At January Capital, we have been studying fraud trends and how modern technologies and business models can circumvent existing fraud detection / prevention tools, which would spark a renewed wave of innovation for incumbents and startups. This article is a summary of our emerging perspectives in the modern world of fraud prevention.
‍

If you are an industry practitioner or founder working in this space, we would love to hear from you. You may reach out to Javier Ng at javier@january.capital.
‍

Fraud has become an increasingly pervasive and sophisticated problem in today’s world. Data theft and exfiltration of personally identifiable information (“PII”) has led to identity theft and account takeovers. Advanced social engineering tactics, such as look-alike phishing pages, are used to capture authentication credentials to overcome multifactor authentication (“MFA”). The cost of fraud in the US alone is estimated to be US$8.8 billion in 2022 (+30% y/y), with account takeovers and phishing being perceived as the #2 and #3 cybersecurity risks globally in 2023 (Cyberthreat Defense Report 2023). In APAC, every $1 of fraudulent transaction costs $4 involving fees, investigation, merchandise replacements, and up to $5 — $6 when it comes to alternative financing solutions (e.g. BNPL, digital bank) as fraud prevention measures against newer models are still nascent.

Over the past 12 months, we have spent time understanding this emerging space better. We have engaged with a broad range of fraud prevention practitioners, including heads of fraud prevention at e-commerce organizations, CISOs, as well as security and fraud prevention vendors to better understand trends and challenges faced. In summary, almost all stakeholders in this space operate with a similar objective in mind — to balance the economic risks and costs of fraud while maintaining a frictionless customer experience. If high value transactions are often blocked, customer experience declines and results a loss in revenue. Conversely, if there is no friction in the customer journey, the business is exposed to a huge amount of fraud risk. Manual reviews of transactions and dispute cases are laborious and costly, which makes it impossible to review every transaction or event interaction by customers. This creates an interesting paradox where merchants tend to have to navigate competing priorities (often led by different functional leads) when considering fraud prevention strategies.
‍

Fraud detection and prevention landscape

‍
Fraud detection and prevention systems provide two main areas of ROI to enterprises and merchants:

1. To reduce costs through improved efficiency by reducing manpower costs for manual reviews, and customer acquisition cost in some instances (e.g. ad fraud, returns fraud) as well as related compliance costs;

2. To increase top line by reducing false positives through identifying genuine transactions and providing frictionless customer experiences.
‍

At its core, fraud detection systems work by identifying anomalies in behaviour/actions or those that match the patterns of known fraudulent techniques. This requires adopting either supervised or unsupervised learning techniques or a combination of both. Supervised learning adopts transaction/event scenarios (e.g. chargebacks, escalated transactions, account abuse) as label data and detect fraud based on known patterns. On the other hand, unsupervised learning detects new fraud patterns and identifies new outliers with anomaly detection.

‍

Disclaimer

January Capital is a licensed fund manager in both Singapore (through Jan Cap Pte. Ltd.) and Australia (through January Capital Pty Ltd) (together, the “Managers”) licensed under the purview of the Monetary Authority of Singapore and the Australian Securities and Investments Commission respectively.

The information contained in this article is being provided to you for informational purposes only and does not constitute in any circumstance an offer to sell, or a solicitation of an offer to buy any interest in the funds managed by the Managers or any other companies mentioned herein. Additionally, the information in this article and any and all forward-looking statements are based upon assumptions that may not prove to be correct or accurate and the actual financial results/performance and opportunities (amongst others) may differ materially from these statements. As such, nothing contained in this article should be construed as or relied upon as financial, legal or taxation advice, or an encouragement to invest in our funds or any companies mentioned in this article. Past performance is not indicative of future results.‍

‍