Which types of funds invest in spatial analysis? How do they use location data in screening & portfolio management? We discuss the role of Data Science.
According to McKinsey private equity's (PE) net asset value has grown sevenfold since 2002 two times faster than global public equities. As a result the number of PE backed companies in the US has doubled from 4 000 to 8 000 from 2006 to 2017 - and a growing number of funds are starting to use more advanced types of analysis to find new opportunities at different stages in the investment process. The search for differentiation has been a constant in the Private Equity world and it has become more apparent in recent years due to high market valuations and an abundance of dry powder waiting to be deployed.
Spatial analysis and Data Science in particular have been cropping up more in Private Equity job postings - with more and more large cap funds looking to attract Data Science profiles to give them a competitive edge against other funds as this could be the difference to gain exclusivity in an auction process or to increase returns by a few percentage points in current investments.
However the appetite to invest in such profiles (and the technology and data they need to be successful) has been traditionally concentrated in funds with investments with clear physical (or brick and mortar) footprints. We are now seeing a change in this trend with funds willing to understand customer behavior patterns and location-related information in a wider range of investments.
Which industries are PE firms investing in?
If we take a look at data on total PE transaction values by sector in 2019 we can see that Information Technology Consumer and Healthcare have been the most active sectors by number of transactions (c.32 000 - according to our internal analysis).
However by transaction size financials and industrials have also been relevant making up c.15 percent of the total capital deployed (c. $600bn worldwide):
Typically Consumer (which includes Retail) Real Estate and Healthcare have been industries with a strong location component and therefore power users of spatial data. The main purpose of the location data has been to understand changing consumer behavior - rather than relying on portfolio company data or on management experience. A more data driven approach has been adopted by Private Equity funds to maximize returns through optimal Site Planning or Investment strategies. One recurring example is around commercial Real Estate acquisition according to development legislation which we can see visualized in this dashboard below:
In fact the use cases are spreading much further across other industries such as Communication Services Energy and Utilities amongst others due to the inherent CapEx intensity present in these industries. Understanding spatial questions relating to infrastructure and networks are crucial to reach target returns - especially relevant given that infrastructure fundraising grew 17 percent globally and 59 percent in Europe - raising valuations and increasing competition for good assets.
Industries such as logistics and consumer product manufacturers with chilled/frozen networks could significantly benefit from geospatial modeling as distribution represents a large share of total opex and any optimization would directly impact EBITDA the key metric in the Private Equity world.
In which phases are funds using Spatial Analysis?
Just as spatial modeling isn't relevant for every type of industry it also isn't relevant for every phase of a deal. Hearing from our Private Equity clients about their experiences typically spatial analysis is most useful in the screening phase and then later when they are managing their portfolio companies. The underlying reason is mainly the nature of these two phases as they are more analysis-heavy and due to the typical constraints of timing during classic M&A processes (deal making exit).
Private Equity professionals involved in screening and/or portfolio management are starting to see that using more spatial data streams (such as human mobility or transaction insights) in their analysis can be pivotal in pushing their assumptions further to support their investment thesis. They can also see that they may be a source of value creation for a portfolio company during ownership increasing potential valuation on exit.
A Retail Whitespace Analysis Case Study with American Securities
Recently at the Spatial Data Science Conference in New York Tim Kiely (Lead Data Scientist at American Securities a fund with approximately $23 billion under management) walked us through a case in the screening phase which focuses on Retail Whitespace Analysis.
"This is a very competitive bidding process we are competing against maybe 10 other PE firms all of whom have the same information we have. And what we're generally able to do using just sort of locations of different stores is to bring a tremendous amount of information that we don't have and our competitors probably don't have.
We sourced 500 spatial variables everything from demographics social-demographics competitive presence a lot of different kinds of demographics variables (Mastercard credit card transactions data) and we matched those all to a 10-15 min drive time radius around each one of the stores and aggregated them using isochrones.
Using spatial data and very simple techniques we can very quickly gain a competitive information edge in this process which allows us to push our assumptions a little bit more write bigger checks and automatically win out some of these bidding processes."
Who uses Spatial Data Science in funds?
Not every fund has a dedicated Data Science team like American Securities however and within the funds themselves operations teams are normally the driving force pushing forward the case to invest in better quality external data (including spatial data) talent and technology.
With many ex-management consultants working in PE operations teams directly with portfolio companies they may have already had significant exposure to how firms like McKinsey BCG and Bain are already bringing the latest and greatest in spatial modeling to clients in a range of industries and use cases through initiatives like Omni and GAMMA.Inevitably these same consultants will have handled information from customer surveys and will have faced the challenges and inconsistencies of survey data. Real data obtained from credit card transactions and foot traffic are now preferred to avoid using the wrong data while proving their hypotheses.
Where (geographically) can spatial analysis be most powerful?
The quality of any analysis depends on the quality of data inputs which tend to be higher in more developed markets (e.g. US UK Germany). For this reason we are seeing an increasing interest from funds investing in these geographies to incorporate spatial modeling techniques to their investment theses. We are observing this trend despite the more obvious use cases for retail-focused investments in geographies where the presence of e-commerce is more limited (such as Latin America or Asia).
Here you can see a split of where the transactions took place in 2019:
A growing number of funds are seeing the immense opportunity presented by spatial modeling as demonstrated by pioneers like American Securities but many don't know where to start.
If you would like to start understanding how this may be of use for your investments you can download our Becoming a Spatial Data Scientist ebook or reach out to one of our experts to discuss your investment priorities for 2020 and how spatial analysis may set your fund apart.