The next time some company tells consumers that an attacker only managed to steal Social Security numbers (SSNs) from a database but has no way of tying those back to customer names, the market should still be worried. Because one of the most successful forms of fraud today comes from a type of identity theft that only needs that number.

Called synthetic identity fraud, the method has criminals taking real SSNs and developing fake identities with them. And according to a new report out today, the criminals' success rate with these false identities is growing.

Consumer risk management firm ID Analytics recently examined new account applications in the financial services and wireless industry over a three-year period to estimate the size of the pool of these fraudulent synthetic identities and what kind of risk they pose to companies. According to its report, since 2010 the crooks have managed to bump up the average fraud rate using synthetic identities by more than 100 percent.

"Synthetic identity fraud is a significant and growing problem as fraudsters continue to find new ways to commit crimes despite technological advances," says Dr. Stephen Coggeshall, chief analytics and science officer for ID Analytics and author of the report. "Our latest research in this area shows that, although the number of synthetic identities is decreasing, the riskiness of those synthetic identities is on the rise."

Coggeshall and his colleagues at ID Analytics postulate that a potential reason for this increased risk is that crooks are now exploiting new SSN randomization practices. The Social Security Administration started issuing SSNs in random order rather than an expected pattern back in 2011 to help offer better protection for citizens beleaguered by growing identity theft practices. But in the case of synthetic identity fraud, this policy may actually help attackers, as it makes it much more difficult for today's anti-fraud technology to detect when an SSN issued just a few years ago is tied to a fake identity of someone supposedly decades older.

And Coggeshall and his team believe that synthetic identity fraud will only grow, following the adoption of the EMV chip-and-pin standard for point-of-sale transactions, which "is expected to force fraudsters to adopt new account fraud strategies," according to the report.

As things stand, Gartner estimates that synthetic identity fraud makes up 20 percent of credit charge-offs today and 80 percent of losses from credit card fraud. ID Analytics says the difficulty with synthetic fraud is that because there is no specific consumer victim in this crime, it often goes undetected for long periods of time. Fraudsters develop a fake identity and open up small accounts to establish credit under that persona until they can pounce and get a larger loan they don't intend to pay.

"This long-term 'con' or fraud is particularly dangerous, because criminals employing this technique for financial gain can often nurture the synthetic identity into generating larger credit limits and larger loss amounts for the lender than the average identity theft scenario," writes Coggeshall.

Rather than consumer victims, it is typically financial institutions and the companies offering fraudulently obtained products and services that are victimized by this form of fraud. According to the report, one analysis at a credit card issuer found that, over three years, about 2 percent of the total application volume was made up of synthetic identities. Even the US government is forced to foot the bill for synthetic identity fraud: Today's report shows that ID Analytics found that, over the course of three years, about 1.4 percent of the tax return population appears to be synthetic, with a total refund amount of $20 million going to these accounts.