Dan Furstenberg from Jefferies, discusses alternative data integration on the buy-side, with an emphasis on quantifying strategy. He highlights how nearly 50 fundamental investors are actively building these efforts and are approaching alternative data from a talent, infrastructure and resourcing perspective. He also covers the data science landscape across investment managers.
I have had the opportunity to see an explosion of data usage over the last 24 to 36 months. Clearly, RavenPack and their team saw it much earlier but having a front row seat at some of the largest mutual funds and hedge funds globally, we saw this inflection two or three years ago. Which is basically the use of data, not strictly relegated to the systematic strategies but also to the discretionary investing side.
So with that, I can give you some context. Over the last year and a half, as we've travelled, we've written a couple of pieces on the space and presented at this point, to over 120 fund managers, both systematic and discretionary investing mutual funds hedge fund the like, both domestically in the U.S. and internationally - Europe and Asia. So we feel like we have a pretty good sample set for the community right now.
It's important to know why do we care about big data now since you're mostly systematic community. I think you know what's happened in the last four to five years which is a real advancement and enhancement of semi's. So semi's haven't locked in all of this, there has been large amounts of data, we even had software for quite some time. So what's changed? It's the advancement of semi's. It's not by accident that we've seen the growth of NVIDIA and the stock performance of AMD as well. And other names that are pure players in Asia over this time. There's secular, it's not cyclical, it's not it's not a coincidence.
So what are some observations in the marketplace? Again typically you all know this, you've experienced some rotation of active and passive, which has been dramatic and crowding of strategies both in the discretionary investing side and we saw in the back half of last year on the systematic side as well.
What does that all lead to? It leads towards what we've always had, but even more keenly, a constant focus on alpha generating activities. As a direct byproduct of that would we see? We saw the discretionary community start to come out, and really in great form, both on multi-managers, which are more common, but also on a single-manager platforms start to experiment with vast amounts of data both small and large.
The Evolution of Investing: Adapt or Die
So Moneyball is a favourite book and movie of mine, evolve or adapt or die. So we're just giving something illustrative here. Discretionary investing we think today, the investment community is made up of two-thirds is discretionary with a very small sliver of systematic and an even smaller sliver is quantamental. What we see, probably two to five years out, is an inversion of the discretionary in the quantamental. We see it happening right now in real time at an asymptotic pace. There are very few discretionary managers right now that are not using data in some form, whether it's traditional, price volume or alternative data as well.
On the systematic side, we expect that to shrink very simply due to economies of scale. We saw a mass amount of spin-offs over the last 36 to 48 months. If you've really taken stock of those spin-offs, most of them have gone away, been folded back in. These are technology companies, at their essence and any paradigm history has proven that the CAPEX is really, in addition to innovation, is really a lead indicator or portend of how things end up, so this is how we think the community will look. Only a year or two out.
Relevant Market Themes
Top Objectives for Acquiring Data Sets
- Alternative data is leveraged as a source for idea generation ONLY one-third of the time.
- It remains earliest days for discretionary managers to be broadly leveraging alternative data or data efforts to source ideas.
- 85% of investors surveyed cited enhancement of pre-existing investment processes as a top objective.
Top Data Science Goals for the Next 12 - 18 Months
- A streamlined process including distinct responsibilities for data acquisition, ingestion, application, application and distribution.
- A practitioner, “bridge-role”, positioned between data science team and investment team is crucial in making data a profit center rather than a cost center.
- Buy-in from the C-Suite and cultural integration within investment teams are key to success.
Building Data Strategies for the Next Decade
The majority of the community, to give you an idea of the sample set, the are the most progressive purchasers and users of data globally. 60% still have shared data leads.
What’s interesting, however, if you take a look at the pie chart on the right, is that about half of those are focused on hiring a dedicated data lead in the next 6 - 12 months. 90% of the progressive players will have dedicated dada leads within the next year.
Despite concerns that news entrants to data science may be “late to the game” or sourcing data sets that have “already been arb’d”, data sets are often used for entirely different purposes.
Layering multiple data sets into a mosaic creates an infinite number of combinations and often entirely new signals for the investment process.
Subscription vs Transitory Use
A number of investors commented that they wished there was better alignment between the tenor of the data subscriptions and the length of the use the investor needs for it.
- Underperformance of Active; Rotation into Passive/”Smart-Beta”
- Crowding of Strategies and Positions
- Search for Alpha using Alternative Data Sets
- Consolidation of the Market Structure i.e. Buy-Side and Sell-Side
Digesting massive and varied data amounts and intelligently deriving insights, trends, and actionable analytics.Request Event Materials