The entire industry of buying and selling ASX shares is based on predicting the future.
How will this business do in the future? Will it keep producing what people want? Will it lose customers to competitors?
To assist the average punter who doesn’t have time to work it out themselves, there is an entire sector of advisors and fund managers that can do the research on their behalf.
But time and time again, evidence comes back that, on average, even those who do it for a living don’t fare much better than the overall share market.
“Professional mutual fund managers fail to outperform the market most of the time,” wrote The Motley Fool’s Catherine Brock this year.
“Over the past 10 years through mid-2020, the S&P 500 Index (SP: .INX) outperformed 82% of large-cap stock funds.”
So it’s clear no one holds a crystal ball.
But this week a team of Australian academics released an algorithm that they claim can predict market movements for up to the next 2 days, based on an analysis of 10 pictures.
How 10 photos can forecast stock market sentiment
A team of researchers at RMIT and Swinburne University of Technology developed the algorithm, building on earlier work from the University of Missouri.
The programme takes in the daily top 10 news pictures from photog service Getty. Machine learning then analyses the photos to judge societal mood, which is extrapolated out to how the share market will turn.
Lead author of the study, RMIT’s Dr Angel Zhong, said one need not read hundreds of news articles to get a feel for where stocks are headed.
“You can get a snapshot of global investment mood by looking at the 10 most popular photos,” she said.
“This technology could have a huge impact for those wanting to get the feel of the day quickly and accurately.”
In general, bad news makes investors buy and sell “impulsively and intensively”.
“When the photo sentiment measure reflects a bad mood, it predicts a large increase in trading volume.”
While the earlier study was based only on US news and markets, the Australian team has broadened the algorithm to apply internationally.
“We can predict stock returns in global markets in 37 countries,” said Zhong.
“If you only look at the text of news articles, you often miss out capturing non-English speaking markets. Analysing images removes that problem.”
The team is now seeking strategic partners to further refine the algorithm to one day be ASX-specific.