Introduction
The alternative investment management community has and is going through a radical transition in the way they research and choose their investments; a transition which will most certainly lead to winners and losers, one I am a part of. Below I will discuss the traditional process in respect to both hedge funds and venture capital firms and the implications of evolving into data-driven investment managers.
What is Alternative Data?
Simply put, data which cannot be found on traditional financial statements and which is leveraged in modeling the key drivers of any given business and/or providing insights into an endless list of factors that impact the performance of a company. Examples of this can be credit card transactions, email receipts, social media, app usage, open-source, geospatial, etc. The type of data and its usage ultimately depend on the business being researched, i.e. consumer credit card data would not be a good candidate for B2B insights.
Example Use Cases: market share, retention, customer overlap, revenue growth, subscriber growth, product composition, discounting, customer mix, AOV, spend per customer; all of which could potentially be cut by granular dimensions (i.e. demographics, geography, etc).
The process is to first find unique and non-commoditized data. This is done through the purchasing of raw assets from data vendors and/or collecting the data yourself. This is likely the easy part, what is significantly more difficult is transforming this often unstructured and unclean data into something that provides tangible insights. This requires a combination of deep business knowledge, highly component data engineers, and even more component data scientists; the essence is a melange of business and technical worlds. The easier it is to acquire data and transform it into something meaningful, the more likely it is or will become commoditized.
The value proposition is clear: producing differentiated research into any given company, but rather than using ubiquitous sell-side research and assumptions every investment manager does, using an alternative means of due diligence. The product of which can be to arrive at numbers on any of the financial statements, arrive at numbers not reported by a public company but critical to their business, and/or to provide a gut check on the health of the business and it’s macro environment (there are more use cases, but these stand out as the most relevant).
With this knowledge, the key quickly becomes how you can source differentiated data and transform it into something that provides insights no other investment manager has. The data is publicly available; but finding the data, trialing it, cleaning it, and modeling it – is an arduous process that requires a significant amount of financial resources, time, and talent. This acts as an enormous barrier to entry for most investment managers who face the proposition of spending tens of millions for an uncertain payoff, simply buying data is not enough; the challenge is more often the talent that transforms something raw into something meaningful. But when there is a payoff, it’s big. The effect is compounding, as while you build your alternative data infrastructure, the repository of proprietary and differentiated data you own continues to grow leaving you with a gold mine.
Hedge Funds
Although the positive implications of alternative data on investment research and management are agnostic of strategy and sector, for conversation purposes I’m going to discuss long/short equity funds, that which I have experience in.
The process of long/short funds can be simplified into developing theses; be it at a macro-level or a company-level: top-down or bottoms-up. These theses can be developed through your domain knowledge on any given sector and/or through research (proprietary or not). And in developing these theses; conducting rigorous due diligence that either supports or negates your hypothesis. Finally, acting upon your findings in determining positions, exposure, and timing.
This due diligence process consists of two facets: qualitative or quantitative. From a qualitative perspective the approach is generally evaluating the management team, which can be done in various ways: physically meeting with management to determine whether you believe they are competent and/or if their vision aligns with your version of reality. From the quantitative perspective, the approach has historically been built on the foundation of traditional financial modeling. The purpose of which can determine several key components of your thesis, most notably: the value of a stock relative to the market which can be driven by inconsistencies in accounting and/or key financial metrics: metrics whose relevance can change depending upon the type of company. The key assumptions driving the financial models are often but not always derived from sell-side research.
Traditional financial modeling is an essential component of the investment process but increasingly commoditized and automatable; dozens of popular software tools exist to assist in this process (i.e. AlphaSense). There is very little value-add in manually going through 10-K forms and sell-side reports to input financial data into arduous excel models that have minimal transparency and are difficult to audit. This is archaic and top-tier hedge funds are no longer utilizing this process in their decision making process as much as they have historically.
This trend of moving away from the traditional process can be thought of as directly resulting from the birth of alternative data. In fact today, top sell-side analysts often use alternative data as part of their research reports (which one could argue becomes definitionally commoditized).
Alternative data can be used in various ways: in both developing a thesis and researching any given stock, oftentimes both use cases are utilized simultaneously. Accurately predicting the stock price movement of a public company before they report earnings is the essence of high returns, and as a result, the essence of a top-performing hedge fund. One can imagine leveraging credit card data to arrive at predicted revenue growth for a consumer company before the market as being valuable. Or perhaps mining email receipts data to understand product-level sales, mix, discount, etc. There is an endless sea of possibilities alternative data provides to public equities research.
Venture Capital
Alternative data impacts the ventures process in a different but equally disruptive manner. As one can expect, depending on stage, the private investment process changes dramatically. Early-stage companies have a greater focus on founders: finding high-quality founders before others and vetting those founders (mostly through a qualitative approach). Your belief in an early-stage founder is oftentimes one of the key factors in determining whether to invest or not.
From a late-stage growth perspective, the process is different: there is greater focus on business strategy and valuation, and as such, a greater focus on quantitative due diligence. Historically this process resembles the public’s approach but with key differences as private companies are not required to publish their financials. Similar alternative data analyses used to evaluate the public markets can be applied to late-stage growth companies i.e. if you believe company x’s valuation is high and you have unique data that paints a picture of revenue growth not inline with their predictions, that would likely be material information in your due diligence process in either negotiating term sheets and/or choosing not to invest.
One of the newest and most notable use cases of alternative data in private investing is exposing your proprietary research to your portfolio companies. In this, you are fulfilling your promise of providing tangible value-add to your investments. Additionally, this can and is used as leverage in getting a hot company to sign a term sheet with you over the competition.
Late-stage growth companies are already well-established businesses, and those which are unicorns likely have every VC knocking at their door. In a sense, the roles flip, oftentimes having the VC desperately attempting to sign a term sheet. And it is precisely this which accentuates the need for a VC to differentiate itself from its peers in the value it can add to any particular company. After all, the premise of venture capital is not simply funding deals, it is foundationally built on adding value to the companies that you invest in. Why would company A choose to partner with VC 1 vs 2? This has historically been driven by a VC’s network and domain knowledge that can assist in business strategies, pricing, exits, etc. But increasingly, private companies are expecting more: analytics they can leverage in various facets of their businesses.
To summarize, the implications of alternative data on private investing is two-fold: First, sourcing founders/deals in addition to differentiated due diligence. Second, providing your research to the companies you invest in, thereby adding value in their business’ growth.
Conclusion
Anyone working within a progressive top-tier firm as part of the alternative investment management community is likely familiar with the concepts I have discussed above, as this transition has been happening rapidly within the past 5 years: a transition I have been an early contributor to. The implications are clear in that hedge fund managers who cannot evolve into an alternative data-driven organization will find themselves in a difficult position to differentiate their theses and as a result, themselves. From the venture capital perspective, it will become increasingly difficult to originate deals before data-driven competitors and will be significantly more difficult to invest with your rolodex as the sole value proposition.