Synthetic identity theft has become the fastest growing type of financial crime in America, according to a new Federal Reserve white paper.

Synthetic identity theft occurs when actors create new identities by combining fabricated and sometimes real personal data, such as birth dates or social security numbers. Perpetrators then obtain credit under the new identity, which they use to run up charges before skipping out on their bills.

According to the white paper, synthetic identities tend to be more prevalent in the United States than in other countries because U.S. identification leans largely on personally identifiable information, which has been vulnerable in recent years due to mass data and cybersecurity breaches.

“Fraudsters increasingly use synthetic identities to commit payments fraud, which can escape detection by today’s identity verification and credit-screening processes,” the paper said. “Because the identity was not real to begin with, there is limited recourse in tracing the perpetrators and holding them responsible.”

Synthetic identity fraud has led to 20 percent of credit losses, $6 billion in costs to U.S. lenders, and an average $15,000 charge-off balance per instance of synthetic identity theft in 2016, the report said.

The theft is also difficult to identify, according to the report, as 85 to 95 percent of applicants identified as potential synthetic identities are not detected by traditional fraud models. Furthermore, synthetic identity fraud is largely unreported because of the types of victims that criminals target.

“It is often unreported, since victims are typically individuals – such as children, the elderly or homeless – who are less likely to access their credit information and uncover the fraud,” the white paper said. “This, combined with gaps in the credit process and potential for large payouts, has made synthetic identity fraud attractive to criminals and crime rings.”

The Federal Reserve said it will work with industry to address synthetic identity theft and to research its detection, controls, and gaps.