Investment in an era of disinformation

At a time when facts are seen by some as blocking specific agendas, it’s difficult to know what and who to believe when faced with the huge amounts of data that permeate our daily lives. This is an increasing challenge for everyone, but specifically for investment analysts who rely on credible information to make investment decisions.

It is critical to differentiate between what is accurate data and what isn’t. It may be that if 10GB of data is being observed, only 1MB is credible and informative in what is being analysed. The Covid-19 pandemic has worsened this distortion of data and, coupled with the sheer volume of data investors need to assess today, the industry has to adapt to navigate this tsunami if it wishes to achieve its investment objectives.

Future generations may look back at Covid-19 and the global vaccination programme as a tipping point in the battle between truth and fake news, as it compounded a trend already set in motion during the 2016 US elections. Caught in the grips of disinformation campaigns and wild conspiracies about the virus, it’s become almost impossible to tell fact from fiction.

Our descent into a world where you can’t believe what you’re told has been swift and severe. Curiously, from a behavioural finance perspective the truth of the information itself matters less than the reaction to the (dis) information. We gain valuable insights by interpreting and observing the behaviour of investors to the (dis) information that is contained in the mass of data.

Behavioural finance is the study of the influence of psychology on the behaviour of investors or financial analysts. A core premise is that investors are not always rational, have limits to their self-control, and are influenced by their own biases. This helps explain why markets are inefficient. As such, behavioural finance underpins many quantitative investment strategies and sits at the heart of our own quantitative investment process on the basis of striving to understand how people respond in different circumstances.

Our departure point when tasked with managing clients’ capital is not determined by what we think should occur or what we think is logical and rational. What is most important is how people interpret and react to the data, what they do with it, and how they behave in making a buy or sell decision based on the data. From here, our first objective is to try to understand people’s sentiment, moods and how they make decisions. Even if you had perfect foresight for every company’s future earnings 12 months forward, this is meaningless unless you know how the market is going to react to the information. Our investment strategy in the quantitative space is far less about predicting far into the future — because as we’ve seen in the pandemic, the future is anything but certain — than assessing how sentiment and the behaviour of market participants on the road to the future are going to affect share price movements.

With a quantitative investment process, which may be aided by artificial intelligence, we are able to extract meaningful information and insights from the standard data published by companies such as financial statements. This allows us to make informative investment decisions, even using the same data set that others might be using.

Speculators are reacting to live news feeds and making decisions in real time. This is specifically relevant to the buying and selling of shares — these investors react immediately to the data that is made available. Our investment strategy applies an investment process to quantify the sentiment and the behaviour of people to these news feeds. To manage this, tools such as machine-readable news and machine-learning algorithms are constantly processing live feeds. These algorithms then analyse the text and type of words to determine if sentiment is net-positive or net-negative around any given area of interest.

We analyse this data with the objective of interpreting investors’ behaviour. From this we are able to determine when there are opportunities to gain from such behaviour. These insights help to inform our decision-making. They serve as an early warning system, which we came to appreciate when Steinhoff collapsed in 2017, and we managed to sell down our holdings and eventually exit the position completely based on the data our systems had analysed, thereby avoiding any substantial losses.

A key component of this approach is to not bet on only one outcome or expectation, but rather focus on a tolerance about the expectations. Consequently, a diversified portfolio is the solution to our particular investment strategy. The process should be data-driven, from which informative and objective investment decisions result.

Fake news continues to pose a risk to the navigation of everyday life, but it needn’t be a challenge for investors. With the value of the growing volume of data being produced by the world only set to rise, technology is going to play an increasingly crucial role in general investment decision-making going into the future, expanding beyond specialised investing such as our quantitative approach.

Ultimately, I believe this could lead to a blend of art and science emerging. While investing is an art rather than an exact science, you can use science to gather information and process it to the best of your ability and gain better value.

• Watson is a senior portfolio manager at Old Mutual Investment Group.