Opinions expressed by Entrepreneur the contributors are theirs.
Alternative data is no longer a secret, especially in the financial sector. Investment firms, hedge funds, venture capitalists, etc. have always had new and unique data sets in mind. A concrete example of this upward trend is ESG data, which seems (so far) to provide extremely powerful signals for investment decisions.
While there are many alternative forms and forms of data, one of the most frequently used and studied is sentiment analysis. Web scraping is relatively easy to access these days, which makes acquiring these datasets inexpensive and efficient.
Sentiment analysis, for example, has already been shown to potentially generate valuable insights. With the hype surrounding data, however, it’s easy to be misled by the supposed potential it holds.
Related: Understanding Market Anomalies Through Alternative Data
Sentiment and other alternative data types
Few articles have discussed the concept of sentiment in depth. In most cases, it is the automated processing of content (usually text-based) to derive emotional states and judgments. Advanced tools like Google’s natural language AI can uncover sentiment, assign entities (objects being talked about), and more.
Emotions and investor sentiment have an obvious effect on the stock market. The momentum effect is a prime example of medium-term market inefficiencies caused by the emotional background of investors. The two forces underlying the momentum effect, under-reaction and over-reaction, are essentially based on emotional states.
One important caveat that I rarely see mentioned, however, is that emotional states don’t last long and can only provide predictions for extremely short windows of time. In fact, as I intend to show later, most, if not all, alternative data captures a short-term forecast (less than 5 years).
Investors and companies
Overall, stock prices are influenced by investors and issuers. The former have a direct impact on the price, as they drive demand. The latter have an indirect effect, because the performance of companies has an impact on the perception of the investor and the valuation of the shares.
There are also potential black swan events that could upset the balance at play. I won’t discuss them, however, as it would add many unpredictable variables to my hypothesis without providing much additional value.
Companies have an indirect impact on prices in the short and long term. Current cash flow, consumer sentiment, marketing strategies, etc., can impact quarterly or annual earnings. Business strategy, growth options and other factors influence long-term performance.
Xerox has always been an example of how a changing business strategy and competitive landscape can dramatically impact an organization’s performance. It has both held dominant market positions and struggled for decades.
Since company performance and stock prices are intertwined, savvy investors need to consider both short-term and long-term prospects. Alternate data, however, mostly contains information about the former.
Related: How to Forecast Revenue and Growth
How far can alternative data go?
Various sources of information can be called alternative data, because they are very loosely defined. However, it will mainly be information retrieved from the web, satellite images and credit card data, as these are among the most used sources.
Each of these sources, whether collecting sentiment from public social media posts or analyzing cash flow through credit card data, indicates the current state of the business. In other words, it is a reflection of current performance. Also, these springs are not meant to last long. These are snapshots of a decidedly short period in the company’s history. Data must be collected over a longer period for any type of forecasting to be made.
Yet they cannot reveal anything about the abstract approach to business. No strategy can be revealed. If a company or organization pivots and changes direction, none of this information can be gleaned from alternative data sources.
They can, however, reveal the effectiveness of these pivots or changes. As such, alternative data can provide valuable information about a company’s expected short-term performance, but not much else.
We have evidence that financial services organizations have started using alternative data for, at least, similar means. A survey by Censuswide and Oxylabs showed that financial services organizations rank web scraping, an alternative method of data acquisition, as one of the most effective in generating revenue.
Related: Top 4 Cash Flow Forecasting Mistakes
Alternative data has a lot of buzz around it. There’s good reason to be excited, especially if you’re involved in the financial industry. This is a whole new source of data that can provide a competitive edge.
I would, however, caution against jumping in headlong. Alternative data is surprisingly good, by its nature, at providing signals that show short-term performance. It’s the same nature that keeps it from being as useful or useful at all in long-term performance predictions.
In the end, most alternative data hides signals about short-term performance. As such, it should be used for exactly that reason alone. Trying to apply longer-term signals, such as five-year forecasts, could even be detrimental.
Finally, the alternative data must also be weighed against the opportunity cost. Extracting signals from them requires more work than with traditional datasets. Working with alternative data means diverting resources from other potential strategies.