How Transparently.AI helps you anticipate short sellers

Posted by Hamish Macalister

The Transparently.AI platform can help equity fund managers identify and avoid companies that might be targeted by forensic short sellers. It can assess whether or not a specific report is likely to be impactful. Short-selling reports can be erroneous. When they are, they usually provide an excellent medium-term buying opportunity.


Forensic short sellers represent an important risk to traditional fund managers


Forensic short sellers represent an important risk to traditional fund managers. They borrow stock from a manager, short the stock, and then issue a report with a view to driving down the stock price. The fund manager’s performance is hurt and their reputation is harmed for having invested in a potentially fraudulent company. It’s a bit like borrowing a neighbour’s car to push it off a cliff so you can post the video on Instagram.


Forensic short sellers had a tough time during the super low interest-rate environment of 2020 when it appeared that the Fed would never allow interest rates to rise again. In the subsequent equity frenzy, inexperienced retail investors formed vigilante groups on social media to buy stocks targeted by short sellers. By early 2021 some, such as Citron Research, closed their doors while most, like Muddy Waters, curtailed their activities.


With interest rates having returned to normal levels, the forensic short sellers are back in business.


Fund managers, with good reason, need to be wary. Our analysis of the historical calls of the three largest forensic short sellers suggests that the overwhelming majority of their calls – more than 90% – involve companies with a high to extreme risk of accounting manipulation, based on our scoring system. In other words, our platform suggests that the more famous short sellers do their homework well.


Our platform suggests that the more famous short sellers do their homework well


What about smaller forensic short sellers? While we have only examined a handful, our findings thus far suggest that these firms tend to short companies with less extreme manipulation risk scores than the more experienced players.


We recently examined the track record of a lesser-known short seller that has issued 10 short-sell reports since the start of 2021. Only two of this company’s shorts had an accounting manipulation risk score above 40%. Our analysis suggests that successful short-seller recommendations typically have a risk score of at least 40%.


Seven of the 10 short sale reports issued by this company have a track record of at least six months. The below chart illustrates the return of each of these short-sell recommendations in the six months subsequent to report being issued. We plot this six-month return against the respective accounting manipulation risk score of each company. The risk score is that generated by our platform for the year prior to the short-selling report. In other words, the Transparently.AI score was available before the short-selling reports were issued.


Chart comparing Transparently's manipulation risk score to stock returns after a short seller's report.
Source: Transparently.AI, Bloomberg

The most successful short fell 50% in the six months following the report. The least successful short gained 16%. Four of the seven shorts lost money.


Remarkably, the six-month return exhibits an 88.2% correlation with the Transparently.AI manipulation risk score. This result is similar to the relationship we found in the other short sellers we have studied.


The link between our risk score and share-price performance appears to be enhanced after a short-sell report is issued


The evidence suggests that  the link between our risk score and share price performance is enhanced after a short-sell report is issued, presumably because investors become more focussed on accounting manipulation risk after a short seller report is issued.


We draw three conclusions in relation to the application of the Transparently.AI platform to forensic short sellers:

  1. First, the platform has a strong track record of anticipating the short-sell reports of the seasoned forensic short sellers
  2. Second, the short-sell reports of smaller players tend to include more companies where the risk of accounting manipulation is more marginal
  3. Third, when smaller players issue reports, our risk score typically gives a good indication of the likely impact of the report on the stock price of the targeted company. If the risk score is above 40%, the report is likely to have a significant detrimental effect on the subsequent 6-month return of the company.