Board of Governors of the Federal Reserve System

2023 Interchange Fee Revenue, Covered Issuer Costs, and Covered Issuer and Merchant Fraud Losses Related to Debit Card Transactions

Preface

The Dodd-Frank Act requires the Board to prescribe rules related to debit card interchange fees and routing. The Board's Regulation II implements this provision of the Dodd-Frank Act.¹

The Dodd-Frank Act also requires the Board to disclose aggregate or summary information concerning the costs incurred, and interchange fees charged or received, by issuers and payment card networks as the Board considers appropriate and in the public interest on at least a biennial basis. Pursuant to these statutory instructions, every two years beginning in 2011, the Board has released a report and accompanying data based on data collections from debit card issuers and payment card networks.

Highlights

Card Use

• In 2023, payment card networks in the United States processed 100.7 billion debit and general-use prepaid card transactions valued at $4.7 trillion. Total debit card transaction volume and value both grew at an average rate of 4.6 percent per year from 2021 to 2023, significantly slower than the average annual growth rates from 2009 to 2021 (7.8 percent by volume and 9.5 percent by value).
• Dual-message networks, which traditionally processed mainly signature-authenticated transactions, accounted for 71.4 and 72.9 percent of the total by volume and value, respectively, in 2023. Single-message networks, which traditionally processed PIN-authenticated transactions, accounted for the rest. Since 2009, the dual-message share of total transactions has progressively increased 9.0 percentage points.
• From 2021 to 2023, annual growth for dual-message transactions (7.0 percent by volume and 6.4 by value) was greater than for single-message transactions, as it had been from 2009 to 2021. In fact, from 2021 to 2023, single-message transaction value barely increased, while single-message transaction volume actually declined over the two years, driven by the first-ever annual decline in volume of 2.2 percent from 2021 to 2022.
• Card-not-present (CNP) transactions reached 34.4 percent of total volume and almost one-half of total value in 2023, having grown consistently faster than card-present (CP) transactions since 2009. From 2021 to 2023, CNP transaction volume grew 8.2 percent per year, faster than the 2.8 percent per year growth recorded by CP transactions, but considerably slower than the 26.5 percent per year CNP growth observed from 2019 to 2021 during the COVID-19 pandemic, or the 17.8 percent per year growth observed from 2009 to 2019.
• In 2023, the average transaction value of CNP transactions was once again higher than that of CP transactions, at $63.96 and $36.99, respectively. However, as e-commerce has grown in popularity over the years and expanded to include smaller value purchases, the average transaction value of CNP transactions has decreased. In 2023, the average transaction value of CNP transactions was 1.7 times the average transaction value of CP transactions, down from 2.3 in 2009.
• From 2021 to 2023, annual volume growth for transactions subject to Regulation II's interchange fee standards ("covered transactions") was slightly higher than for transactions not subject to the interchange fee standards ("exempt transactions"), at 4.8 and 4.2 percent, respectively.
• From 2021 to 2023, annual growth for prepaid card transactions (11.5 percent by volume and 7.0 by value) was once again faster than for non-prepaid card transactions (2.7 percent by volume and 3.3 percent by value). In fact, prepaid card transactions surpassed CNP transactions as the fastest growing category for the first time since 2015, having grown 19.7 percent from 2022 to 2023.

Interchange Fees, Network Fees, and Incentives

• In 2023, interchange fees across all debit and general-use prepaid card transactions totaled $34.12 billion, an average increase of 3.9 percent per year since 2021.
• For transactions covered by Regulation II, the average level of interchange fees has not changed materially since the regulation took effect in the fourth quarter of 2011. In 2023, the average interchange fee for covered transactions processed over single-message networks was $0.24, while that for covered transactions processed over dual-message networks was $0.22.
• The average interchange fee for transactions exempt from Regulation II was $0.52 in 2023. Across data collections, the average interchange fee for exempt transactions gradually increased from $0.43 since the regulation took effect to $0.52 in 2021, and remained steady through 2023. For exempt transactions processed over dual-message networks, the average interchange fee fell from $0.64 in 2021 to $0.62 in 2023, marking a reversal of the upward trend that saw this value gradually increase from $0.51 in the fourth quarter of 2011 to $0.64 in 2021. By contrast, for exempt transactions processed over single-message networks, the average interchange fee gradually fell after Regulation II took effect, from $0.31 in the fourth quarter of 2011 to $0.25 in 2017, and has since gradually increased to $0.27 in 2023.
• Network fees to all parties increased to $12.95 billion in 2023 from $11.49 billion in 2021. The average network fee per transaction increased to $0.129 in 2023, up from $0.125 in 2021 and $0.098 in 2011. By contrast, the average network fee per transaction expressed as a percentage of average transaction value, 0.28 percent in 2023, has not changed materially since Regulation II took effect.
• In 2023, acquirers and merchants paid 64.9 percent of network fees, up from 63.9 percent in 2021; issuers paid the rest. This increase is a continuation of a long-term trend, which has seen the proportion of network fees paid by acquirers and merchants increase gradually from 44.3 percent in 2009.

Fraud Losses

• In 2023, across all transactions for covered issuers, fraud losses to all parties as a share of the transaction value were 17.6 basis points, or $17.63 per $10,000 in transaction value. This value reflects a steady increase in fraud losses from 7.8 basis points in 2011.
• In 2023, merchants absorbed 49.9 percent of losses from fraudulent transactions reported by covered issuers, up from 46.9 percent in 2021 and 38.3 percent in 2011. By contrast, issuers absorbed 28.3 percent of losses in 2023, down from 33.4 percent in 2021 and 59.8 percent in 2011. Cardholders, who absorb the remaining fraud losses, saw their share of fraud losses rise from less than 1.8 percent in 2011 to 21.8 percent in 2023.
• In 2023, across all transaction categories, prepaid transactions experienced the highest level of fraud losses as a share of transaction value, as well as the highest incidence of fraudulent transactions as a percentage of total transactions (312.6 basis points and 1.12 percent, respectively). Fraud levels were second highest for dual-message transactions (17.9 basis points and 0.12 percent, respectively) and were once again the lowest for single-message transactions (6.7 basis points and 0.03 percent, respectively), as they have been since 2011.

Issuer Costs

• Across all covered issuers, average per-transaction authorization, clearing, and settlement (ACS) costs, excluding issuer fraud losses, were $0.041 in 2023, approximately one-half of the 2009 value, and a slight increase from $0.039 in 2021.
• In 2023, the base interchange fee cap in Regulation II excluding the fraud-prevention adjustment—$0.21 plus 5 basis points multiplied by the value of a transaction—exceeded the average per-transaction ACS costs, including issuer fraud losses, for 80.1 percent of covered issuers and for 99.2 percent of covered transactions.

Data Collections

Payment Card Network Survey

The Board collected data for 2022 and 2023 through the two most recent Payment Card Network (PCN) Surveys, conducted in 2023 and 2024, respectively. All 13 card networks that processed debit card transactions in 2022 and 2023 completed the survey for each of those years.² In addition to using data from these latest PCN Surveys for this report, the Board used the data to calculate the annual information it published on the average interchange fees received by issuers across different networks.³

As in previous years, the most recent PCN Surveys asked respondents to report information separately for issuers covered by the interchange fee standards (covered issuers) and issuers exempt from the interchange fee standards (exempt issuers). The surveys further asked respondents to distinguish between prepaid card transactions that were covered by or exempt from the interchange fee standards. Starting with the 2013 data collection, the PCN Surveys asked respondents to distinguish exempt prepaid transactions performed with cards issued by exempt issuers from those performed with cards issued by covered issuers. This distinction allows transactions reported in the PCN Surveys to be categorized as either exempt or covered.⁴

As figure 1 illustrates, the difference between the transaction volume of covered issuers, which includes transactions performed with exempt prepaid cards issued by covered issuers, and the volume of covered transactions, which excludes those transactions, was equal to 3.0 percent of total transaction volume of covered issuers or 1.9 percent of total transaction volume of all issuers in 2023, in line with previous years. Depending on the context, this report decomposes the data either by transaction status under Regulation II (that is, exempt versus covered transactions) or by issuer status (that is, exempt versus covered issuers).

Debit Card Issuer Survey

The Board collected data from covered issuers for 2023 through the latest Debit Card Issuer (DCI) Survey, conducted in 2024. As in previous data collections, respondents ranged from the largest debit card issuers in the United States to issuers with small debit card programs and consolidated assets greater than $10 billion.⁵ The DCI Survey asked for information related to the volume and value of debit and prepaid card transactions; authorization, clearing, and settlement costs as well as other costs associated with particular debit card transactions, including the incidence of and losses related to debit and prepaid card fraud; fraud-prevention and data-security costs; and interchange fee revenue. The DCI Survey instructed respondents to provide the requested information separately for debit card transactions processed over dual-message networks, debit card transactions processed over single-message networks, and prepaid card transactions for issuers with prepaid card programs.⁶

A total of 166 covered issuers responded to the 2023 DCI Survey, compared with 162 respondents for 2021.⁷ Figure 2 shows the percentage that high-, mid-, and low-volume issuers represented out of total covered issuers and the total number and value of covered issuers' transactions in 2023. In particular, 34.1 percent of the respondents processed more than 100 million debit card transactions (high-volume issuers), 53.9 percent processed between 1 million and 100 million transactions (mid-volume issuers), and the remaining 12.0 percent of respondents processed fewer than 1 million debit card transactions (low-volume issuers). As in previous data collections, high-volume issuers accounted for the vast majority of transaction volume and value. In 2023, high-volume issuers accounted for 94.0 percent of transaction volume and 93.9 percent of transaction value, mid-volume issuers accounted for 5.9 percent of transaction volume and 6.1 percent of transaction value, and low-volume issuers accounted for 0.01 percent of transaction volume and value.

Detailed Discussion

Card Use

The PCN Surveys have provided information about total card usage on a yearly basis since 2009.⁸ Over this period, the volume of total card usage progressively grew from 37.6 billion transactions in 2009 to 100.7 billion transactions in 2023, corresponding to an average growth rate of 7.3 percent per year (see figure 3). Total debit card transaction volume and value both grew at an average rate of 4.6 percent per year from 2021 to 2023, significantly slower than the average annual growth rates from 2009 to 2021 (7.8 percent by volume and 9.5 percent by value). The total value of purchase transactions also increased, growing consistently from $1.43 trillion in 2009 to $4.66 trillion in 2023, corresponding to an average growth rate of 8.8 percent per year.⁹

Figure 4 examines transaction volume growth rates across different transaction categories from 2022 to 2023 compared with average annual growth rates from the first year the data were collected to 2022. The growth rate for transactions processed over dual-message networks from 2022 to 2023, at 8.4 percent, was in line with the average annual growth rate of 8.3 percent per year observed from 2009 to 2022. By comparison, the growth rate from 2022 to 2023 for transactions processed over single-message networks was 0.7 percent, once again lower than the corresponding value for dual-message networks, and significantly lower than the average annual growth rate of 5.6 percent per year from 2009 to 2022 for single-message transactions.¹⁰ Moreover, single-message transaction volume decreased 2.2 percent from 2021 to 2022, the first time when an absolute decrease in the number of single-message transactions was observed. This decline led to a decrease in the average annual growth rate since 2009, lowering it from an average of 6.3 percent per year from 2009 to 2021, to 5.6 percent per year from 2009 to 2022.

As shown in the second set of bars in figure 4, growth in the volume of CNP transactions slowed significantly from an average annual growth rate of 18.1 percent per year from 2009 to 2022 to 10.7 percent from 2022 to 2023. From 2022 to 2023, CNP transactions was the second-fastest growing transaction category—the first time since 2015 that CNP transaction volume did not grow faster on an annual basis than all other transaction categories.¹¹ CNP transaction volume grew at an average annual rate of 19.2 percent per year from 2009 to 2021, with a substantially higher growth rate experienced from 2019 to 2020, at 33.8 percent. Since 2021, the CNP growth rate slowed, dropping to 5.8 percent from 2021 to 2022 and reverting slightly upward, to 10.7 percent, from 2022 to 2023. This slower growth following 2021 reverted the average annual growth rate of CNP transaction volume to its pre-COVID rate. From 2009 to 2023, the average annual growth of CNP transaction volume was 17.5 percent, which is essentially identical to the average annual growth rate observed from 2009 to 2019 (17.8 percent). By contrast, card-present (CP) transaction volume increased at a rate of 4.9 percent per year from 2009 to 2022 and at a slower rate of 3.8 percent from 2022 to 2023.

As shown by the third set of bars in figure 4, the growth rate in the volume of transactions processed by covered issuers increased slightly, from an average annual growth rate of 6.0 percent per year from 2011 to 2022 to 7.1 percent from 2022 to 2023. Growth in the volume of transactions processed by exempt issuers was slightly lower than for their covered counterparts, slowing down from an average annual growth rate of 9.6 percent from 2011 to 2022 to 9.1 percent from 2022 to 2023.

Finally, as shown by the final set of bars in figure 4, prepaid card transactions recorded the highest growth rate by volume from 2022 to 2023 among all transaction categories. Growth in prepaid card volume accelerated from an average annual growth rate of 15.4 percent per year from 2009 to 2022 to 19.7 percent from 2022 to 2023. By contrast, growth in non-prepaid transaction volume slowed from an average annual growth rate of 6.7 percent per year from 2009 to 2022 to 4.9 percent from 2022 to 2023.

While figure 4 examines the growth rates in transaction volume across different transaction categories, figure 5 focuses on the composition of total transactions within each of these categories in 2023. To better understand how the composition itself evolved, figure 5 also illustrates the corresponding breakdown at the beginning of the data collection, and the associated change in composition relative to that initial breakdown.

In 2023, transactions processed over dual-message networks accounted for 71.4 percent of debit card transaction volume, with transactions processed over single-message networks accounting for the remaining 28.6 percent. The proportion of transactions processed over dual-message networks increased 9.0 percentage points from 2009 to 2023. The share of CNP transactions out of total transaction volume reached 34.4 percent in 2023, up from 9.6 percent in 2009. This increase reflects the continuing increase in popularity of online retail, as well as the significant increase in CNP transaction volume of 33.8 percentage points from 2019 to 2020 alone. The share of transactions processed by exempt issuers reached 38.8 percent in 2023, an increase of 6.0 percentage points from the corresponding share in 2011. Finally, the share of prepaid transactions has been growing steadily and reached 9.2 percent in 2023, almost three times higher than the share in 2009.

As shown in figure 6, average transaction values in 2023 did not differ substantially across most transaction categories, which has been the case across data collections. In 2023, average values for most transaction categories were higher than they were in the first year the data were collected. Unlike the steady patterns observed in other data series, this increase was primarily the result of larger increases observed from 2019 to 2021.

CNP transactions were a notable exception in both cases. In 2023, the average value of CNP transactions ($63.96) was nearly three-quarters greater than the corresponding value for CP transactions ($36.99). While the average value of CNP transactions was still considerably higher than the average value for all other transaction categories in 2023, it fell by over $14 from 2009 to 2023. However, after a steady decline from 2009 to 2019, reaching the lowest observed average value of $61.56 in 2019, the average value of CNP transactions has progressively increased since then, reversing the declining trend.

Interchange Fees, Network Fees, and Incentives

The PCN Survey requests information about interchange fees; certain network fees; and payments and incentives (P&I) that are set, charged, or paid by payment card networks.¹² These fees and transfers vary from network to network; thus, the totals and averages reported in this section serve only as a general characterization of network practices. Moreover, P&I are usually bilateral arrangements between a network on one side and a merchant, acquirer, or issuer on the other. The statistics reported in this section, calculated from network-reported totals, do not reflect the heterogeneity of these bilateral arrangements.

Figure 7 shows the main fees, as well as P&I, exchanged among networks, issuers, and acquirers/merchants in 2023. The total value of interchange fees transferred from acquirers to issuers was by far the largest: $34.12 billion. Network fees paid by issuers and acquirers/merchants were $4.54 billion and $8.40 billion, respectively. In each case, these network fees were considerably higher than the P&I issuers and acquirers/merchants received from networks: $1.98 billion and $3.71 billion, respectively.

Figure 8 illustrates the evolution of average interchange fees over time, by network type and transaction status (whether transactions were covered by or exempt from the interchange fee standards). Average interchange fees for covered transactions, both those processed over single-message and those processed over dual-message networks, did not change materially after Regulation II took effect in the fourth quarter of 2011. In 2023, these fees stood at $0.24 and $0.22, respectively. In both cases, the fees were slightly less than the maximum that an issuer could receive on an average covered transaction under the regulation.¹³ While average interchange fees on covered transactions were higher for dual-message networks immediately after the regulation took effect, their subsequent slow-but-steady decline means that, since 2013, average interchange fees for covered transactions have been higher for transactions processed over single-message networks.

Average interchange fees for exempt transactions also declined in the years after Regulation II took effect, albeit at a much slower rate than interchange fees for covered transactions—by 2016, fees for exempt transactions processed over dual-message networks were $0.51, down from $0.53 before the regulation took effect, while fees for exempt transactions processed over single-message networks were $0.26, down from $0.32. After 2016, the paths of the two average fees diverged. In particular, while average interchange fees for exempt transactions processed over single-message networks have changed little since 2016 (and equaled $0.27 in 2023), average interchange fees for exempt transactions processed over dual-message networks rose from $0.51 in 2016 to $0.64 in 2021, before declining slightly to $0.62 in 2023.

Figure 9 illustrates, for 2023, network fees paid by acquirers and issuers as well as P&I received by acquirers/merchants and issuers from networks. Overall, for both single-message and dual-message networks, acquirers/merchants and issuers paid more in network fees than they received from the networks in P&I. Across network types, dual-message networks consistently charged higher network fees and disbursed higher P&I than single-message networks.

Looking across different parties and types of networks, per-transaction network fees paid by acquirers were considerably higher than those paid by issuers for both dual-message and single-message networks. Network fees charged by dual-message networks were considerably higher than the fees charged by single-message networks for both acquirers and issuers.

For single-message networks, the amount of P&I that acquirers/merchants received corresponded to around 39.7 percent of the network fees that they paid, over twice as much as the corresponding value for issuers, 18.9 percent. For dual-message networks, by contrast, the amount of P&I that acquirers/merchants and issuers received was similar and higher than for single-message networks, at 45.0 percent and 47.6 percent, respectively. Notably, the amount that acquirers/merchants received in P&I relative to the network fees paid increased substantially from 2021 to 2023 for both network types, reaching its highest value since the promulgation of Regulation II.

Figure 10 shows trends over time in network fees, as well as in P&I. Focusing on network fees first, single-message network fees have changed little since 2009, rising a bit for acquirers/merchants and falling a bit for issuers. By contrast, dual-message network fees have changed more over time, almost doubling from 2009 to 2023 for acquirers/merchants, and substantially decreasing for issuers. Turning to P&I, payments and incentives received by issuers from either network type has not changed materially since 2009. By contrast, P&I received by acquirers/merchants have increased substantially since 2009, especially payments and incentives from dual-message networks.

Figure 11 examines how network fees and P&I differed between covered and exempt issuers in 2023. Exempt issuers paid higher per-transaction network fees and received higher per-transaction P&I than covered issuers, with these differences occurring for both dual-message and single-message networks.¹⁴ Moreover, P&I paid to exempt issuers by dual-message networks remained substantially higher than those paid by single-message networks in 2023. P&I paid to covered issuers by dual-message networks were more than 12 times higher than those paid by single-message networks in 2023. While covered issuers received P&I equal to 74.7 and 13.6 percent of the network fees that they paid to dual-message and single-message networks, respectively, P&I for exempt issuers were 40.0 percent of network fees paid to dual-message and 20.4 percent of fees paid to single-message networks.

Figure 12 examines how network fees, as well as P&I, for covered and exempt issuers changed after 2011. As previously noted, network fees paid by issuers to single-message networks as well as P&I received by issuers from both types of networks have remained relatively stable over time. By contrast, network fees paid by issuers to dual-message networks gradually declined since the beginning of the data collection. That overall decline was driven by a substantial decline in dual-message network fees paid by covered issuers, which outweighed the gradual increase in dual-message network fees paid by exempt issuers over the same period.

Fraud

Fraud is ever-evolving and shifts toward new areas of vulnerability as new methods to prevent fraud are developed and deployed. For example, with the introduction of increased security for in-person card payments through the issuance of chip-based EMV cards, card fraud shifted from in-person fraud toward CNP, or remote, fraud.¹⁵

The fraud data presented in this section are from the DCI Survey and, therefore, only apply to covered issuers.¹⁶ Because these issuers constitute a particular segment of the total population of debit card issuers, their fraud experience may not be representative of all debit card issuers. As a result, estimates of aggregate fraud that might be derived from these data could differ from those based on information that reflects a broader set of issuers.

Figure 13 shows that overall fraud losses as a share of transaction value rose from 17.5 basis points in 2021 to 17.6 basis points in 2023, a substantially smaller increase than in previous years. However, this small increase masks substantial differences across different transaction categories and is driven primarily by a small increase in fraud losses for dual-message transactions, the most common transaction type. By contrast, fraud losses as a share of transaction value increased substantially for single-message transactions, from 4.2 basis points in 2021 to 6.7 basis points in 2023, while those for prepaid transactions rose from 222.1 basis points in 2021 to 312.6 basis points in 2023.

Figure 14 shows total fraud losses in 2023 to all parties expressed as a share of transaction value, broken down into four types of fraud: (1) lost and stolen fraud, (2) counterfeit fraud, (3) CNP fraud, and (4) other fraud.¹⁷ As figure 14 illustrates, CNP fraud accounted for more than one-half of overall fraud in 2023, at 54.3 percent. The second-largest fraud source, lost and stolen fraud, resulted in less than one-half as much fraud, at 22.2 percent. The third-largest fraud source, counterfeit fraud, accounted for a similar share of fraud, at 17.5 percent. Other fraud accounted for the remaining 5.9 percent.

The main fraud type driving observed fraud losses depends on the transaction category. CNP fraud was the main driver for dual-message transactions, making up over two-thirds of total fraud losses. Lost and stolen fraud made up the majority of prepaid fraud losses, at 72.9 percent. By contrast, the fraud experienced by single-message transactions can be accounted for by CNP, lost and stolen, and counterfeit fraud in almost equal shares. Additionally, other fraud constituted a small share of fraud losses for each transaction category; however, in absolute terms, the value for other fraud experienced by prepaid transactions was more than an order of magnitude higher than for the other two transaction categories.

On July 1, 2023, the final revision to Regulation II specifying that the regulation's routing provisions apply to CNP transactions went into effect.¹⁸ While single-message networks were still used relatively rarely for CNP transactions in 2023, a modest but clear increase in usage was observed. In particular, the share of volume and value of single-message CNP transactions out of all CNP transactions increased from 6.1 and 4.5 percent, respectively, in 2021, to 6.6 and 5.0 percent in 2023. With increased routing of CNP transactions over single-message networks, the share of total single-message fraud attributable to CNP transactions in 2023 increased to 29.7 percent but remains less than one-half of the equivalent share for dual-message networks (69.4 percent).

Figures 13 and 14 focus on fraud losses as a share of transaction value and the fraud loss per fraudulent transaction (figure 16). The product of these two factors yields the average fraud loss per transaction.¹⁹

Figure 15 shows that overall fraud incidence displayed a similar directional trend from 2009 to 2023 to that of fraud losses as a share of transaction value (figure 13) but with a greater magnitude. Overall fraud incidence flattened from 2021 to 2023, putting an end to the upward trend observed from 2011 to 2021. Fraud incidence more than tripled from 2011 to 2021, an even sharper increase than the one observed in fraud losses as a share of transaction value over the same period (figure 13). As had been the case since 2009, prepaid and dual-message transactions exhibited a considerably higher fraud incidence than single-message transactions in 2023.

Figure 16 shows that the average loss per fraudulent transaction across all transactions remained relatively stable from 2021 to 2023. This was driven by two opposing forces: (1) an increase in the average loss per fraudulent transaction for both single- and dual-message transactions and (2) a decrease in the average loss per fraudulent transactions for prepaid transactions. Unlike fraud losses as a share of transaction value (figure 13) and fraud incidence (figure 15), the average loss per fraudulent prepaid transaction is not an outlier relative to the other transaction categories. In particular, the average loss per fraudulent prepaid transaction decreased slightly from 2021 to 2023 and remained above the equivalent value for dual-message transactions but below that for single-message transaction. As had been the case since 2009, in 2023 the average loss per fraudulent dual-message transaction was fairly similar to the average loss across all transactions. Finally, the average loss for single-message transactions increased slightly from $122.89 in 2021 to $128.42 in 2023, remaining nearly double the corresponding value for dual-message transactions and higher than the corresponding value for prepaid transactions.

Figure 17 breaks down fraud losses that were absorbed by merchants, cardholders, and issuers in 2011 and 2023.²⁰ In both years, the majority of fraud losses for all transactions were absorbed by issuers and merchants. However, the composition of fraud losses for all transactions, as well as for particular transaction categories and fraud types, evolved significantly over this period.

From 2011 to 2023, for all transactions, the share of fraud losses absorbed by issuers fell from 59.8 percent in 2011 to 28.3 percent in 2023, while the equivalent shares rose for merchants (from 38.3 percent in 2011 to 49.9 percent in 2023) and for cardholders in (from less than 1.8 percent in 2011 to 21.8 percent in 2023).

These changes reflect both shifts in the composition of fraud losses by fraud type and transaction category and varying growth rates across different transaction categories. In particular, the rise in cardholders' share of fraud losses is driven mostly by the former, with shares rising across all fraud types and fraud categories, and especially for prepaid transactions and counterfeit fraud. By contrast, the rise in merchants' share of fraud losses is driven in large part by the rise in popularity of CNP transactions (as illustrated in figure 5) for which merchants' share of fraud has been much higher than for other categories, despite that share falling slightly from 2011 to 2023.

Focusing on the composition of fraud losses in 2023, the share of losses absorbed by merchants, cardholders, and issuers varied significantly across transaction categories and fraud types. Merchants absorbed the majority of fraud losses for transactions dual-message transactions (64.4 percent) and CNP fraud (65.1 percent). Issuers absorbed the largest share of fraud losses for single-message transactions (45.3 percent) and lost and stolen fraud (49.3 percent). Cardholders absorbed the largest share of fraud losses for prepaid transactions (56.3 percent) and counterfeit fraud (30.7 percent).

Issuer Costs

Like the fraud data presented in the previous section, issuer cost data presented in this section come from the DCI Survey and, therefore, apply only to covered issuers. Figure 18 presents average per-transaction authorization, clearing, and settlement (ACS) costs over time for different transaction categories.²¹ After declining consistently from 2009 to 2017, average ACS costs for all transactions stabilized in recent data collections, reaching $0.041 in 2023, a slight increase from 2021 ($0.039). Looking across transaction categories, in 2023, the average ACS cost of a prepaid card transaction ($0.129) was more than three times the cost of a dual-message debit transaction ($0.039) and more than four times the cost of a single-message debit transaction ($0.030). This pattern, in which prepaid card transactions are the most costly and single-message transactions are the least costly, has not changed since 2009. The average ACS costs for all three transaction categories converged from 2009 to 2019, although the average ACS cost of a prepaid card transaction has since diverged, rising from $0.076 in 2019 to $0.129 in 2023. Despite this increase, from 2009 to 2023 the highest-cost transaction category (prepaid) has exhibited the largest overall decline (50.2 percent), while the lowest-cost transaction category (single-message) saw the smallest decline (35.7 percent).

Figure 19 presents the average ACS costs over time by issuer size. Overall, for all years, issuers who processed more transactions consistently had lower per-transaction ACS costs on average.²² In 2023, the average per-transaction ACS cost for mid-volume issuers ($0.121) was over three times higher than the cost for high-volume issuers ($0.036), whereas the cost for low-volume issuers ($1.088) was 30 times higher than the cost for high-volume issuers. From 2021 to 2023, the average per-transaction ACS cost increased for issuers of all sizes.

The overall decline in the average per-transaction ACS cost over the period from 2011 to 2023 has been driven primarily by the decline in the average pre-transaction ACS cost among high-volume issuers, which fell from $0.047 in 2011 to $0.036 in 2023. Over the same period, the average per-transaction ACS cost for mid-volume issuers increased slightly above its 2011 level. Lastly, the average per-transaction ACS cost for low-volume issuers oscillated around $0.60 from 2011 to 2021, but its value in 2023 ($1.09) was almost double the corresponding 2021 value.

Figure 20 breaks down ACS costs in 2023 across three categories: (1) in-house costs, (2) third-party processing fees, and (3) network fees.²³ Overall, in-house costs constituted almost one-half of total ACS costs, at 45.2 percent, third-party processing fees constituted 29.6 percent, and network fees made up the rest. This split varied substantially across transaction categories and issuer size. For example, in-house costs constituted 53.9 percent of total ACS costs for dual-message transactions but only 32.9 percent for prepaid card transactions, while third-party processing fees constituted 60.8 percent of costs for prepaid cards, almost three times the share for dual-message transactions. Similarly, the share of in-house and third-party processing costs for high-volume issuers was 52.5 and 27.3 percent, respectively, while the corresponding shares for low-volume issuers were 5.6 and 75.6 percent.

Figure 21 illustrates how in 2023 the average per-transaction ACS costs borne by issuers compared with other costs borne by issuers in connection with their debit card programs. Overall, per-transaction ACS costs excluding issuer fraud losses constituted around one-third of average costs borne by issuers as reported in this data collection. Across transaction and issuer categories, three types of costs varied more than others: ACS costs excluding issuer fraud losses, cardholder inquiry costs, and issuer fraud losses.²⁴ Unlike the pattern with ACS costs, there is no clear negative relationship between fraud-prevention costs and issuer volume, but high-volume issuers tend to incur a higher share of fraud-prevention costs. In particular, while mid-volume issuers had fraud-prevention costs that were 9.7 percent of total costs, those costs were more than twice as high, at 20.6 percent, for high-volume issuers, and slightly higher, at 10.3 percent, for low-volume issuers.

Issuer fraud losses as a share of total costs were higher for prepaid transactions and lower for low-volume issuers relative to other transaction and issuer categories. Cardholder inquiry costs were comparable across transaction categories; across issuer categories, cardholder inquiry costs were 31.3 percent of total costs for high-volume issuers but constituted a much smaller share of total costs for mid- and low-volume issuers: 6.3 percent and 0.6 percent, respectively. Reward program costs constituted just 0.2 percent of overall debit card costs and were equal to or below 1 percent for all transaction categories and issuer volume tiers.²⁵

Figure 22 illustrates how average per-transaction ACS costs and other costs borne by issuers changed over time. Overall, total issuer per-transaction costs gradually decreased from $0.169 in 2011 to $0.127 in 2023.²⁶ After Regulation II took effect, reward program costs substantially declined, with the largest drop observed from 2011 to 2013. ACS costs excluding issuer fraud losses, cardholder inquiry costs, and nonsufficient funds handling costs all progressively fell from 2011 to 2023. By contrast, fraud-prevention costs and issuer fraud losses were both higher in 2023 than in 2011.

The first two columns in figure 23 show the percentage of covered issuers and transaction volume of covered issuers for which the sum of ACS costs and issuer fraud losses was less than or equal to the interchange fee permitted by the interchange fee standards. In 2023, the percentage of issuers with per-transaction ACS costs, including issuer fraud losses, less than or equal to the permissible amount was 80.1 percent, a small increase from 78.0 percent in 2021. Meanwhile, transactions reported by these issuers represented 99.2 percent of all transactions in 2023, down slightly from 99.6 percent in 2021.²⁷ The difference between the percentage of issuers under the maximum and percentage of transactions under the maximum reflecting the fact that issuers under the maximum processed, on average, more transactions than issuers above the maximum.

As figure 23 further shows in the middle two columns, the percentages of both covered issuers and covered transactions with per-transaction fraud-prevention costs below the $0.01 permitted by the fraud-prevention adjustment were substantially lower than those for which per-transaction ACS costs, including issuer fraud losses, were lower than the interchange fee standards. In 2023, 26.3 percent of covered issuers had fraud-prevention costs less than or equal to $0.01, down from 42.0 percent in 2021. Similarly, 10.6 percent of covered transactions had fraud-prevention costs less than or equal to $0.01, down from 34.5 percent in 2021.²⁸

Finally, as shown in the last two columns in figure 23, in 2023, the percentage of issuers with combined costs lower than or equal to the total amount allowed by the interchange fee standards and the fraud-prevention adjustment was 78.1 percent, a small increase from 76.5 percent in 2021. Meanwhile, transactions reported by these issuers represented 99.2 percent of all transactions in 2023, down slightly from 99.6 percent in 2021.²⁹

Background

The Electronic Fund Transfer Act (EFTA) requires the Federal Reserve Board to biennially publish data on the costs incurred, and interchange fees charged or received, by debit card issuers and payment card networks.³⁰ The Board conducted its first data collection in 2010 on a voluntary basis, collecting data from payment card networks and issuers from 2009. The information from the first voluntary data collection assisted the Board in developing Regulation II and was included in a report published in 2011. Six subsequent reports summarized data collected on a mandatory basis since the initial voluntary data collection. The most recent report, released in 2023, contained information through 2021.³¹ The present report is the eighth in the series and contains data through 2023.

Regulation II establishes interchange fee standards under which, for a debit card transaction subject to the standards, an issuer may not receive or charge an interchange fee greater than $0.21 plus 5 basis points times the value of that transaction.³² For a transaction subject to the standards, an issuer may be eligible to receive an additional $0.01 adjustment to the interchange fee amount permitted under the standards if the issuer meets fraud-prevention standards specified in the regulation. Certain transactions—namely those performed using certain debit cards issued pursuant to government-administered payment programs, and those performed using certain reloadable general-use prepaid cards—are exempt from the interchange fee standards.³³ Additionally, an issuer with consolidated assets below $10 billion is not subject to the interchange fee standards, provided the issuer holds the account that is debited.

Data Collections

The Board conducts two surveys to collect information about the debit card industry, both of which are mandatory under Regulation II. The Payment Card Network (PCN) Survey is conducted every year and surveys payment card networks that process debit card transactions. The Debit Card Issuer (DCI) Survey is conducted every two years and surveys issuers that are subject to the interchange fee standards in Regulation II.

The surveys ask respondents to report information on all debit card transactions issued by that issuer or processed by that network in a given calendar year.³⁴ The surveys distinguish between general-use prepaid cards and other debit cards.³⁵ Although Regulation II defines prepaid cards to be a subset of debit cards, for ease of exposition, this report uses terminology that distinguishes between prepaid cards and non-prepaid debit cards. In particular, the report uses the term "debit cards" to mean non-prepaid debit cards. It further uses the term "prepaid cards" to mean general-use prepaid cards that are covered by the definition of debit card in Regulation II.

The surveys also distinguish between dual-message and single-message networks. A dual-message network typically uses separate messages to authorize and clear a transaction. Traditionally, these networks processed signature-authenticated transactions.³⁶ Increasingly, however, transactions processed over these networks may not require signature authentication or may involve other methods of cardholder authentication, such as entry of a personal identification number (PIN) or biometric authentication. By contrast, a single-message network typically uses a single message to authorize and clear a transaction. Traditionally, these networks processed PIN-authenticated transactions.³⁷ Increasingly, however, transactions processed over single-message networks, such as low-value purchases and CNP transactions, may not require PIN authentication. Over time, the distinction between single-message and dual-message networks has become less clear from a functional perspective, as methods of cardholder authentication change and both types of networks develop new functionalities.³⁸ This report continues to categorize networks as single-message or dual-message because this categorization serves to distinguish networks into groupings that are widely used by the industry.

Appendix: Tables

Table 1. Purchase transactions, 2023

Table 2. Card-present and card-not-present transactions, 2023

Table 3. Covered and exempt debit card transactions, 2023

Table 4. Interchange fee revenue, 2023

Table 5. Interchange fee revenue from covered and exempt debit card transactions, 2023

Table 6. Payments and incentives paid by networks, 2023

Table 7. Payments and incentives paid to issuers by networks, 2023

Table 8. Network fees, 2023

Table 9. Network fees paid by issuers, 2023

Table 10. Covered issuer fraudulent debit card activity, 2023

Table 11. Covered issuer fraud losses, 2023

Table 12. Covered issuers by 2023 volume

Table 13. Covered issuer average authorization, clearing, and settlement (ACS) costs, excluding issuer fraud losses, per transaction (dollars), 2023

Table 14A. Covered issuer costs per transaction (dollars) and fraud losses as share of transaction value, 2023

Table 14B. Covered issuer costs per transaction (dollars) and fraud losses as share of transaction value, 2023

Table 15. Covered issuers with costs/losses below the levels permitted by the interchange fee standard and fraud-prevention adjustment, 2023

Table 16. Covered issuers average authorization, clearing, and settlement (ACS) costs, excluding issuer fraud losses, per transaction (dollars), by transaction percentile range, 2023

Footnotes