Earning calls being targeted for vagueness
One of the '' parents develop with young children is vagueness. 'Mum, dad, can we go swimming on Saturday?', chirp the little ones. Our reply , when we haven't decided yet, is often vague, 'we'll see'. They didn't say 'yes', they didn't say 'no'. They didn't say, 'stop' and they didn't say 'go'. Children quickly work out that vagueness means indecision and uncertainty
It seems that vagueness can now be picked up by algorithms searching for vague company executives remarks. So, all you comply execs out there, 'watch out'. It's not only your children who can detect our decision.
In the Middle East's Financial Times paper edition last Friday it was reported that a study by Bank of America Merrill Lynch found that using algorithms to decode the dry quarterly calls to decide on whether to buy or sell a company's stock can add an extra 6.3% to annualised returns. The technology is known as 'natural language' processing and outperformed a basket of 2,500 stocks over a 9year period.
The article underscored that this kind of approach is part of a wider move onto what is called the use of 'alternative data'. This is using things like satellite images of number of visitors to a shopping mall in order to use that as a trading signal. In 2017 the total spend on alternative data was $446m in 2017. By the 2023 that is expected to double to around $900m. The $2.6tn fund Fidelity investments, for example, is using 'alternative data' to inform a portion of 170 strategies that the firm launched internally last year with $500m of it's own capital. Alternative data is growing with Nasdaq announcing a deal in December to acquire the leading alternative data provider Quandl and Bloomberg has been looking at to have several alternative data providers on their terminal.