Quants must think like amateurs in world’s wildest market

Bloomberg (subscription)

Mom-and-pop investors drive more than 80% of mainland China equity trades, so it’s important to understand what drives such ‘amateur’ investors. In this article, quants share how they use social sentiment and other alternative data to gauge the market’s direction. Read More

How has machine learning changed the game for Goldman?

Quantminds

ML has changed Goldman Sachs’ advertising material, but not much else, says Michael Steliaros, Global Head of Quantitative Execution Services (though eFinancialCareers thinks he may be trying to talk down job applicant expectations). Nevertheless, in this video the firm’s top quant remains measuredly optimistic, noting that his team can carry out calculations that were impossible “even two or three years ago.” Watch Here

Alternative data firms attract M&A, private equity, interest

Forbes

Companies that specialize in making and distributing unique datasets have attracted strong equity investment and M&A interest from strategic and private-equity investors. Mergermarket looks at potential deals and tie-ups. Read More

The AI sharecroppers

Axios

Without humans, AI can’t tell between a skyscraper from a spider. To give algorithms power, an invisible underclass of low-wage workers are painstakingly making inventories of millions of pieces of data. Read More

Alternative data in fixed income, real estate, FX, and crypto

Various

When looking at investors’ use of alt data and technology we often focus on equities, but several articles this week looked at applications in other areas: Institutional Investor asks if ML will replace active bond managers, Estates Gazette notes useful alternative datasets for commercial real estate, Cuemarco looks at alternative datasets for forex, and Bloomberg takes an alt-data dive into Bitcoin.

Learning to love the AI bubble

Sloan MIT Management Review

Are we in an AI bubble? How bad would it be if it were to burst? Boston Consulting Group says if there is a bubble it has more in common with the dot.com bubble, which helped finance the internet backbone, rather than the housing bubble behind the 2008 meltdown. Read More

Peering into peer selection: quantifying company similarity

MSCI

Is it possible to identify and measure relationships between companies from a systematic analysis of various types of data, including fundamentals and company filings? This MSCI paper looks at whether alternative data can be used to get a better understanding of inter-company relationships and risks. Read More

Quartz

Data show Game of Thrones is taking a historic critical nosedive

Examining IMDb data on 82 television shows that lasted at least five seasons, Quartz found that the show’s nosedive is nearly unmatched, with only one other former critical darling having a sharper drop. Read More