In this news, we discuss the COVID-19 pandemic deals body blow to quant models, study shows.
LONDON (Reuters) – The coronavirus pandemic has taken a heavy blow to the style of investing based on a quantitative model, with the majority of companies using such strategies having a negative impact, a Refinitiv study has found.
In a report, financial data provider Refinitiv said 72% of those investors had been affected by the pandemic. About 12% declared their models obsolete and 15% were building new ones.
Machine learning refers to the use of complex mathematical models and algorithms based on historical data in order to make predictions without being explicitly programmed to do so.
While these auto-powered models have been successful in the past, with historical correlations between different asset classes held firm, they have suffered in the aftermath of the pandemic as those links were severed.
These quantitative models have also suffered in 2020, as the amount and complexity of inputs that go into these algorithms to generate trading signals has exploded in recent years.
“COVID-19 presented a significant shift in most market dynamics and many institutions would have had to revisit much of the models they had in order to cope with extreme market events,” said Amanda West , Global Head of Refinitiv Laboratories at Refinitiv.
A majority of respondents said the main areas of focus over the next two years in data strategy will be to extract more value from data and accelerate processing speed. The average size of data science teams in companies has more than tripled, from 2.7 in 2018 to 7.1 in 2020, according to the study.
The survey was conducted through 423 telephone interviews with senior executives and data science practitioners at various financial services companies between June 29 and August 14, 2020.
Machine learning has long been the mainstay of deep-pocket hedge funds, which have combined complex algorithmic strategies with financial data to make big bets in the markets.
But the coronavirus pandemic has accelerated the adoption of new technologies in the financial sector, although the lack of quality data will be the main distinguishing factor between companies in the years to come.
The number of companies that only use unstructured data climbed to 17% in 2020 from 2% in 2018, while only 3% of companies surveyed said they did not use alternative data sources compared to 30% in 2018.
“Those who have instituted careful data governance processes are much more likely to be successful in this game than those who haven’t because trash is trash in the world of machine modeling,” he said. declared Refinitiv’s West.
Reporting by Saikat Chatterjee; Edited by Hugh Lawson
Original © Thomson Reuters
Originally posted 2020-10-25 18:56:10.