Snowflake Data Clean Rooms for M&A
The purpose of this post is to describe how a Snowflake Data Clean Room can address some of the challenges faced today in performing pre-deal analytics for M&A use cases.
Within the world of Mergers and Acquisitions there is an extensive body of analysis occurring prior to the actual deal where two or more business entities will perform due diligence on the other entities or parties to determine whether a M&A transaction will make good business sense. The underlying process will often leverage 3rd party data sources in addition to 1st party data. The 1st Party data may include financial metrics that can be used to calculate valuation metrics or customer master data so that the customer bases of the parties may be compared. Whatever the case, the underlying processes supporting these activities often rely on emailing of spreadsheets between these parties. This can be inherently problematic for the following reasons:
- Providers of 1st party data lose control over who is using their data, how it is being used, and who has visibility to it.
- When working with spreadsheets, it becomes challenging to standardize and enforce format and structure of the data
- Relying on email or even ftp to transmit and share data introduces latency and additional security concerns
- Workflows involving spreadsheets and email are very difficult or impossible to automate
- Data residency & geo-location compliance comes as a barrier in data exchange.
- Fragmented data spread across multiple systems make it difficult to ensure the availability of data.
The principal actors in an M&A transaction can be varied but will often include
- Private Equity firms (PE)
- Investment Banks
- Venture Capitalists
- Venture or Corporate Strategy organizations within larger companies
- Any company or business entity that is party to the M&A transaction
Since M&A is a very broad area of study, we will be focusing on three specific areas in describing how Data Clean Rooms can help: Valuation Metrics, Customer Overlap, and Peer Comparison.
Baseline information for any M&A transaction must include an analysis of the valuations of the parties under consideration. There are many different methods and techniques for estimating valuation that are well documented. The truth is there is no single metric that provides the entire picture, and it often requires a trained and experienced eye to consider multiple metrics in arriving at the best possible valuation estimate. It is also true that each metric has its own set of pros/cons that impact their usefulness. For example, PE Ratio (Price-Earnings Ratio) is probably the most widely used and best understood valuation metric, but it does not consider debt and is not useful when valuing private institutions. So, we might also look at EBITDA (Earnings before Interest, Taxation, Depreciation, Amortization) which does consider balance sheet components like debt. However, consensus estimates for computing EBITDA are not always available. A good discussion of some of these metrics and their respective tradeoffs can be found here:
10 Stock valuation metrics and their pros and cons
Valuation metrics and models can be invaluable when assessing stocks to invest in. These ratios are by no means…
In addition to the principal parties under consideration in the M&A process, it is often useful to compare these organizations to a select set of peer companies. These peer companies may be in the same industry and have similar:
- Business activities
- Demand Factors
- Cost structures
- Access to capital
There are several data providers that offer this data as a product such as S&P and FactSet. Calculating the same valuation metrics for these companies and comparing them alongside the principal parties under consideration can yield valuable insights that can help increase confidence in your valuation estimate.
Beyond looking purely at financial metrics like valuations, M&A analysts may also look at the size and composition of the market. The parties being analyzed each have a customer base. Knowing the size of the respective customer bases, the number of customers in common, the revenue associated with the common customers, and their high-level demographics can provide valuable insights to the M&A analyst.
The problem with legacy solutions
Most M&A workflows today are plagued with challenges stemming from a fragmented system architecture where multiple technologies must be manually stitched together. Data must often pass through multiple phases before it is ready for consumption creating multiple copies along the way. Moving spreadsheets and PDFs via email is not the recipe for a robust, automated, and reliable workflow. Legacy approaches often lead to stale, untrustworthy data with high cost and complexity.
Snowflake Data Clean Rooms
To help in use cases where multiple parties have a need to securely collaborate on shared data, Snowflake has developed a Data Clean Room framework. The Snowflake Data Clean room allows two or more Snowflake customers to analyze data without exposing it to one another.
Several differentiated features specific to Snowflake come together to make this possible:
In addition to the Snowflake features listed above, Data Clean Rooms can be implemented with an explicit validation step whereby a data provider must authorize each consumer request against its data before it can be executed. This capability, as well as those listed above, are fully configurable and can be customized to meet the governance/security requirements of all parties within a Clean Room.
Why Data Clean Rooms for M&A?
With all of that as a backdrop, it should be clear why Snowflake Data Clean Rooms and M&A pre-deal use cases can be a perfect match.
- Multiple Parties
- Needing to collaborate on shared data
- With a need for high-quality 3rd Party data
- And a requirement to protect access to sensitive data for each party in the Clean Room.
Addressing these requirements is foundational to why Data Clean Rooms were created in the first place!
We’ve recently developed a demo within Snowflake that illustrates all of these capabilities and how they can be applied to M&A. This is not a complete solution but rather an “art of the possible” example of how Snowflake’s unique cloud-native architecture can make your M&A experience better, faster, and more secure. The demo includes the following capabilities:
- Data Sharing between 2 parties on 2 Snowflake Accounts
- Governance Policies (Row Access Policy, Conditional Masking, Explicit Query Validation with Snowflake Stored Procedures)
- Valuation Metrics Calculated in a Secure View (PE Ratio, EV/EBITDA, Discounted Cash Flow, Peer Valuation Comparison)
- Customer Overlap Analysis showing customer counts and top-line customer revenue using double-blind joins
- Synthetic Peer Data accessed from a direct share
- Streamlit UI controlling query execution and visualization of the results
With multiple parties collaborating on shared data sets in Snowflake, the possibilities are limitless! Snowflake imposes no limits on scale of storage or compute so the customers are free to drive whatever amount of complexity and scope into their analytics. Additional 3rd party data can enhance the experience by providing not only current and historical valuations but also projecting future valuations based on macro-economic conditions, market data, or industry guidance. These can be developed into models that are trained and deployed on Snowflake using Python or AI/ML partner technologies.
If you would like to see the M&A Clean Room in action, check out the demo in the video below:
If you would like further information or would like access to the code behind this demo, please contact your Snowflake representative.
I hope you found this article useful. If you enjoyed it and would like additional info, please follow me on my profile or hit me on social media at Linkedin.
All information and opinions expressed in this post are solely my own and do not represent the views or opinions of my employer.
Very special thanks to Vikash Kumar for his contributions to this article and to the construction and testing of the DCR demo!
Check out this link to run a Snowflake DCR in your own environment: Snowflake DCR Quickstart
Here’s a great article providing more detail into the DCR framework and why its important What is a Data Clean Room and Do You Need One?