Take On Payments, a blog sponsored by the Retail Payments Risk Forum of the Federal Reserve Bank of Atlanta, is intended to foster dialogue on emerging risks in retail payment systems and enhance collaborative efforts to improve risk detection and mitigation. We encourage your active participation in Take on Payments and look forward to collaborating with you.
Federal Reserve Web Sites
Other Bank Regulatory Sites
June 19, 2017
Calculating Fraud: Part 2
Part 1 of this two-part series outlined an approach for whittling down credit card transactions to the value or number of authorized and settled payments as the denominator for calculating a fraud rate. This post reviews the elements needed to quantify the numerator.
To summarize from the previous post, when analyzing credit card fraud rates, you should consider what is being measured and compared. To calculate a fraud rate based on value or number, you need a fraud tally in the numerator and a comparison payment tally in the denominator. The formula works out as follows:
Fraud Rate = Numerator
Where, for any given period of time
Numerator = Value, or number of fraudulent payments across the payments under consideration,
Denominator = Value, or number of payments under consideration.
Before calculating the numerator value, you must first decide what types of fraud to include in the measurement. One stratification method divides fraud into the following two categories:
- First-party payments fraud results when a dishonest but seemingly legitimate consumer exploits a merchant or financial institution (FI). That is, the legitimate cardholder authorizes a credit card transaction as part of a scam. One manifestation of this is "friendly fraud," whereby a consumer purchases items online and then falsely claims not to receive the merchandise.
- Third-party payments fraud occurs when a legitimate cardholder does not authorize goods or services purchased with his or her credit card. Besides the victimized cardholder, the other two parties to the transaction are the fraudster and the unsuspecting merchant or FI.
Sometimes no clear delineation between first-party and third-party fraud exists. For example, a valid cardholder may authorize a payment in collusion with a merchant to commit fraud.
The 2016 Federal Reserve Payments Study used only third-party unauthorized transactions that were cleared and settled in tabulating fraud. The study measured and counted fraud as having occurred regardless of whether a subsequent recovery or chargeback occurred. Survey results had to be adjusted because some card networks report gross fraud while others report net fraud, after recoveries and chargebacks. Furthermore, the study made no effort to determine which party, if any, in the payment chain may ultimately bear the loss. Finally, the study did not measure attempted fraud.
Excluding first-party payments fraud
The study excluded first-party fraud due to the greater ambiguity around identifying and measuring it along with the idea that it is difficult to eliminate, given that controls are relatively limited. One control option would be to place repeat offenders on a negative list that, unfortunately, might not be shared with other parties. As a result of excluding first-party fraud, the study focused on fraud specific to the characteristics of the payment instrument being used.
Paraphrasing from page 30 of the 2013 Federal Reserve Payments Study, first-party fraud, while important, is an account-relationship type of fraud and typically would not be included as unauthorized third-party payments fraud because the card or account holder is by definition authorized to make payments. Consequently, first-party fraud can occur no matter how secure the payment method.
As with tallying payments, you could follow a similar process for tallying fraudulent payments for other types of cards payments, with more questionnaire definitions and wording changes needed for other instruments such as ACH and checks.
By Steven Cordray, payments risk expert in the Retail Payments Risk Forum at the Atlanta Fed
June 12, 2017
Watching Your Behavior
Customer authentication has been at the core of the Retail Payments Risk Forum's payments risk education efforts from the beginning. We've stressed not only that there are legal and regulatory requirements for certain parties to "know your customer," but also that it is in the best interest of merchants and issuers to be sure that the party on the other end of a given transaction is who he or she claims to be and is authorized to perform that transaction. After all, if you allow a fraudster in, you have to expect that you or someone else will be defrauded. That said, we also know that performing this authentication, especially remotely, has several challenges.
The recently released 2017 Identity Fraud Study from Javelin Strategy & Research estimated that account takeover (ATO) fraud losses in 2016 amounted to $2.3 billion—a 61 percent increase over 2015's losses. (ATO fraud occurs when an unauthorized individual performs fraudulent transactions through a victim's account.) Additionally, new-account fraud on deposit and credit accounts has increased significantly and generated several public warnings from the FBI.
In payments, the balancing act between imposing additional customer authentication requirements and maintaining a positive, low-friction customer experience has always been a challenge. Retailers, especially online merchants, have been reluctant to add authentication modalities in their checkout process for fear that customers will abandon their shopping carts and move their purchase to another merchant with lower security requirements. Some merchants have recently introduced physical biometrics modalities such as fingerprint or facial recognition for online orders through mobile phones. Although these modalities have gained a high acceptance rate, they still require the consumer to actively participate in the authentication process.
Enter behavioral biometrics for online transactions. Behavioral biometrics develops a pattern of a user's unique, identifiable attributes from when the user is online at a merchant's website or using the merchant's proprietary mobile app. Attributes measured include such elements as typing speed, pressure on the keyboard, use of keyboard shortcuts, mouse movement, phone orientation, and screen navigation. Coupled with device fingerprinting for the customer's desktop, laptop, tablet, or mobile phone, behavioral biometrics gives the merchant and issuer a higher level of confidence in the customer's authenticity. Another benefit is that behavioral biometrics is passive—it is performed without the user's involvement, which eliminates additional friction in the overall customer experience. Proponents claim that while it takes several sessions to develop a strong user profile, they can often spot fraudsters' attempts because fraudsters often exhibit certain recognizable traits.
Behavioral biometrics is still fairly new to the market but over the last couple of years, some major online retailers have adopted it as an additional authentication tool. Like any of the physical biometric modalities, no single behavioral authentication methodology is a silver bullet, and multi-factor authentication is still recommended for moderate- and higher-risk transactions.
By David Lott, a payments risk expert in the Retail Payments Risk Forum at the Atlanta Fed
June 5, 2017
Responsible Innovation Part 1: Can Community Banks Remain Competitive?
The Atlanta Fed's Retail Payments Risk Forum recently co-hosted a summit with the United Kingdom's Department for International Trade to discuss faster payments and their effects on community financial institutions (FIs). In a series of three posts, I will share summaries of the lessons and implications that payments industry stakeholders discussed at the summit. A major theme of these discussions was whether community FIs can remain competitive independent of how they access a faster payments network. This post tackles this theme.
|United States||United Kingdom|
|ACH (NACHA)||ACH (Bacs)|
|Real-Time Payments (The Clearing House)||Faster Payments (Faster Payments Scheme Ltd.)|
The Faster Payments Scheme, or FPS, opened in the United Kingdom in 2008. The summit was a good opportunity to hear first-hand from one community banker's experience with the still-new system. A panelist from the first retail community bank to join the FPS discussed how access options played a role in the bank's ability to compete with large FIs.
- In the beginning, the only way a community bank could access the FPS was through a sponsoring bank.
- This option was expensive, hindering, and much like a newborn baby who needed attention all day and night (even on weekends), according to the panelist.
- The FPS sends messages 24/7, in near-real time, but her bank's access model often caused a delay of 15 to 30 minutes, making the bank less than competitive.
- Last year, the bank was able to join as a "Direct Participant" under the New Access Model,, an experience that the panelist compared to parenting a toddler who allows her to sleep through the night, even as it runs 24/7/365. The new model was also much more affordable and provided her community bank the near-real time model larger banks received. (The New Access Model that gives payment service providers and community FIs direct connection began in 2014, six years after the FPS began.)
- The panelist did note a serious obstacle to this access model for the smaller banks: the onerous 12-month certification process to become a Direct Participant is tailored to large banks. The process required significant resources and strained other areas of her bank. She suggested that the certification take a risk-based approach.
Two developments on the way may affect future access options: (1) plans are set to consolidate Bacs, FPS, and Cheque; and (2) the Bank of England plans to grant settlement services to nonbank payment service providers.
The United States is facing a similar challenge: community FIs will have to choose how to access faster payment systems. Some community FIs have begun to offer same-day ACH and will likely consider real-time payments later this year.
Representatives from the Clearing House's Real-Time Payments initiative shared some details on their access model:
- FIs of all sizes will be able to connect directly or through third-party service providers.
- Regional payments associations will play an important role as they collectively represent all U.S. financial institutions plus third-party processors.
- The speed will be the same for all participants.
- Indirect participation will not be available.
- Payments can be made 24/7/365.
While direct access is available for both same-day ACH and Real-Time Payments, some FIs may choose to use a sponsor or correspondent access model. To remain competitive, community FIs will have to understand the advantages and limitations that each access model provides.
The next installment in this series will discuss the U.S. market appetite for faster payments; the one after that will look at the impacts of adoption.
By Jessica Washington, AAP, payments risk expert in the Retail Payments Risk Forum at the Atlanta Fed
May 22, 2017
The Year(s) of Ransomware
I remember, as a child, despising the neighborhood kid who would always say, "I told you so." Well, let's move ahead some 30-odd years to the WannaCry ransomware attack—I now feel like that despised child. You see, on March 29 of this year, I emailed the following note to my colleagues in the Risk Forum:
Just a few high-level and interesting notes from the conference.… 2017 & 2018 will be the Year of Ransomware (I can elaborate on this when we are all together—pretty fascinating business models developed here).
Too bad I kept my thoughts to our little group here at the Atlanta Fed and didn't get the message out to the masses (or at least to our Take on Payments readers) prior to the WannaCry ransomware attack that began on May 12. So why did I (and still do) think 2017 and 2018 will both be the "Year of Ransomware"?
Those who know me know that I am not a very technical person. I see things more strategically than technically and usually sprint away from conversations that become technical. After viewing a demonstration on how to launch a ransomware attack, I was shocked to learn that hardly any technical expertise is required to pull off an attack. This is all made possible by the "pretty fascinating business models" that I referred to in my note, business models known as Ransomware as a Service (RaaS).
I'd always envisioned that serious technical code writing capabilities would be a requirement for developing the code to send the malicious files involved in ransomware. And while coding is needed, that is where the RaaS comes into play. You pay someone else to create the malicious code, which you then use to launch a ransomware attack. And to make the attack even more successful, there are simple tools available that allow you to not only test the code against the market-leading antivirus software detection programs but also to tweak the code embedded in the malicious file to ensure that none of the antivirus software programs will detect it. Antivirus software protects users only from known malicious code, which is the reason the software must be constantly updated.
With the undetectable code in hand, you can now launch a ransomware attack through either an embedded file or a link within a phishing email or social media post to a legitimate-appearing, but malicious, website. And this costs little or nothing up front! The cost for the RaaS is only realized once a successful attack occurs, with a portion of the collected ransom paid to the RaaS provider.
Which brings me back to why I think ransomware attacks will continue to escalate, leading to 2017 and 2018 becoming "The Year(s) of Ransomware." They are simple to execute, low cost, and proving to be highly lucrative. (According to the FBI, an estimated $209 million was paid in ransom in the first quarter of 2016.) Expect a future blog post on how to plan for and defend against attacks.
By Douglas A. King, payments risk expert in the Retail Payments Risk Forum at the Atlanta Fed
- Calculating Fraud: Part 2
- Watching Your Behavior
- Responsible Innovation Part 1: Can Community Banks Remain Competitive?
- The Year(s) of Ransomware
- What Canada Knows That We Don't
- Calculating Fraud: Part 1
- Additional Authentication: Is the Protection Worth the Hassle?
- Would Consumers Ever Give Up Their Passwords?
- Will the Password Ever Die? Part 1
- Catch Me If You Can
- June 2017
- May 2017
- April 2017
- March 2017
- February 2017
- January 2017
- December 2016
- November 2016
- October 2016
- September 2016
- account takeovers
- ATM fraud
- bank supervision
- banks and banking
- card networks
- check fraud
- consumer fraud
- consumer protection
- cross-border wires
- data security
- debit cards
- emerging payments
- financial services
- identity theft
- law enforcement
- mobile banking
- mobile money transfer
- mobile network operator (MNO)
- mobile payments
- money laundering
- money services business (MSB)
- online banking fraud
- payments risk
- payments study
- payments systems
- phone fraud
- remotely created checks
- risk management
- Section 1073
- social networks
- third-party service provider
- trusted service manager
- Unfair and Deceptive Acts and Practices (UDAP)
- wire transfer fraud
- workplace fraud