The Bull Case for Alternative Data

There's a palpable pessimism about Alternative Data's impact in financial markets, currently. While justified currently, the newsletter offers a path forward to long-term success for the industry.

Before outlining the bullish¹ path forward for the Alternative Data industry, first we need to acknowledge the current state of the industry. The Gartner Hype Cycle^TM² provides a useful paradigm to frame the discussion.

I believe we are in the trough of disillusionment for Alternative Data. Gartner’s Hype Cycle^TM is a paradigm that explains how emerging technologies, applications and ideas go from Technology Trigger, Peak of Inflated Expectations, Trough of Disillusionment, Slope of Enlightenment and Plateau of Productivity.

As an example of the Hype Cycle^TM, e-commerce in the late 90s was at the peak of inflated expectations, followed by the dot com bubble bursting. Of course, today e-commerce is on the Plateau of Productivity.

Alternative Data is in the trough of disillusionment. There are pressures from a growing supply of data sources and difficulty generating insights from data for many participants as well as difficulty connecting the ROI impact to the work and data.

On April 2nd two well written, bearish Substack articles were released on the same day, which articulated the current state of the data market place and the data science team’s impact on ROI.

1 - Scaling Data Ops - Why are so many data science initiatives failing? - Vin Vashishta

Why Are Data Science Initiatives STILL Failing So Often?

2 - On the Mark Data SDO Opinion - Do we really need so many data companies? - Mark Freeman

SDO Opinion - Do we really need so many data companies? - Mark Freeman

While I don’t agree with all the points made, I believe these well written articles do very well to capture the current state of affairs. For my newsletter I’m going to focus on Alternative Data as used by financial markets, which is a subset of what both articles cover. The Alternative Data industry faces the same challenges: there’s a huge number of data providers selling to financial markets and not all users of data have figured out how to drive high business impact from data. In the financial markets, similar stories exist as outlined in the referenced Substack posts, as data initiatives on the buyside and sellside³ have come and gone over the past decade as have many alternative data providers.

The future path to the bull case

This isn’t a lost cause. Perhaps it’s the value investor⁴ in me, but at the times when the consensus opinion is so negative, there’s really a catalyst for longer term opportunities with much more realistic expectations. But one can’t simply hope for better days at the bottom. There has to be a path forward to success.

So how does alt data get out the trough and to the period of (successful) plateau of productivity?

Let’s first draw from what we know works. At present time, it’s not all doom and gloom. There are clear success stories where asset managers are generating alpha powered by data driven insights (both fundamental and systematic investors⁵) and there are data, software and insight companies providing products and services that have been proven to power alpha generation⁶ on the buyside.

As one of the first 10 employees of UBS Evidence Lab, I fortunately have experience as part of building a success story in turning data into positive returns. My views below are based on my past experience as a purchaser of alternative data, creator of new alternative data, generator actionable insights from alternative data, and collaborator with end users of alternative data, who benefited directly from data driven insights. These views are general views covering the wide range of experiences working with different businesses and teams - they are not about any specific business or team.

Outlining the path to long term success

There are 3 key factors that when appropriately achieved will move the Alternative Data industry and companies forward through the Slope of Enlightenment to the Plateau of Productivity.

• Realign data and teams with critical client outcomes.

• Make it easy to value a data practice.

• Build network effects.

Realign data and data teams with critical client outcomes.

Data by itself, on its own, is actually useless. The insight from the data is what’s valuable and the insights are only as good as the decisions they shape and the direct impact it has on financial outcomes.

Start with the outcome needed, work backwards to solutions that solve critical problems. “I just want the raw data” is not the actual outcome needed by an end user of data, they are seeking insights with a more specific goal in mind. The importance of understanding the goal behind the request for data can’t be understated. The “why” behind the request is what should be used to realign data products and teams.

Avoid data projects and ideas that are just “interesting”. Interesting is not good enough. The client outcomes need to be material to the business mission.

Too many times the technique or the novelty of the data source becomes the focus rather than understanding what actually needs to be achieved by the business. This happens at all levels of the data to decision supply chain including the end users.

Make it easy to value a data practice.

The articles I reference clearly outline the case that the return on asset (ROA) and return on investment (ROI) are not clear at the moment. I don’t disagree with their analysis.

Data businesses and teams need to be clear about the value of the decisions made that leveraged insights derived from the data. Further, make it easy for users to get to the insights in a way that empowers their decision making. In order to create valuable insights, data practitioners need to have a deeper economic understanding of the decisions made by the users of data insights.

The most egregious, shockingly common, display of misunderstanding of how financial decision makers use data is data companies showing a chart with the correlation of an alternative data point directly to share price. If there’s one thing I can do to help data companies better position the impact of their product well, it’s the simple guidance that it’s extremely rare that one data point can, on its own, explain share price movements. So, please don’t try to show your data’s value that way. It’s a much more complicated relationship between single data points and share price movements. There are other ways to show the alpha of a data point (maybe a future newsletter post if your feedback suggests it’s a priority to understand this at a deeper level).

Product market fit is found by helping end users achieve the outcomes they need, invest in actually understanding this. This builds on realigning the data with outcomes. If the data product, analytics and insights are more directly answering questions that shape decisions, leaving less work to do for the end user to achieve the outcomes, it is easier to attribute the decision to the data insight generated by the data team/business.

Data teams and businesses need to understand the economic impact of the data on end user decisions.

• This is perhaps more straight forward for systematic end users as the alpha generation of the signal can be measured as an added factor. The data is either incorporated in the model or not, and if incorporated how did the performance change.

• There are attribution challenges on the fundamental side. The data may specifically answer an unknown question, and the answer to that question will shape an investment decision that generates alpha, but there is a shared attribution between the investment decision maker and the data business/team. It’s less direct, but the impact on alpha generation can be measured and converted to ROI measures.

• Stepping away from investing for a moment, data teams in the corporate world also have this issue. Understand how the business KPIs are related to the revenues, profits and cash flows of the business which ultimately drive the value of the business. This isn’t very different to the fundamental approach financial market participants would take when deciding the value of a company from the outside. Armed with more data on the inside as a data team guiding business strategy decisions, model the impact of the data on the overall business to measure the impact.

Build network effects.

There is not a technique or data source type that will be impervious to direct competition. I believe whatever data set or technique you feel is proprietary today, it won’t be in the long run. If your data product is solving a valuable outcome, substitute products will enter the market. Therefore, it’s critical to build network effects that generate the feedback loops and scale that allow the data business to remain relevant long term.

A common adage talked about in sell-side equity research is that good research leads to more questions, which leads to more good research. In that sense. valuable data insights lead to more questions, which leads to more valuable data insights. I believe this network effect has been the most important part of the UBS Evidence Lab business model. Great analysts are experts in early identification the key debates about uncertain outcomes, that once known would change the value of the companies and markets. Evidence Lab partnered with the analysts to understand the key questions, breaking them down into smaller questions which would be addressed with data driven insights not yet understood by the broader market. In turn, this would move the markets forward in their understanding, leading to new critical questions to understand and more need for insights driven by proprietary data.

I appreciate that most data businesses are not able to benefit directly from the networking effect a global bulge bracket investment firm generates in synergies with a data insights team. However, the client - provider relationships can provide the necessary feedback loop.

There are other network effects to seek.

• Becoming the consensus⁷ measure / table stakes for the outcome needed. In the beginning the alpha generation is the driver of adoption, but eventually if the whole market uses the data source and method it becomes the beta input that market participants need to have.

• Being part of data marketplaces that provide the ability for end users to access many dataset types for their insight needs. When both supply and end markets are fragmented, there’s network effects within the distribution channel that provide the breadth of supply and demand together. This seems to be the case for the data landscape with so many data insight providers and varied end users of data insights.

Finding opportunities to be in other types of network effects is critical. Being in the flow, sharing data, methodology and insight openly and proactively within data communities and organized networks, can kick start the network effects if managed thoughtfully.

TL;DR

I agree that not all data companies and data science teams will survive the trough of disillusionment. But that doesn’t change the long term story for alternative data insights as an industry. The data businesses that work backward from critical client outcomes, align their product with the value created by decision makers using the data and set up their businesses to benefit from network effects have the best chance of sustainable long term growth and reaching the plateau of productivity.

The newsletter is just the start of the conversation.

• Fellow data creators, which parts of the path forward are most challenging for you?

• For fellow insight generators and data scientists, where do you see the biggest blockers to generating a positive impact?

• For fellow insight seekers, what would help you more easily get value from data?

- Jason DeRise, CFA

1. “Bullish” and “Bearish” are financial market jargon for positive (Bullish) and negative (Bearish) opinions about what will happen next for an industry or investment.

2. For more info on the Gartner Hype Cycle^TM check out: https://www.gartner.com/en/research/methodologies/gartner-hype-cycle also note that “Gartner does not endorse any vendor, product or service depicted in its research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner’s research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose.”

3. More financial market jargon - Buyside typically refers to institutional investors (Hedge funds, mutual funds, etc) who invest large amounts of capital and Sellside typically refers to investment banking and research who provide execution and advisory services to institutional investors.

4. As a simplified definition, value investors prefer a style seeking underappreciated, under-valued businesses that will be more appropriately valued in the future.

5. More financial market jargon. Simplistically fundamental refers to institutional investors that leverage portfolio manager’s judgement and decision making to allocate capital (leveraging varying degrees of statistical, data driven analysis). Systematic refers to a quantitative (quant) approach to portfolio allocation based on advanced statistical models, and machine learning (with varying degrees of human involvement “in the loop” or “on the loop” managing the programmatic decision making).

6. Yes, more financial jargon - A simple way to think about alpha is that its a measure of the outperformance of a portfolio compared to a pre-defined benchmark for performance. Investopedia has a lot more detail https://www.investopedia.com/terms/a/alpha.asp

7. Last financial jargon term for today (or maybe I missed one because I’m so used to talking in financial jargon) “The Consensus” is the average view of the sell-side for a specific financial measure. Typically it refers to Revenue or Earnings Per Share (EPS) but can be any financial measure. It is used as benchmark for what is currently factored into the share price and for assessing if new results/news are better or worse than expected. However, it is important to know that sometimes there’s an unstated buyside consensus that is the better benchmark for expectations. We’ll leave that for another newsletter as a topic to demystify.