Moving into the Alt Data Universe

Apr 10, 2018

Topics: Alt Data |

As part of the never-ending quest for alpha, investment management firms are increasingly using alternative data to inform their investment decisions.

Alt data, or non-traditional sources of data, is often fed into other technologies such as machine learning solutions. At the moment, there are three major types, as classified by JP Morgan in a research report published in 2017:

  • Data generated by individuals – social media posts, product reviews, search trends, etc.
  • Data generated by sensors – satellite image data, foot and car traffic, ship locations, etc.
  • Data generated by business processes – company exhaust data, commercial transaction, credit card data, etc.

It’s possible that as the alt data trend develops, the types of data sets will increase, or new data sources within these three groups will evolve.

Regulation is spurring on the push into alt data. Recent rules on both sides of the Atlantic Ocean – and in particular, MiFID II’s research-related requirements as well as the US Securities and Exchange Commission’s (SEC) restrictions around the use of material non-public information (MNPI) – are making it more difficult for investors to generate alpha from traditional sources of information. As a result, investors are looking to alt data to identify trends, ideas, or signals that could help produce the performance that will allow an asset management firm to stand out from its peers.

Unsurprisingly, dozens of RegTech firms, both established and newly-launched, are moving into this space. Some focus on new ways to harvest the data generated by individuals on the web, and then analyze that data to produce tradeable insights. Others are focused on structuring the data produced by sensors into similar, tradeable information.

The use of the third category of data (data created by business processes) is also rapidly gaining in popularity. According to Forrester’s 2017 Data and Analytics survey, there was a large jump in the number of companies that are turning their data exhaust into a new revenue stream: 48% of respondents said they were commercializing their data, which is up from 32% in 2016.

Investors are using this data in a wide variety of ways to support the research process.  For example, credit card data could point to consumer spending trends, or indicate the performance of a retailer long before official quarterly sales figures are published. Other asset managers are mining alt data to do intensive analysis on a particular company. Schroders, for example, used a team of 27 data scientists to determine how many branches of Ladbrokes and Coral – two UK betting shops – would have to close when the firms merged in 2016. Schroders’ analysis showed that more shops would have to close than the firms originally indicated.

With all the hype around alt data, it’s hardly surprising that there are legal and compliance concerns about the use of this kind of data. Asset management firms need to be certain they are using legal sources of alt data in ways that do not violate insider trading, privacy, or market manipulation rules. Potential questions advisers should be asking themselves when considering the use of alt data include:

  • Is there a firm-wide approach to alt data? – What has senior management and the board said about the use of alt data? How would its use align with existing firm priorities, policies and procedures, as well as its risk appetite?
  • How was the alt data obtained? – The same laws that apply to traditional information sources apply to alt data as well. If the data is obtained in a manner that is considered improper or deceptive, then steer clear. If in doubt, ask compliance or your general counsel.
  • Who is the supplier? – A wide range of firms – from well-known data giants to small RegTech start-ups – are now supplying alt data. It’s important to ask about the source of the data, review it periodically, and monitor the activities of the vendor on an ongoing basis. Be sure all of the normal vendor onboarding checks are performed initially and on an annual basis. If the vendor is a start-up, find out if additional due diligence is a good idea.
  • What does it say in the contract? – Does the contract specify that the data came from a legitimate source, that the vendor did not violate any internal policies, and that it is legal for the vendor to sell the data? Get the vendor to supply a written explanation of how the data is obtained and why it is authorized to sell it.
  • What will the alt data be used for? – It’s possible to use good data to make decisions that are not compliant. Firms should make sure that their intended use of alt data complies with existing regulatory requirements.
  • What does compliance think? – In today’s regulatory environment, significant outperformance will generate scrutiny. Compliance should be comfortable with the source and use of alt data, and may need to review the data periodically. And if new compliance processes, or controls, need to be put in place to use the alt data, this should be discussed up front. Lastly, document, document, document! It’s important to document the analysis of the issues, due diligence of vendors, and any steps taken to address the firm’s use of alt data.

It’s hardly surprising that regulators are watching how the alt data space develops closely. For the most part, the approach they are taking at the moment aligns with how they are managing regulatory issues around other hot topics, such as cryptocurrency. If something is prohibited now, it remains prohibited when using new products, tools, or technologies.

One area to take particular care in is when using data the company has generated itself – such as its own data exhaust, or data from another division. In one enforcement case, the SEC prosecuted two employees of Capital One for using credit card data to buy and sell stocks. Both returned to China when they were fired by Capital One, and one of the pair was fined more than $13 million by the regulator.

In short, alt data presents tremendously exciting potential for investment management firms to generate alpha. However, firms should use alt data responsibly and intelligently. A good starting point would be to discuss the use of alt data at the board or senior management level, and develop a strategy suitable to the firm’s overall goals. The next step is of course to speak with compliance, to gain a better understanding of what controls or processes should be put in place to keep the firm on the right side of the regulators.