- Bloomberg LP is hoping to improve how it assists customers who want to use alternative data but lack experience with complex data sets.
- Naz Quadri, the head of alternative data and machine learning at Bloomberg, told Business Insider that included answering specific questions about alt-data usage for investment strategies via machine-learning techniques.
- Bloomberg is also helping clients make connections between alternative and traditional data sets.
- Click here for more BI Prime stories.
One of the largest data providers in the world wants to make it easier for investors to digest complex, unique data sets.
Bloomberg LP is working to help investment firms, such as hedge funds, use alternative data despite lacking the resources or experience often required to digest it. The efforts come just over six months after Bloomberg announced a move into the booming market for data such as stats on drug approvals, retail foot traffic tracked through cellphones, and construction permits.
Bloomberg is now working to make alternative data even more mainstream by making it easier for a wider variety of its customers to use the unique information as they try and find new ways to beat the market. That also includes helping customers draw connections between data sets, whether they be alternative, traditional, or both.
Naz Quadri, the head of alternative data and machine learning at Bloomberg, told Business Insider it takes a certain level of sophistication and resources to be able to digest the nontraditional data sets. Oftentimes, even if a client wants to test out using alternative data, it's too difficult for them to try.
"[They] don't have the investments already needed to be able to process these types of huge data sets that are complex, and maybe don't want to make the investments," Quadri said. "One of the things that we have been thinking about is how do we make complex things easier for clients to ask questions against."
From satellite images to credit-card transactions, alternative data offers users insights into companies and industries they don't typically get from traditional data. However, it's not a fool-proof process, as even the most advanced firms still sometimes struggle to get an edge using the data sets.
While originally a field dominated by only the most advanced quantitative hedge funds, in recent years, alt-data providers have helped to level the playing field for less-sophisticated firms. And now with the entrance of mainstream players such as Bloomberg and S&P, efforts to make the data available for an even broader clientele have begun.
For Quadri, it's about Bloomberg having the ability to go beyond just providing data.
"What we are thinking much more about is, how can I actually answer the questions a client may have using more than just that underlying data and just giving that to them?" Quadri said. "How can I do that for a variety of different kind of clients across the technical or complexity curve?"
Quadri declined to get into specific details but said the use of machine learning and natural language processing were key tools in how Bloomberg would be able to better understand which type of data would be beneficial based on a client's questions.
That's not the only work Bloomberg is doing in the alt-data space. It's already working on helping customers draw correlations between data sets, whether it's between multiple alternative data sets or making connections with more traditional data.
For the latter, Quadri gives the example of combining global and regional crude-oil-volume estimates via satellite imagery with fundamental data from filings to help predict where particular oil companies rank among their peers.
Connecting alternative data sets can be even more tricky, Quadri said, but offer big benefits. As an example, a customer could combine information on the upcoming drugs a pharmaceutical company is planning to release, such as which stage they are in, with how often specific drugs are being prescribed.
"Knowing what we have, make suggestions on the data sets that we think they should trial, and really have them understand how these things link together so they can minimize the amount of time it takes to figure out, 'Do I see value in all of these,'" Quadri said. "In a subset, how do I make a decision about what to move forward with?"
To be clear, Quadri said Bloomberg wasn't interested in offering advice on how clients should be investing, or creating some type of uber data set that should be used by everyone. Those efforts, he added, would likely be fruitless, as the industry is far too idiosyncratic to be able to generalize it.
Instead, he said he wanted to make it as easy as possible for consumers to find the appropriate data.
"Identify things that can go together, and then it is really up to the end investors to say, 'You know what, that makes sense, I should give that a shot. I found it actually gives me value,'" Quadri said. "We can get them to the point where with the smallest amount of investment they can make decisions on what work with them and what doesn't."