The investment community is starting to gorge on richer sources of technology-generated data to inform its decision-making, as Hilary Eastman explains
Studying this technical article and answering the related questions can count towards your verifiable CPD if you are following the unit route to CPD and the content is relevant to your learning and development needs. One hour of learning equates to one unit of CPD. We'd suggest that you use this as a guide when allocating yourself CPD units.
This article was first published in the May 2018 UK edition of Accounting and Business magazine
If you ask a fund manager or a research analyst at an investment firm about what information they use to make or advise on investment decisions, you might get an answer that will surprise you. What used to be a conversation primarily about their analysis of audited financial statements now often turns into one about artificial intelligence and ‘alternative data’ (non-financial corporate information created outside a company, and not reported by that company).
The proliferation of data, combined with the internet and social media as channels for distributing it, means that the investment world is changing – and quickly. And in their attempt to gain an advantage over their peers (and the market), investors are eating it up.
Investors are looking at the impact of technology from two perspectives. First, they want to know how they can assess whether a company’s business model will be disrupted, and how and when that might happen. Second, they are looking at how they can improve their investment analysis by using these new forms of data – particularly if that is something their peers are already doing.
This isn’t happening just with the quants (investment funds that select stocks using quantitative analysis) and hedge funds; mainstream investors are starting to use the growing sources and types of data to inform their decision-making, gain insights into the fundamentals of the business, and understand market and economic trends. They can also triangulate relationships between the data to discover the story the data is telling.
Many wonder what the fuss is about – after all, the use of computers in investment analysis is not new. But there are three changes coming together that were not there before: more data is being generated, more processing power is available, and there is more data storage capacity. Those factors, combined with the power the internet brings, are changing how people are able to make decisions. That includes decisions by companies (eg about operational efficiencies and optimal pricing levels), and by shareholders and other stakeholders (eg about whether they want to invest in, buy from, partner with or work for that company).
Of course, there is the question of how much the investor can trust the data being generated and the algorithms processing it. Whenever there is new information, there is a new need for trust in that information.
The issue of trust becomes particularly important as the audited financial statements become a relatively smaller component of the investment decision. So far, most people hesitate to predict there will be no need in the future for audited financial statements. But some do go as far as to say that if other metrics become reliable indicators of performance, they will shift toward using that information instead. And there are some who say the volume of data they use is so large that the reliability of each data point is not an issue – as long as enough of it is right overall.
The key, then, seems to be making sure that the material (or most relevant) data points are reliable, or at least that none of them is so unreliable that it messes up the model and the results. Also, investors who are using big data find the accounting data useful for corroborating their model’s results to fine-tune their algorithms. Even a company’s boilerplate disclosures can be useful if an analyst uses natural language processing to ‘read’ them to see if something has changed or if they differ to any great degree from other companies’ disclosures.
There are many examples of how investors and others are using alternative data, such as satellite images of shopping mall car parks. And the sources of such data are growing – there is geolocation data from mobile phone apps, sensor data from shipping containers, and weather pattern data, to name a few. What’s notable is that investors are not using this data to predict a company’s sales or profits, but to try to see whether the data tells them something different from what the market is predicting the company will report, and to inform their knowledge of whether something is going on in the business that may mean the investor should sell or buy more shares.
For this type of data analysis to work, though, the key is the linkage between what the data is saying and how stock prices are derived. Not only are data and algorithms useful for those wanting to find profitable trading patterns, but they can help other types of investor understand the business’s fundamentals better – and expectations about those fundamentals. Such investors are letting the algorithm do the heavy lifting so they can make decisions about the data without having to sift through it all to make
sense of it. Companies therefore need to think about what data is out there about them, their relationship with third parties, how information about them will be consumed and used, and what conclusions people – or computers – might draw from that.
We seem to be in a twin-track phase. Some people are working to improve the reporting model, with changes to financial reporting standards, new non-financial reporting standards, etc. Others are working to find ways not only to make more informed and better decisions using corporate reporting data, but also to supplement it with other data and to use technology to analyse it. At some point those tracks should converge, but it takes a lot to change the regulated reporting model, and radical changes are unlikely – until we reach the point that it is inevitable.
Clearly, something is happening and change is coming – although it is hard to say just yet what our corporate reporting model will look like in the future or how quickly we’ll get there.
Hilary Eastman, head of global investor engagement, PwC
CPD technical article
"Investors are letting algorithms do the heavy lifting so they can make decisions about the data without having to sift it all to make sense of it"