Making investment decisions with alternative data
How can asset managers gain an information advantage in today’s world? One way is to use unstructured data. They provide valuable signals and can thus offer an important investment advantage.
Google registers around 5.5 billion search queries worldwide. Per day. On Twitter, more than 500 million tweets are sent daily, that’s more than 6,000 tweets per second. And in 2021, the number of Wikipedia articles in all available languages was around 57 million. These are just three examples that impressively show how unimaginably huge the amount of data has become – and it continues to grow.
According to calculations by the market research company International Data Group (IDC), the amount of data generated worldwide was about 64 zettabytes in 2020. To put this in perspective: a zettabyte is a trillion bytes or 1,000,000,000,000 megabytes. According to IDC estimates, data volumes will continue to grow strongly, by almost 25 percent per year.
“We expect more and more unstructured data to find its way into active fund management.“
Dr Markus Ebner
Head of Multi-Asset
IDC experts predict that around 80 percent of all global data will be unstructured by 2025. This data, also known alternatively as Big Data, is unstructured text or non-financial data that can be found everywhere in the digitalised world. The big advantage: once systematically aggregated and analysed, it can provide valuable insights into the growth prospects of individual companies, sectors, countries or topics.
“We expect more and more unstructured data to find its way into active fund management,” says Dr Markus Ebner. The Head of Multi-Asset and his colleagues at Quoniam Asset Management are among the very few experts in Germany who have been collecting and analysing unstructured data for years and using it for investment decisions.
In search of more stable and faster information
Evaluating and analysing data in the financial sector is not new, on the contrary. But until now, the majority of data has been structured. Increasingly, however, the aim is to tap into new, possibly unstructured data sources in order to obtain more diverse, more stable – and also faster – information during the analysis. The prerequisite, however, is that the data contain useful information for the respective issue and that this information is sufficiently valid.
Unlike conventional macroeconomic data and the well-known financial market figures, unstructured data can, for example, provide indications of the underlying mood of a message. For example, how does the number of words with positive connotations compare to the number of words with negative connotations? In other words, is the news more positive or negative? Sentiment indicators can thus be derived from the articles that capture these dynamics and thus provide important signals for asset management.
For this purpose, the search programmes of special data providers continuously scour news sites worldwide or even social networks such as Twitter and Facebook. They count predefined terms such as “recession”, “inflation” or “consumer prices” and thus provide a real-time picture of the mood of consumers and investors. The result is a lot of useful information that cannot be found in the balance sheet or past share prices.
“In my opinion, processing an ever-increasing amount of available information can only be done sensibly with quantitative methods. And it takes a certain company size of asset manager to be able to do all the analysis at all,” says Ebner and sees his house at an advantage here. After all, Quoniam traditionally uses a large amount of data from different sources. In any case, the asset manager has always been data-driven, which means that the experts continuously and systematically analyse all relevant company data and then transfer them into investment models.
“Our Multi-Asset Global Data Sentiment strategy is based on modern, data-based research technologies such as Big Data and News Flows.“
Dr Markus Ebner
Head of Multi-Asset
Investor sentiment drives prices
“Our internal studies show that there is a causal link between media reports and the financial markets. Incorporating news sentiment into trading strategies can increase portfolio returns,” Ebner is convinced. This means: Not only facts, but also the classification of facts by investors determine financial market development.
Based on these findings, Ebner and his team have developed a multi-asset strategy that relies purely on the evaluation of unstructured data for return signals. “Our Multi-Asset Global Data Sentiment strategy is based on modern, data-based research technologies such as Big Data and News Flows,” explains the portfolio manager. For the investment concept, the Quoniam experts extract the sentiment on the capital markets from a multitude of global news. The evaluation is carried out on the basis of statistical methods that help to identify return opportunities.
Diversifier for the portfolio
The strategy went “live” at the beginning of 2022 with a 50 million euro mandate. However, the approach has already proven itself in the past years as part of another quant strategy. And practice has shown that the strategy has a very low correlation to other investment concepts. “The approach is therefore an interesting diversifier for the portfolio,” says the Head of Multi-Asset.
This is another reason why he is certain: Big Data – the evaluation of gigantic amounts of data – can become the decisive success factor in asset management in the long term. Those who are able to use the flood of information to generate reliable models for systematic investment decisions will be ahead in the long run and offer investors decisive added value.
