In (Part 1) of this whitepaper, I discussed some basic tenants and fundamentals that must be established before an analyst can apply value methods and techniques to determine the fair market value of a database. The gist is, a database is a collection of datasets which is appropriately considered an intangible asset. Databases are being used more and more in AI applications as well as drug development and value-based arrangements. More often, organizations are willing to buy and/or sell their de-identified patient data, and the question of fair market value will begin to emerge more often for compliance with the Anti-Kickback Statute.
There are three main approaches to valuing a database; the income approach, market approach and cost approach.
Income Approach – The income approach to valuing a database is a technique whereby the analyst estimates the amount of future income the user/buyer of the data would earn from having access to the data and discounts the future cash flows to the present value (i.e. discounted future cash flows). Normally, the analyst would forecast the future income to the owner of the database. However, the future income stream accruing to the owner or the buyer/user of a database, specifically in the healthcare/life science space is largely unknown and/or highly imprecise to forecast, which makes the income approach unusable. For example, one of the most likely uses of de-identified patient data or case outcome data is for the development of pharmaceuticals or other new medical technologies. De-identified patient data is only one elemental input to the development of a drug or medical device and sometimes companies invest large sums of money into drug discovery and nothing ever develops. Conversely, trying to ascertain the future income to the owner of a de-identified patient database or a collection of case outcomes is equally as difficult because the market is just now evolving and there is not enough willing buyers and sellers to culminate into an observable market. For these reasons, the income approach is not an applicable technique to determine the value of a database.
Cost Approach – The cost approach to valuing an intangible asset is a technique that uses the concept of replacement cost or reproduction cost as an indicator of value. The premise is that an investor would pay no more for an asset than the amount for which the utility of the asset could be replaced, plus a required profit/return to incent a third party to replace the asset. More specifically, the analyst looks at the reproduction cost of the data which is estimated as the cost of recreating a replica of the data. Or the analyst estimates the replacement cost which is equivalent to the cost of replicating the utility of the data rather than creating an exact copy of the data.
The cost approach can be a viable valuation approach for intangible assets but it largely depends on the specific characteristics of the asset. In the case of patient records, it will depend on the types and amount of parameters each record (i.e. patient) includes. The analyst would attempt to estimate the human hours incurred to manually compile the specific data from patient files or case outcomes. Normally computer systems and EHR contain this data, but the specific data/information has to be extracted and assimilated into a database. Furthermore, in some situations, cases are naturally occurring in real-time and data is being compiled as it occurs. Based on the highly unique nature and characteristics of a health record, the application of a cost approach is feasible but very challenging.
Market Approach – The market approach is considered the most feasible and viable method to estimate the market value of a healthcare database. the market approach to valuing databases is based on research on comparable transactions in the public domain either for raw data (i.e. customer leads, contact lists, or demographic information) or licensing and royalty deals. The sale of a healthcare database is similar to the use of other intangible assets where the buyer/user pays a licensing fee or royalty rate. An owner of a dataset, trademark, trade name, or technology (such as software) charges a royalty or licensing fee to its users. The same transaction could be executed for de-identified patient files that a health system uses to establish benchmarks in a value-based arrangement.
For example, a salesperson can purchase a dataset of the name, address, email, company name, etc on individuals who might be considered leads for his/her product. Understanding there may be special circumstances where data is sold to a buyer and the seller is unallowed to re-sell or use the data again.
There exist in the market royalty rates for uses of technology. Again, the challenge to applying a royalty rate to value a healthcare dataset is the future revenue to the user of the data is largely unknown. A license fee could be applied for use of the healthcare database, but finding a sufficiently comparable licensing transaction is a challenge, but feasible. There are various transaction databases in the market for information like business sales transaction data, customer leads, royalty rates, compensation survey data, public company financial data to name a few. The challenge deepens because some of these organizations charge a flat fee to access all of the data or they charge a fee per record.
It’s the valuation analyst’s job to sift through the various market comps to identify the most suitable and appropriate one to apply. The market approach is currently the most viable method for a valuation analyst to use to determine the fair market value of a de-identified patient database.
More to come on this topic….