Aspiring coronavirus models are also spreading

Coronavirus modeling as a hedge play and a weapon

With the spread of coronavirus now creeping into the financial markets psyche, amateur and semi-professional modelers are ramping up their efforts by leveraging social platforms and access to media opportunities to forecast often apocalyptic scenarios based on questionable data.

Most often, the people pushing their scenarios are equity and hedge fund managers using home-brew models that often diverge wildly from current traditional scientific forecasts.

At the fringe, some are employing social media, such as Twitter, to amplify their theories and “model outputs” that target a particular stock. They often incorporate a short thesis for particular companies that have a heavy reliance on Chinese supply chains, such as Tesla.

Hedge fund message boards and online communities are a fertile playground for coronavirus model chatter, including the ZeroHedge post that threw into question traditional model estimates.

Is the epidemic truly far less serious than conventional epidemiological models predicted? Or is China merely hiding the full extent of the problem?

However, other amateur modelers are making their way into more traditional media, including Forbes, where they gain notoriety (albeit with some skepticism) when describing the methodologies that predicts 2.5 Billion coronavirus infections and 53 million deaths.

"I started with day over day growth,” he told me, using publicly available data released by China. “[I then] took that data and dumped it into an AI neural net using a RNN [recurrent neural network] model and ran the simulation ten million times. That output dictated the forecast for the following day. Once the following day’s output was published, I grabbed that data, added it to the training data, and re-ran ten million times.”

More traditional equity investors are also trying their hand at virology, especially Asia funds like Hong Kong-based Gavekal Capital which told the Financial Times that it published a “quantitative model” that showed the spread of coronavirus would accelerate in the coming weeks before peaking “in early March.”

Mr Darcet said the model used in the analysis was “unsophisticated” but added that it “provides a yardstick with which to assess the day by day evolution of the outbreak”. If the total number of infections exceeded 52,000 by Sunday, his forecast would be “reassessed and scaled up significantly”.


Below is a list of what RMN paid subscribers read this week…

Underwriting "Three-Peat"

Aon's Reinsurance, Cat Model Bundle

Consider becoming a full subscriber at the special offer rate to get access to daily emails and exclusive features, along with discussion access.