More coverage. More stress testing. More transparency. More flexibility. More consistency.
“For asset managers the analytics arms race is not about either vendor solutions or wholesale outsourcing. It is about differentiation and superior competitive performance. Access to superior analytics is the key to those objectives”
And that means more is being demanded of analytics. Analytics must be able to consume and digest far more data and more history than ever before, and that requires even more computing power.
The upshot is this: to get the analytics they need, asset managers essentially have two choices. Either outsource the entire problem, or engage with a variety of vendors to create a custom ‘ecosystem’, with analytics at the core.
For outsourced solutions, BlackRock is the dominant provider by far. Having invested heavily to develop its own ecosystem, BlackRock quickly realised that it had created a solution that it could sell. According to a recent The New York Times article, BlackRock’s risk analytics platform, Aladdin, helps “firms trade, analyse and keep a compliant eye on the assets they manage,” noting that “75 firms – including Deutsche Bank’s asset management unit and Freddie Mac – managing $10trn, now use it”. Yet, the question for asset managers is: If I outsource my risk technology to, say, BlackRock, what is left? And do I really want to use the same analytics platform as my competitors?
Alternatively, asset managers who choose to build their own solutions will need a package of vendor-supplied tools that not only play nicely with each other, but are fuelled by consistent and reliable analytics.
Unfortunately, most of the solutions offered by leading analytics vendors were developed in the 1990s and rely heavily on legacy applications and technologies that lack the necessary processing muscle.
Some analytics vendors have responded to this challenge by saying, “OK, we’ll beef up our existing IT infrastructure, we’ll build more data centres, we’ll hire armies of people to cleanse data, and so on. If clients and regulators want more, we’ll adapt our existing solutions to the new requirements and, of course, charge more for it.”
There are two fundamental problems with this approach. First, legacy technology – beefed up or not – cannot handle the computing requirements, in terms of either volume or complexity, of state-of-the-art risk-management analytics. Second, clients will resist paying more for services where costs scale with complexity.
But it goes even beyond that, because the analytics that clients want today must be customisable and flexible, and that, too, requires more computing power. This is why the role of technology in multi-asset analytics is often misunderstood – because even the biggest data centre has finite capacity and still has to scale for peak.
For asset managers the analytics arms race is not about either vendor solutions or wholesale outsourcing. It is about differentiation and superior competitive performance. Access to superior analytics is the key to those objectives.
The next generation of ‘super analytics’ will issue forth from above: the cloud. Those asset managers willing to embrace the latest technology will reap the rewards. The race is on.
Sebastian Ceria, PhD, is chief executive officer and founder of Axioma