Improving ACH Network Quality by Reducing Exceptions Request for Comment and Information

Transcription

1 Improving ACH Network Quality by Reducing Exceptions Request for Comment and Information Executive Summary and Rules Description November 11, 2013 RESPONSES DUE BY MONDAY, JANUARY 13, 2014 NACHA requests comments on a proposal to amend the NACHA Operating Rules that is intended to improve ACH Network quality by reducing the incidence of ACH exceptions. Generally speaking, an exception is an ACH transaction: 1) that an RDFI cannot straight-through process due to some incorrect data within the transaction; or 2) the validity of which is disputed by a Receiver customer of an RDFI. 1 These types of exceptions impose costs on RDFIs and impact their Receiver customers. Exceptions can be caused by poor origination practices, as well as through use of the ACH Network by high-risk Originators. A reduction in exceptions would provide a cost-reduction benefit to RDFIs, improve their Receiver customers satisfaction with ACH payments, and shift some of the cost of exceptions to the DFI in the best position to avoid those costs. The proposed rule is one component of NACHA s overall risk management strategy, which also includes industry education and information sharing, the development of tools and services, as well as risk management-related Rules and enforcement. A related and complementary rule proposal on ACH Network Risk and Enforcement is also open for comment concurrently (see Comments are due by Monday, January 13, 2014 at 5:00 p.m. Eastern Time. NACHA STAFF CONTACTS Return comments to: Questions: Maribel Bondoc, Manager, ACH Network Rules Fax: (703) mbondoc@nacha.org Michael Herd, Senior Managing Director, ACH Network Rules mherd@nacha.org 1 For the purpose of this proposal, an ACH debit which is not paid due to insufficient or uncollected funds ( NSF ) is not included within the scope of ACH exceptions. Additional discussion on NSFs is included later in this document.

2 Rule Proposal Description; November 11, 2013, Page 2 Part I: Proposal Summary This proposed rule ( Rule ) is intended to improve ACH Network quality by reducing the incidence of ACH exceptions. Quality in the context of the ACH Network is generally considered to be an environment with reasonably low rates and incidences of returned transactions. For RDFIs, ACH quality can also be considered to be an environment with reasonably low rates and incidences of exceptions, which frequently involve customer service contacts and a manual resolution, and which also may result in the RDFI transmitting a returned entry or a Notification of Change ( NOC ) to the ODFI. The Rule would improve ACH quality and reduce exceptions by establishing a system of economic incentives for ODFIs to improve the quality of the ACH transactions they originate, and for both RDFIs and ODFIs to use and comply with the existing NOC process within the ACH Network. The proposed economic incentives are as follows: 1. A fee for certain Return Entries for which an RDFI returns an ACH transaction to an ODFI due to incorrect account data within the transaction ( Erroneous Data Fee ); 2. A fee for certain Notifications of Change (NOCs) for which an RDFI corrects information within an ACH transaction and transmits an NOC to the ODFI of the transaction ( NOC Fee ); 2 3. A fee for certain Return Entries and Extended Return Entries for which an RDFI returns an ACH transaction to an ODFI due to a problem with the Receiver s authorization, such as a Receiver s dispute of the validity of the transaction ( Unauthorized Entry Fee ). For each of these incentives, the fee would be paid by the ODFI and passed through to the RDFI to partially offset the RDFI s costs for exception processing and customer service. 3 Under this proposed Rule, NACHA is not the beneficiary of any of the proposed fees. No portion of any fees proposed under this Rule would be passed to NACHA for NACHA s benefit. The primary objective of this set of incentives is to improve ACH Network quality by reducing the incidence of returns and exceptions. By establishing fees for specific types of exceptions, ODFIs would have a direct financial incentive to implement processes and tools with their Originators to reduce the number of such returns and exceptions. ODFIs with the highest return rates would have the greatest incentives to work with their Originators to reduce those rates. Economic incentives applied to certain returns would provide additional impetus for ODFIs to address ACH origination practices that result in returns, such as poor data capture or unclear, confusing or deceptive authorizations. Economic incentives applied to certain NOCs provide additional impetus for RDFIs to send NOCs, and for ODFIs to encourage their Originators to use the corrected information. 2 The NOCs covered by this proposal are those that represent incorrect information sent by the ODFI, and not those that represent changes to routing and account information due to RDFI mergers and acquisitions. 3 An NOC bears an SEC Code of COR. The RDFI of the original Entry retains its role as the RDFI, even though it is initiating the COR Entry; i.e., it is not considered an ODFI by virtue of transmitting an NOC. See Section 3.9.

3 Rule Proposal Description; November 11, 2013, Page 3 Regardless of the previous adoption of risk-management related rules, the ACH Network is still used by some high-risk originators 4, such as debt collectors, payday lenders, credit repair services, sweepstakes, travel clubs, and online and telemarketers, occasionally resulting in episodes of high rates of customer disputes, returned transactions, consumer and business harm, and damaged reputations for both the ACH Network and RDFIs alike. More recently, such episodes have resulted in attention by certain regulatory and law enforcement agencies, as well as the news media, and improper assignment of responsibility to RDFIs for solving high-risk practices that are better addressed by ODFIs. A system of Quality Fees would serve to provide an additional incentive for ODFIs of such high-risk Originators to adequately monitor these Originators ACH activity and returns, and to provide some economic relief to RDFIs if and when any such episodes occur in the future. If such a system is successful in improving ACH quality, a reduction in exceptions would directly lead to cost savings by RDFIs. Additionally, economic incentive fees passed through to RDFIs would provide some cost recovery for a portion of RDFIs costs for exception processing, and would shift this cost to the ODFIs that are responsible for the transactions that caused the exceptions. Those ODFIs that are responsible for generating the greatest volume of ACH exceptions, such as those that process for high-risk originators, would incur the greatest impact from the shift in these costs. To the extent that an ODFI does not reduce its unauthorized and administrative return rates in response to these financial incentives, it would be appropriately and fairly bearing some of the costs that these exceptions impose on RDFIs. These quality incentives should not be seen as the complete solution for ACH quality issues. They do not replace the NACHA System of Fines and enforcement actions already in place to address Rules violations, nor do they replace other tools to improve quality. Rather, the proposed ACH quality incentives should be seen as part of a process of continuous improvement in the ACH Network. As a complementary approach to reduce exceptions in the ACH Network, NACHA has concurrently issued a related proposed rule that would improve the ability to identify and enforce the Rules against outlier Originators that are responsible for generating the most disproportionate levels of exceptions (see Part II: Background NACHA s Previous Risk Management Strategy, and Unauthorized and Administrative Returns NACHA s previous risk management strategy covering the time period focused mainly on unauthorized ACH debits. Individual components of that strategy included: the Network Enforcement Rule; the Company Name Rule; Direct Access Registration; the Terminated Originator Database; and many other rules and industry educational efforts. During that time period, the overall unauthorized debit rate on the ACH Network decreased from 0.06 percent (i.e., 6 out of 10,000 debits) in 2004 (the year before the strategy began) to For purposes of this RFC, high-risk originators generally means those originators in certain lines of business that result in exceptions or returned transactions substantially higher than the average across the ACH Network. Additional discussion of high risk activities and examples or high-risk industries can be found in OCC Bulletin at

4 Rule Proposal Description; November 11, 2013, Page 4 percent in Further, the number of unauthorized debits fell slightly from 3,062,114 in 2004 to 2,942,349 in A closer look at unauthorized volume shows that a significant portion of the decrease occurred during , which coincides with the period immediately following the effective date of the Network Enforcement Rule and Company Name Rule. More recently, the decline in the unauthorized debit rate has leveled off, with the rate of 0.03 percent in 2012, and projected for 2013, being identical to that of Further, the number of unauthorized debits is increasing again in proportion with increases in ACH volume. Regardless of the past success, unauthorized debits remain a significant pain point for RDFIs. Inherent to any unauthorized debit is a Receiver customer whose account has been debited, and for which the Receiver is disputing the validity of the debit, resulting in customer contact with his/her financial institution, and attendant Regulation E dispute resolution procedures. All this results in costs for the RDFI to handle the customer s dispute, and frequently can result in dissatisfied customers as well. While focusing on unauthorized transactions, the risk management strategy of did not substantially address administrative returns and other exceptions such as NOCs. High administrative return rates also can be associated with fraud, or indicative of poor-quality origination that also imposes costs on RDFIs. 5 As measured by the number of debits returned bearing the R-codes R05, R07, R10 and R29.

5 Rule Proposal Description; November 11, 2013, Page 5 In contrast to unauthorized debits, the number of administrative returns 6 increased during the time period of the risk management strategy, from 21,962,881 in 2005 to 29,953,930 in 2012, although the rate of return has not increased. A closer look at the data shows that the administrative return rate for ACH debits declined from 0.30 percent in 2005 to 0.23 percent in 2009, largely due to the rapid adoption of ARC transactions, which inherently had very low return rates. There has been no further rate decline since 2009, which is consistent with the plateau and subsequent decrease in ARC volume. For ACH credits, the administrative return rate has increased slowly but consistently throughout the period from 0.09 percent to 0.11 percent. To further drive down both the rate and the number of administrative returns, the risk management strategy updated in 2012 includes a focus on improving ACH quality, potentially through the implementation of economic incentives. Part III: RDFIs Costs to Handle Exceptions To investigate RDFIs actual costs of handling exceptions, NACHA collected data from receiving financial institutions of various types and sizes. NACHA asked these RDFIs to provide cost information related to their ACH operations labor, systems, and office space allocated to full-time equivalents that handle exception processing. Thirteen financial institutions provided data regarding account data exceptions, and fourteen provided data regarding authorization-related exceptions. These costs were then divided by the volume of exceptions to arrive at a cost per exception. Note that this data covers only the costs to RDFIs of their ACH operations, and not additional costs related to customer contact and service via branch, phone, Internet, etc., nor any compliance costs for Regulation E dispute resolution and 6 As measured by the number of debit and credit transactions returned bearing the R-codes R03 and R04.

6 Rule Proposal Description; November 11, 2013, Page 6 NACHA rules for obtaining a customer s written statement. Therefore, the cost calculations and estimates below should be considered very conservative, understating all-in costs for these exceptions. Perhaps not surprisingly, reported costs varied widely, but also inversely correlate very strongly with size (i.e., unit cost goes down as volume increases). Costs for handling administrative errors and corrections ranged from a low of less than one cent per exception for a large RDFI (that presumably has automated much of this work), to a high of $63.20 per exception for a small RDFI (with a low volume of administrative exceptions over which to spread its fixed costs). The average cost of per-exception handling among this group of RDFIs is $9.40. The weighted average cost which factors in each RDFI s volume is $0.26. This significant difference between average and weighted average again shows the strong inverse correlation of unit cost and volume. As might be expected, the costs reported for unauthorized exceptions are significantly higher than for administrative exceptions, but still correlate with size. Costs ranged from a low of $2.30 per unauthorized exception for a large RDFI, to a high of $ per unauthorized exception for a small RDFI. The average cost of per-unauthorized exception handling among this group of RDFIs is $ The weighted average cost which factors in each RDFI s volume is $4.99. As with administrative exceptions, this significant difference between average and weighted average shows the strong inverse correlation of unit cost and volume. Based on extrapolating the cost data reported by these 13 RDFIs, the cost of ACH operations to all RDFIs of handling all 46,381,371 7 account data-related returns in 2012 that would be covered by the proposed Rule was $12.1 million on a weighted average basis. Also based on extrapolating the cost data reported by these 13 RDFIs, the cost of ACH operations to all RDFIs of handling all 11,527,812 8 NOCs in 2012 that would be covered by the proposed Rule was $3.0 million on a weighted average basis. Again, based on extrapolating the cost data reported by the 14 RDFIs, the cost of ACH operations to all RDFIs of handling all 6,433,162 9 authorization-related returns in 2012 that would be covered by the proposed Rule was $32.1 million on a weighted average basis. The total cost of ACH operations to RDFIs in 2012 for these three categories combined was $47.2 million using a weighted average basis. As noted above, these reported costs do not include other cost areas such as customer care. A study in 2006 by Global Concepts (primarily of large and regional banks, using 2005 data) estimated a median labor cost of $8.54 for an RDFI to obtain a customer s written statement 7 This figure does not precisely match that of the research firm because they cover different time periods, though both time periods reviewed cover 12 months. 8 This figure does not precisely match that of the research firm because they cover different time periods, though both time periods reviewed cover 12 months. 9 This figure does not precisely match that of the research firm because they cover different time periods, though both time periods reviewed cover 12 months.

7 Rule Proposal Description; November 11, 2013, Page 7 contesting the validity of a transaction. If this cost figure is still valid, then RDFIs experienced an additional $25.1 million in customer service costs in 2012 for these customer service contacts related to ACH debits that consumers disputed as unauthorized. This raises RDFIs total costs for exception handling in 2012 to approximately $72.3 million. Via this RFC, NACHA solicits comments from other RDFIs on their costs to handle exceptions, and welcomes the submission of cost data from additional RDFIs. A template for RDFIs to report the cost to their ACH operations to handle exceptions will be provided on NACHA s web site accompanying this RFC. Part IV: Fee Ranges for Evaluation For the purpose of evaluating this proposed Rule, NACHA asks respondents to consider the following ranges for each fee type. The proposed types and ranges of fees were developed and recommended through a significant research effort conducted by an expert payments consulting firm that included the RDFI cost data cited above. The firm s methodology, analysis and recommendations are appended to this rule proposal in the Appendix beginning at page 15. For each type of fee and the suggested ranges, NACHA requests comment on whether a fee that is within the suggested range, or some other specific fee amount, would have an appropriate incentive effect to improve origination practices while not unduly impacting use of the ACH Network (i.e., the fee should be viewed as fair to both the ODFI and RDFI). 1. Erroneous Data Fee Evaluate a range of $ $0.40 per account data-related return For each return using the R-codes R02, R03, R04, R20, R24, R39, and R50, the ODFI would be assessed a fee estimated to be in the range of $ $0.40, which would be passed through to the RDFI. In 2012, there were a combined 46,381,371 returns using these R- codes. An Erroneous Data Fee in this range would result in fees of $4.64 million to $18.55 million annually assessed to ODFIs and passed through to RDFIs. Based on the industry average return rate for these combined R-codes in 2012 of 0.28 percent, the fee range would result in average cost per entry across all ODFIs of $ to $ ODFIs with returns rates significantly higher than the industry average would bear proportionately higher per-entry costs. For example, an ODFI with an administrative return rate of 1 percent would bear a cost of $0.001 to $0.004 per entry. 2. NOC Fee Evaluate a range of $ $0.75 per NOC 10 For each NOC using the change code reasons C01, C05, C06, C07, C08, C09, C13 and C14, the ODFI of the underlying entry would be assessed a fee estimated to be in the range of 10 The change code reasons C02 and C03 relate to a change in routing information due to an RDFI s merger or consolidation, and so therefore are not proposed to be included within the scope of NOC Fees.

8 Rule Proposal Description; November 11, 2013, Page 8 $ $0.75, which would be passed through to the RDFI of the underlying entry. In 2012, there were a combined 11,527,812 NOCs using these C-codes. 11 An NOC Fee in this range would result in fees of $2.9 million to $8.6 million annually assessed to ODFIs and passed through to RDFIs. Based on the industry average rate for these combined C-codes in 2012 of 0.07 percent, the fee range would result in average cost per entry across all ODFIs of $ to $ per entry. ODFIs that receive NOCs at rates significantly higher than the industry average would bear proportionately higher per entry costs. For example, an ODFI with an NOC rate of 1 percent would bear a cost of $ to $ per entry. 3. Unauthorized Entry Fee Evaluate a range of $ $2.50 per authorization-related return For each return using the R-codes R05, R07, R08, R10, R29, R37, R38, R51, R52 and R53, the ODFI would be assessed a fee estimated to be in the range of $ $2.50, which would be passed through to the RDFI. In 2012, there were a combined 6,433,162 returns using these R-codes. An Unauthorized Entry Fee in this range would result in fees of $9.71 million to $16.08 million annually assessed to ODFIs and passed through to RDFIs. Based on the industry average return rate for these combined R-codes in 2012 of percent, the fee range would result in average cost per entry across all ODFIs of $ to $ ODFIs with returns rates significantly higher than the industry average would bear proportionately higher per entry costs. For example, an ODFI with an unauthorized return rate of 1 percent would bear a cost of $0.015 to $0.025 per entry. Total Cost Recovery of All Three Fee Categories As described above, the total of all three categories of proposed quality fees ranges from a low of $17.1 million annually to a high of $43.3 million annually. Compare this to the total estimated annual costs to RDFIs of ACH operations described in the previous section of $47.2 million, not including non-ach operations costs, such as customer service. As recommended by the research, the estimated fee levels do not represent full cost recovery for RDFIs, but do acknowledge and attempt to ameliorate the significant cost of dealing with exceptions. On a per-entry basis, the total of all three fees amounts to a range of $ to $ In other words, an ODFI whose overall return/noc rates are exactly at the industry average would incur quality fees that amount to between $ and $ per entry. ODFIs with higher rates would incur a greater amount of fees. An ODFI with a rate of 1 percent in each of the three categories would incur fees on a per-entry basis of $ to $ The change code C14 was added in March 2013; therefore there was no volume of C14 in 2012 included within this total.

9 Rule Proposal Description; November 11, 2013, Page 9 Part V: Other Topics A. Insufficient Funds High return rates for insufficient funds can often be a red flag for fraudulent or otherwise problematic origination activity. Most ACH debit Originators, however, experience a much lower, more routine level of NSFs over which they likely have very little control. Therefore, returns of insufficient and uncollected funds are not proposed to be covered by the system of incentive fees. NACHA requests comment on the appropriateness of excluding returns for NSF from the proposed system of Quality Incentive Fees. B. Check conversion The existing NACHA rules for check conversion (ARC, BOC, and POP) require Originators to use a reading device to capture the routing number and account number from the Receiver s check. With respect to the proposed Erroneous Data Fee, one view could be that ODFIs should not be assessed a fee when the Originator has accurately obtained the routing and account number from a check provided by the Receiver but subsequently receives a return with a code of R03 or R04. A different view could be that most R03 and R04 returns are due to misreads, and are not related to transactions with accurate information. Additionally, under the NOC Fee proposal, RDFIs would not receive fees due for NOCs that relate to merger and acquisition activity; therefore, they would not be compensated for NOCs related to check conversion transactions originated after the change in routing or account information. On balance, NACHA believes the benefits to the ACH Network of assessing the Erroneous Data Fee to discourage poor origination practices that are likely to result in R03 or R04 returns outweigh the burden on Originators; and so check conversion transactions are proposed to be included within the scope of the Erroneous Data Fee. NACHA specifically requests comment on the applicability of the proposed Erroneous Data Fee to check conversion transactions. C. Stopped Payment As proposed, ODFIs would be charged an Unauthorized Entry Fee for each Return using the R08 (Payment Stopped) R-code. This is because high numbers of R08 Returns are often indicative of a fundamental and pervasive problem with an ODFI s origination practices such as deceptive authorization processes or language, and inconvenient or unavailable cancellation mechanisms. However, NACHA recognizes that in some cases, an R08 Return may result from the consumer Receiver s behavior without any fault of the ODFI or Originator. For example, an Entry may be returned pursuant to the R08 R-code when a consumer stops the payment of an otherwise valid Entry but fails to notify the Originator to prevent transmission of the Entry. In determining whether to apply the Unauthorized Entry Fee to R08 Returns, NACHA must balance the risk that ODFIs (and ultimately Originators) will be asked to pay some costs for unavoidable Returns versus the significant burden that is imposed on the ACH Network and

10 Rule Proposal Description; November 11, 2013, Page 10 other participants by Originators who trigger large volumes of R08 Returns as a result of poor origination practices. On balance, NACHA believes the benefits to the ACH Network of imposing a fee to discourage poor origination practices that are likely to result in R08 Returns and compensate RDFIs outweigh the burden of a relatively small amount of R08 Returns that any individual Originator might be asked to absorb as a result of occasional consumer error in failing to notify the Originator. It is worth noting in this regard that the proposed fee would not fully compensate RDFIs in any event. Nonetheless, NACHA specifically requests comment on whether inclusion of the R08 Return strikes the right balance for enhancement of ACH Network quality without unduly burdening ODFIs and Originators. D. Collection and Disbursement of Fees In order for a system of Quality Fees to function effectively, the fees must be assessed, collected and disbursed accurately and timely. NACHA is currently exploring a number of options for such a system, understanding that there may be specific features of any system that would make the system either viable or unworkable for ACH Network participants. Part VI: Examples The following examples are provided to illustrate how Quality Fees might work in practice. A. Payroll originator / ACH credit An employer adds a newly hired employee to its payroll Direct Deposit file. The first Direct Deposit intended for this employee has an incorrect account number (perhaps key-entered incorrectly). The ACH credit in not postable at the RDFI, and is automatically placed in an exceptions review queue. The RDFI s ACH Operations staff determines the correct account for posting the credit, and sends an NOC to the ODFI of the ACH credit. The ODFI of the ACH credit would be assessed an NOC Fee, which would be passed to the RDFI. Under this proposed Rule, the NOC Fee would be in the range of $0.25 to $0.75. Across all its 500 employees, the employer experiences just five Direct Deposit errors each year (for example, an employee closed a bank account and didn t inform the employer). The total amount of the NOC Fees and Erroneous Data Fees assessed to its ODFI for the year is likely to be no higher than about $2.50. Spread across 12,000 Direct Deposits annually, the fees amount to $ per transaction. B. Utility bill / ACH debit A utility company customer goes to the company s web site to pay a bill. The customer enters his bank account number inaccurately. The resulting ACH debit is returned by the RDFI as an R03 (No Account/Unable to Locate Account). The ODFI of the ACH debit would be assessed an Erroneous Data Fee for this account data-related return, which would be passed to the RDFI. Under this proposed Rule, the Erroneous Data Fee would be in the range of $0.10 to $0.40. Across all its customers, the utility company has an administrative return rate of 0.05 percent (i.e., 1 in every 2,000). It originates 1 million ACH debits per year, resulting in 500 administrative returns. The total amount of the Erroneous Data Fees assessed to its ODFI for

11 Rule Proposal Description; November 11, 2013, Page 11 the year would be between $50 - $200. Spread across its 1 million origination volume, the fees amount to $ to $ per transaction. C. High-risk Originator / ACH debits An online payday lender collects customers routing and account numbers while enrolling them for payday loans. During a calendar month, the lender originates 10,000 ACH debits to collect loan payments, of which 15 percent are returned collectively as R02, R03 and R04, and another 1 percent are returned as R10. Additionally, the lender resubmits all NSFs, and 500 of those are returned R08 (stop payment). 12 Collectively, within this calendar month, the lender has received 1,500 account data-related returns and 1,000 authorization-related returns. The ODFI of the lender would be assessed Erroneous Data Fees on the 1,500 administrative returns; and Unauthorized Entry Fees on the 600 authorization-related returns. Under this proposed Rule, the Erroneous Data Fee would be in the range of $0.10 to $0.40, so the total to the ODFI would be between $100 and $400. Under this proposed Rule, the Unauthorized Entry Fee would be in the range of $1.50 to $2.50, so the total to the ODFI would be between $900 and $1,500. If this level of return activity persisted across a full calendar year, the total amount of the Erroneous Data Fees assessed to its ODFI would be between $1,200 and $4,800; and the total amount of the Unauthorized Entry Fees assess to its ODFI would be between $10,800 and $18,000. Spread across the lender s annual volume of 120,000 debits, the total fees amount to $0.10 to $0.19 per transaction. D. Bill payment scam / ACH debits A scam is perpetrated in which unsuspecting consumers are informed that President Obama will pay your bills if they would just provide their Social Security Numbers in exchange for the correct routing and account numbers to use. These consumers use these routing and account numbers on several billers web sites to pay their bills. Because the billers do not receive returned ACH debits for 2 days, the consumers think their bills really have been paid successfully, and through social media platforms encourage their families and friends to do the same thing. Before too long, 50,000 ACH debits from several dozen national and local billing companies have been originated using a single RDFI s otherwise valid routing number. The RDFI is able to return all 50,000 debits timely as R04s. The ODFIs of the debits are assessed Erroneous Data Fees on each of the R04s, all of which would be passed to the single RDFI. Under this proposed Rule, the Erroneous Data Fee would be in the range of $0.10 to $0.40, so the victimized RDFI in this case would receive recompense of between $5,000 and $20, A concurrent and complementary proposed rule would establish return rate thresholds for certain categories of returns, and for an overall return rate (see The ODFI of an Originator with return rate as shown in this example could also be subject to enforcement and sanctions or fines for exceeding those return rate thresholds. 13 Billers could seek to avoid Erroneous Data Fees in such a circumstance by validating account information when a customer uses new routing and/or account information, or by performing velocity checks to detect multiple or repeated use of the same routing and account information across multiple billing accounts.

12 Part VII: Impact of the Proposal Benefits of the Proposal Request for Comment: Improving ACH Network Quality by Reducing Exceptions Rule Proposal Description; November 11, 2013, Page 12 Arguably, all ACH Network participants will experience at least some indirect benefit through higher quality, fewer exceptions, greater customer satisfaction, and lower reputational risk for the Network. Any beneficial impact of the proposed Rule on ACH Network quality should be visible through monitoring of the return and NOC activity covered by the proposed Rule. While ODFIs will directly bear the costs of the Quality Fees, the system of fees might encourage ODFIs to perform enhanced risk management and monitoring of Originators with high volumes or rates of entries that result in exceptions. Conceivably, service providers will bring new tools and services to market to enable ODFIs to avoid some Quality Fees. ODFIs might be able to provide new tools and services to Originators. Ultimately, the goal is to incent different actions to improve overall Network quality, recognizing that this is only one aspect of risk management and quality focus of the ACH Network. RDFIs will directly benefit from the Quality Fees that are passed to them, which will provide a partial cost recovery for externally-imposed costs. Of perhaps greater benefit, any reduction in the number of exceptions will also directly benefit RDFIs through the direct avoidance of costs that would have been caused by those exceptions, and indirectly through fewer customer service contacts and less dissatisfaction. Receivers would directly benefit from fewer unauthorized entries or other entries that result in exceptions. While Originators would likely bear at least some of the costs initially borne by their ODFIs, it is conceivable that as a result of this focus on exceptions their ODFIs and other service providers would bring new tools and services to market to reduce the number of transactions that cause exceptions and help discourage bad actors from using the ACH Network. Costs to Comply with the Proposal ODFIs would directly bear the costs of the Quality Fees. Further, ODFIs would have costs to the extent that they work with their Originations to implement systems, processes and tools to lower return rates. Success in lowering return rates, however, would lessen the cost burden of fees. The proposed Rule does not require or envision that RDFIs would need to make changes to their exception handling. Therefore RDFIs would not incur direct costs due to changes in exception handling. RDFIs would likely have some costs to implement systems to account for the receipt of Quality Fees. Either the ACH Operators or NACHA, or both, likely would incur costs to implement a system to assess, collect and disburse Quality Incentive Fees. Receivers likely would not incur any costs as a result of this proposed Rule.

13 Rule Proposal Description; November 11, 2013, Page 13 Originators would likely bear at least some of the costs initially borne by their ODFIs, such as through transaction, returns, and other service fees. Alternatively, Originators might incur other costs in an attempt to improve the quality of their origination, for instance, by implementing new processes to detect fat-fingering of account information. Part VIII: Effective Date This Rule is proposed to become effective on March 19, 2015, which should provide sufficient time for the industry to prepare and implement. NACHA and the ACH Operators will need time to generate the data to determine the collection and disbursement of the fees. ODFIs and RDFIs will need time to properly account for fees paid and received. ODFIs will need time to work with their Originators, if desired, on implementing systems, tools and processes to reduce returns and thereby minimizing fees they incur. Part IX: Rules Framework The proposed Rule would add a new Section in Article One (General Rules) to establish Quality Fees. One subsection establishes the obligation of the ODFI to pay the fees to the respective RDFIs. A second subsection establishes the authority of the NACHA staff, and the methodology used, to determine the specific amounts of the Quality Fees. The Quality Fees would further be defined in Article Eight (Definitions).

14 Rule Proposal Description; November 11, 2013, Page 14 Request for Information Part X: NACHA Requests Information on Additional Concept of a Prenote Validation Fee In addition to the proposed Rule on Improving ACH Network Quality described above, NACHA also requests information on a related but separate idea that would also improve ACH quality by providing an incentive for ODFIs and RDFIs to make fuller use of the existing prenotification transaction and process. The concept is as follows: A fee paid by the ODFI to the RDFI for a new type of NOC in which an RDFI responds to a prenote when the account information in the prenote is valid ( Prenote Validation Fee ). The new NOC would be defined to mean that the RDFI has validated the account information in the prenote and that it is correct. For such an NOC, a fee would be paid by the ODFI of the prenote and passed through to the RDFI. 14 This concept would support NACHA s efforts to address Originators stated desire for the ACH Network to support account validation. This desire for account validation has been documented in several recent NACHA surveys. Originators today might be hesitant to use a prenotification because, among other reasons, they do not receive a response when the account information in the prenote is valid. The basic framework is structured as no news is good news. Some Originators might use, or increase their use of, prenotes if they knew they would receive a response regardless of whether the information was valid or invalid. An initial step in a separate rulemaking will reduce an Originator s waiting period for initiating live entries after sending a prenote, from six to three Banking Days. 15 Additional future steps could include ways to further quicken the prenote process and/or provide validation of additional account or balance information. These future steps could result in additional Rules proposals or other projects. Each of these measures could provide progressively more information to more robustly provide value associated with account validation. While a Prenote Validation Fee itself is not directly an incentive for ODFIs to improve quality based on a return or a correction to a specific transaction, functionally it is nearly identical to the NOC Fee being proposed, and so therefore could be implemented utilizing the same NOC Fee rules and implementation framework. Further, a Prenote Validation Fee could be an incentive for Originators to make fuller use of prenotes if the value of receiving an explicit, positive response from the RDFI justifies the amount of the fee. Under such a concept, an Originator sending a prenote would be assured of receiving a response, which would take one of three forms: An NOC meaning account number is valid ; 14 As with the NOC Fee, the Prenote Validation fee would be paid by the ODFI of the prenote, and passed to the RDFI that initiates the NOC. 15 Effective September 19, 2014.

15 Rule Proposal Description; November 11, 2013, Page 15 An NOC meaning account number is not valid and corrected data is included; or A return. For purposes of evaluating the concept of a Prenote Validation Fee, NACHA suggests the same range as for the NOC Fee i.e., $0.25 to $0.75 per NOC sent in response to a prenote meaning that the account information is valid. Because the RDFI is sending a NOC to indicate either that the account information is valid or invalid, presumably the same fee amount would be both appropriate, and easier to implement that some different fee amount.

16 Rule Proposal Description; November 11, 2013, Page 16 Appendix A Part XI: Research Performed by Payments Industry Consultants NACHA engaged a payments industry consulting firm to perform research into the benefits, costs, impacts, and feasibility of a system of ACH quality fees. It is worth quoting directly from the research paper at length: 16 Executive Summary The purpose of the proposed incentives is to reduce the incidence of ACH exceptions which is a goal of NACHA s Risk Management Strategy while providing some cost recovery to RDFIs for the expense of handling returns and exceptions caused by Originator practices. By ensuring the efficiency of operations and fairness for all participants, ACH Quality Incentives should help the ACH Network remain relevant and competitive in an increasingly demanding payments environment. These concepts should not be seen as the complete solution for ACH quality issues. They do not replace the NACHA system of fines and enforcement actions already in place to address rules violations. The ACH Network should also continue to develop operational, technical and commercial approaches to improve quality. Rather, the proposed incentives should be seen as part of a process of continuous improvement. Background The ACH Network relies on ODFIs and their Originators to ensure the accuracy and validity of ACH entries submitted for clearing. Originators are also responsible for obtaining and maintaining authorization from Receivers for ACH debit entries, and for obtaining routing data from Receivers for both debits and credits. While the vast majority of Originators cause few problems, others create problems for ODFIs, Receivers and especially RDFIs. In some instances, the problems result from poor operational practices such as not checking for accurate routing and account data, or the failure to discontinue pre-authorized Direct Payments when requested by Receivers. Other problems have less benign causes such as fraudulent debit origination or payments related to scams. Some Originators have high return rates because they are in high-risk businesses. Poor customer communications can result in high return rates for some Originators if Receivers do not understand that they will be debited, or do not recognize the name of the parties debiting their accounts. Regardless of the cause, poor quality of origination can result in inconvenience for Receivers, return and regulatory risk for ODFIs, and additional cost for RDFIs. 16 Most of the substantive portions of the research paper are quoted here, with some editing and paraphrasing for length and clarity.

17 Rule Proposal Description; November 11, 2013, Page 17 The ACH Network has made remarkable strides in reducing unauthorized entries. Nonetheless, nearly 3 million ACH debits are returned unauthorized every year, and an additional 3 million ACH debits are returned for other reasons that are related to authorization. In addition, more than 45 million entries are returned every year because of problems with account data. ACH Quality Framework When designing a program to incent different behavior or actions, it is essential to balance several factors. The first is to define what the program is intended to accomplish, i.e., its objectives. It is also important to understand and mitigate the potential for unintended consequences. The program should not be overly complex, which can lead to difficulty in implementation. And the program should be perceived as fundamentally fair which can be difficult to achieve when there are multiple points of view. (We) employed the following framework for developing and evaluating quality incentive options including objectives, design considerations and potential incentive types. Quality Incentive Objectives Improved Quality: The primary objective is to reduce ACH exceptions, either by increasing the cost of poor quality (disincentive), or potentially decreasing the cost of higher quality (incentive). Cost Recovery: The program should also help financial institutions recover some of the cost of poor quality. In most cases, this means providing for at least some cost recovery to the RDFI for the work they are obligated to perform to correct problems caused by the Originator. Successfully reducing exceptions would also provide RDFIs with a second benefit reducing exception costs in proportion to that reduction in the volume of exceptions. Simplicity: The program should be easy for participants to understand, easy to implement, and easy to maintain. Maintain Ubiquity: The program should encourage continued RDFI participation in the ACH Network, and should not in any way compromise the ubiquity of the ACH Network by placing unreasonable restrictions on origination, creating excessive costs for Originators or ODFIs, or imposing technical requirements that become a barrier for participation. Minimize IT and Operational Requirements: The program needs to be universal. Any change or investment ODFIs or RDFIs need to make becomes an impediment to adoption, and Operator IT modifications delay implementation. Proactive Solutions: The program should encourage the development of network and third-party solutions to quality problems, either directly by providing funds for solution development, or indirectly by creating a business case for ODFIs to improve quality. For example, ODFIs might pay for an account validation service if it allows them to avoid paying fees for incorrect data.

18 Rule Proposal Description; November 11, 2013, Page 18 Don t discourage the use of ACH: Financial incentives or disincentives should not significantly increase the cost of origination, either directly or indirectly, to the extent that they undermine the attractiveness of ACH as a payment option. While the program will provide an economic incentive for improving the quality of forward work, penalizing bad practice is not the primary objective of the quality incentive program. NACHA already has a system of fines and enforcement actions to deal with rule violations and other harmful behaviors. Potential Incentive Types * * * Return Fee: a surcharge on each returned entry (may apply to only some SECs or return codes); Exception Fee: a charge to the ODFI for certain types of RDFI exceptions caused by poor origination data regardless of whether the item is returned; NOC Fee: a surcharge on each NOC (COR SEC Code) received by a financial institution; Threshold-Based Incentive: a fee that only applies to ODFIs that exceed a designated return rate for a certain class of returns; Tiered-Incentive: a fee that varies based on quality metrics, exception/return rates or exception/return volumes, and that may include tiers that are wholly exempt from a fee; Discounts/Allowances: a deduction from the volume of returns subject to a fee, in order to account for exceptions that are beyond the control of the ODFI or Originator; Forward Item Cost Tiers: surcharging forward item fees based on Originator return rate. Return fees and NOC fees are preferable to exception fees because they are simpler to administer. Returns and NOCs are already handled through the network. Exceptions that do not result in returns or NOCs would need to be tracked, and would be subject to dispute by ODFIs. Thresholds and tiers make it possible to tailor fees to enhance their incentive effect, address concerns about fairness and mitigate the cost of normal levels of returns. However, thresholds and tiers are much more complicated to administer, and would require additional technical implementation. Analysis of ACH Returns At the Network level, ACH exceptions generally result in returned entries. The reason for an ACH return is denoted by a return reason code. There are 55 return reason codes, some of which apply to all types of ACH entries, and others that are only used for designated SEC codes.

19 Rule Proposal Description; November 11, 2013, Page 19 Most returns can be grouped into three categories of return codes, according to the underlying cause of the return: 1. Funding Issues returns due to insufficient funds, uncollected funds or frozen accounts 2. Account Data Issues returns due to incorrect account data or a problem with the account, including invalid account numbers, inability to locate the account, closed accounts, deceased account holder or deceased beneficiary 3. Authorization Issues returns due to a problem with authorization, including unauthorized, revoked authorization, stopped payments or RDFI customer disputes. This look at return rates points to a several key insights with respect to the three main categories of returns: Funding Issues (i.e., insufficient funds) are the majority of returns, in most cases reflecting an inherent aspect of debit payments. o Unlike other returns, RDFIs are usually compensated for insufficient funds returns by the Receiver. o Among ACH debits, however, some have much lower return rates notably ARC, CCD and POP. o The low return rate for POP is likely because these transactions are typically screened at the point of sale using check verification or guaranty services. o More widespread use of verification, guaranty or online authorization may reduce the NSF return rate for other types of ACH debit transactions. Account Data Issues are the most widespread cause of quality exceptions (as opposed to NSF) occurring at a rate of 0.28 percent of ACH entries originated. In addition, another 0.11 percent of entries result in account-data corrections NOCs and thereby avoided a return. In most cases, both administrative returns and NOCs are the result of Originator errors. As such, these are a prime focus of ACH quality incentive. Authorization Issues are the other significant cause of quality-related returns. As a point of comparison, the authorization-related return rate for ACH (0.05 percent) is slightly lower than reported fraud rates for credit cards (0.07 percent) and signature debit cards (0.06 percent). This is impressive when you consider that the ACH Network has no online authorization process. However, unauthorized debits can cause harm to RDFIs and Receivers, and are a significant source of costs to RDFIs. These three categories accounted for 97 percent of all ACH returns in the 12 months ending in the 2 nd quarter 2012 (the period we studied); we focus our attention on quality incentives on the latter two categories.

20 ACH Quality Incentive Options Request for Comment: Improving ACH Network Quality by Reducing Exceptions Rule Proposal Description; November 11, 2013, Page 20 (We) evaluated three potential incentives that meet the operational imperatives of: (1) addressing issues that have a significant impact on ACH participants, and (2) allowing for Originator or ODFI actions to modify behavior and reduce their exceptions. ODFI fees for Account Data returns o Fees charged to ODFIs and passed through to RDFIs related to the cost of handling exceptions caused by account data issues o Assessed on returns with reason codes R02, R03, R04, and R20. ODFI fees for Authorization debit returns o Fees charged to ODFIs and passed through to RDFIs related to the cost of handling disputed or revoked authorization only for debit entries o Assessed on returns with reason codes R05, R07, R08, R10, R29, R38, R39, R50, R51, R52 and R53. ODFI Fees for Correction Entries (NOCs) o Fees charged to FIs that receive NOCs, passed through to the FI originating the NOC, related to the cost of correcting account data and avoiding a return. For each of the two return categories (Account Data and Authorization), (we) developed five alternative types of fees to assess: 1. A fee charged for all returns in the return category; 2. A fee charged only for returns involving select SEC codes; 3. A fee minus a discount, to allow for returns outside the control of the Originator; 4. A tiered fee, with no charge for an ODFI s returns below a target return rate, and a fee for all returns in excess of that rate; 5. A threshold fee, with no charge if an ODFI s returns are below a target return rate, but if the ODFI s return rate exceed the threshold, the fee applies to all of its returns (not just those that exceed the threshold). As a general principle, a quality incentive fee should be commensurate with the costs incurred by RDFIs due to exceptions. More expensive types of exceptions should be subject to higher return fees. The optimal fee will be different for each type of exception. Account data exceptions occur about once in 400 entries (i.e., approximately 0.25 percent), and much of the handling is automated. Authorization exceptions occur much less frequently, affecting about one in 2,000 entries (i.e., approximately 0.05 percent). They are also relatively costly for RDFIs to handle, because they usually involve a disputed transaction with a customer service interaction, and frequently a Written Statement of Unauthorized Debit. These factors call for a relatively higher fee for authorization returns. The incidence of NOCs falls in between these the two return categories, occurring once in 900 entries (11 basis points). NOCs require the RDFI receiving an error-containing