Riskified (RSKD -1.33%) is a fintech company that uses artificial intelligence to help businesses predict and prevent e-commerce fraud. The company went public during the IPO frenzy of 2021, initially commanding a $3.3 billion valuation. But the stock has since sold-off sharply, fueled by weak third-quarter results and the prospect of rising interest rates, the latter of which tends to blunt economic growth. Riskified currently trades 81% below its all-time high.

However, some analysts are forecasting significant share price appreciation in the near term. For instance, Josh Beck of KeyBanc puts Riskified's price target at $25, implying 260% upside. Better yet, Terry Tillman of Truist Securities puts the price target at $35, implying 405% upside. Of course, you should never build an investment thesis around short-term price targets -- but given the conviction shown by these analysts, it's worth taking a closer look at Riskified.

Person analyzing price charts using her digital tablet and paper documents.

Image source: Getty Images.

The bull case

E-commerce fraud is a significant problem, and existing risk management systems tend to be ineffective. In 2021, lost revenue due to false declines totaled $443 billion in the U.S. alone, while lost revenue due to fraudulent charges will exceed $25 billion by 2024. And those numbers are likely to get bigger as e-commerce continues to take market share from brick-and-mortar retail.

Compared to traditional solutions, Riskified integrates more deeply with its merchants' infrastructure. It captures hundreds of data points per transaction, pulling information from websites themselves, as well as back-end systems for payments, shipping, and order management. Riskified then leans on artificial intelligence to correlate those variables with past transactions, identify behavioral patterns, and predict fraud. To that end, Riskified can automatically approve or deny charges in real time with 99.8% accuracy, according to SEC filings.

More importantly, its AI-powered risk management system helps merchants boost sales and reduce fraud-related operating expenses. In fact, a study of the company's 10 largest merchants revealed an average uptick in revenue of 8%, and an average decrease in fraud-related operating expenses of 39%.

Not surprisingly, that value proposition has made its platform very sticky. In 2020, the company kept 98% of its customers, and the average customer spent 17% more. And if you exclude merchants from travel and ticket industries (i.e., markets heavily impacted by the pandemic), Riskified kept 99% of its customers, and the average customer spent 58% more.

The bear case

Generally speaking, Riskified has posted solid top-line growth. In the third quarter, gross merchandise volume (GMV) rose 28% and revenue jumped 26%. The problem starts with gross profit, which rose only 10% during the quarter, as Riskified's gross margin fell seven percentage points to 46%.

That's a problem, because Riskified assumes responsibility for fraudulent charges that slip past its AI, and those expenses -- classified as chargeback guarantee expenses -- are the primary component of the company's cost of revenue. In other words, if gross margin is falling, then fraudulent charges are rising, meaning Riskified's AI models may not be working.

Of course, there are other factors that could affect the performance of those AI models. During the most recent earnings call, management noted that Riskified onboarded several new merchants in new industries during the third quarter, meaning the company lacks sufficient data in those markets. If that's the underlying cause, the problem will resolve itself as Riskified's data becomes more robust. However, some investors are clearly skeptical, as the stock has been in a downward tailspin for several months.

A plausible price target

Looking ahead, shareholders have plenty of reasons to be excited about Riskified. Internal data shows that its platform is more effective than traditional risk management systems, and that value proposition has helped it win several high-profile clients, including three of the 10 largest internet retailers. Moreover, Riskified's AI models should benefit from a network effect, because each new data point should make them a little more effective, thereby bringing more clients (and data) to the platform over time.

However, investors also have reason to be leery. Riskified faces competition from payment giants like PayPal, Mastercard, and Visa, meaning its solution has to be highly effective if it hopes to succeed. In other words, if Riskified's AI models are indeed faulty -- or just moderately effective -- I think that would break the pro-investment thesis.

So could Riskified's price really rise 260% (or more) in the next 12 months? Absolutely. If gross margin reverses course and starts climbing, that would be a very positive sign. And if Riskified continues to grow GMV and revenue quickly, it's entirely plausible that shareholders could see triple-digit returns in the near term.