What is traditional data?

Traditional data is a term which has been used in many different ways over the years. It encompasses all the datasets not considered alternative. Given Neudata’s role in the marketplace, we are in a unique position to bring a standard definition.

Traditional data can be viewed in two ways

• First, any source that is part of the “traditional roster” of data sources asset managers typically use in their standard market analysis.
• Second, any source that is widely accessible and structured, typically by convention or regulation.

What does this mean in practice?

Think about the data provided by the large stock exchange groups — on market pricing, trading volumes and other information relevant mainly to equities. This data meets both of the criteria above and falls firmly into the traditional market data bracket.

It is not just about stock prices; the rise of quantitative trading techniques and the importance of “momentum” as a factor in market moves means the underlying data provided by exchanges has never been so prized. But it remains traditional, because the data is available by virtue of regulation. All exchanges must publish their feeds in a given format to meet transparency requirements set down by securities regulators.

Another example of datasets sitting squarely in the middle of the traditional data sector are those provided by the big aggregators. These companies will aggregate data, much of it available by regulation and convention, and provide it in a central one-stop service.

User sophistication is not relevant

The users of this type of data might be using very sophisticated techniques and technology as they put it to work, for instance in their investment trading. But it is important to note that sophistication of the data, or the sophistication required to use it, is not relevant to the alternative/traditional data split.

For instance, the trading data sold by stock market exchanges gives incredible precision, sometimes to the nanosecond. You need sophisticated technology to deploy that data. But it remains traditional in type, because it is both part of the “traditional roster” of data sources that asset managers use for standard market analysis and is highly structured.

Traditional data vs alternative data

Traditional data differs from alternative data. We would classify alternative data as:

1. Any unconventional source of data.
2. Data that is less accessible and less structured than established sources.
3. Data that often involves crowdsourcing or web-scraping collection methods.

Take satellite and aerial data. Though its use among investment managers is growing in prominence as technology improves, it still remains an unconventional source of information. It is often less accessible and structured than the established sources of data you might get from a market feed. 

Alternative data does not have to meet all three of those criteria; for instance, satellite/aerial data is not crowdsourced or web-scraped. More than half of the providers on Neudata’s platform fall into one of those two groups, however, from online payments data to economic data.

A taxonomy for traditional data

With all of that in mind, Neudata has developed a taxonomy of traditional data categories and subcategories to help the industry understand the difference between traditional and alternative types of data.

At a primary level, the taxonomy for traditional data includes dataset categories like:

1. Company-reported data, such as fundamentals and financials.
2. Data management data, like security master files.
3. And market data, such as pure-play pricing and reference data.

Some categories fall right on the border between the alternative and the traditional data universe. Examples include flow and short interest data, which may form part of the traditional roster of data products procured by asset managers, but are often collated through less accessible sources and crowdsourced models.

Given that such products are typically sourced alongside all the other traditional sources, we will include them on the list.

We’ve compiled a list of traditional data categories, along with examples of datasets that would fall into each bucket.

• Company reported data (such as financial statements you might find on Companies House or corporate event and filings notifications from listed companies).
• Market data
• Flows (such as fund, asset or retail flows).
• Human capital (such as information on directors’ dealings in their own company stock).
• Risk & compliance data (such as information on global sanctions or cyber risk).
• News (financial reporting coverage).
• Data management (such as information on corporate hierarchies/structure).
• Business relationships (such as information on supply chain relationships).
• Economic (such as commodity market data).
• Private markets (pricing data in non-public markets)
  

Areas of innovation

Traditional data has been around for much longer than alternative data, with some exchanges starting to sell their data in the last century.

But developments, most notably in asset management, mean there is growing demand, and also more use cases.

There's been a lot of talk about the growth of systematic credit trading recently, which is forming an ever-greater part of the fixed-income market.

For systematic macro and credit funds, traditional data is an integral part of their trading systems that they’ve relied on for many years. But while shares have been traded on national stock exchanges, bonds are mainly traded over-the-counter (OTC), which means the associated data has been harder to come by.

The largest OTC trading venues are now beginning to evolve their approach by making a big push into the data market, following the traditional exchanges. They have realised they have so much data and a large audience to sell to. The venues have therefore been making big investments in data in the past few years, expanding their reach to highly sophisticated parts of the traditional data market.

Fund managers say the rapid electrification of corporate bond markets has opened up more quant approaches in fixed-income. Improvements in execution were described as the “main game-changer” in this area by Paul Kamenski, co-head of credit with Man Group’s Numeric unit, this month.

After moving from voice execution to portfolio trading to single-name electronic, the fourth stage in execution advances will be algorithm development, which Kamenski sees on the horizon.

‍Conclusion

This piece has set out to define the parameters of traditional/market data. In one way, that is straightforward: we define it as anything not falling under alternative data.

It comes down to the conventionality of sources, how accessible/easy are they to find and whether they are available by regulation. The more unconventional the source, the harder to procure, the less available by regulation, then the more likely they are to be alternative.

The UK Company's House register of corporate filings is a good example of the reverse. The information is freely accessible and available because of regulation. Services which aggregate that data are traditional providers.

There could be a more complex product which uses the filings we see on Companies House and derives some signals off it, which a hedge fund then uses. That would fall under alternative data.

Now we have introduced traditional data, the rest of this blog series will explore the pain points users can face and the opportunity for improvements in the future.

‍

This is the first blog in two-part series a series of traditional/market data topics from Neudata. Neudata has launched Neudata Ranger to address the needs of users and vendors in the traditional/market data space.

‍

About Neudata

Neudata helps institutional investors, corporations and leading global organisations find the most relevant data sources to use in their internal data ingestion processes. Neudata offers a data research catalogue with objective and neutral assessments of alternative/traditional data vendors and datasets.

Neudata doesn’t buy or sell data, or require data providers to pay a revenue-share or commission in exchange for recommending their products to data buyers. That means users get unbiased intelligence that’s tailored to their specific research goals and strategies.

It also means Neudata is ideally placed to consult on data monetisation and help companies create new revenue streams in this field to future-proof their business in a tech-driven world.

‍