Duplicate Payment Detection Paying Off For Bank of the West

Historically a centerpiece of basic bank services, safe and secure payment processing is a critical element of the banking industry’s future.

By Flavia C. Muttera, Bank of the West
Bank Systems & Technology
August 24, 2012

Historically a centerpiece of basic bank services, safe and secure payment processing is a critical element of the banking industry’s future. Fortunately, transformational innovation in digital payment forms, including the last decade’s tectonic shift from check to image processing and sustained growth in native electronic payments (ACH), has worked in banks’ favor.

Banks’ reputation for being good stewards of payments has positioned them to benefit from the advent of new options in electronic and mobile payments — a market advantage worth preserving. Consumers and commercial enterprises trust their financial institutions with their payments — trust that will continue unless eroded through financial institutions’ operational error or fraud — or unless banks move too slowly in offering new payment options, siphoning customers’ payment loyalty to alternative providers.

Banks have discovered that one of the leading risks involved with processing higher volumes of all types of electronic payments has been the unintended consequence of duplicates and their close cousin, false positives. Banks and banking experts agree that duplicate payments create a significant operational impediment and that weeding them out is in banks’ best interests.

One Bank’s Story

A top 25 US bank with more than $60 billion in assets and more than 700 locations in 19 states, Bank of the West will process more than 6 million inbound checks and more than 10 million outbound checks per month in 2012. Bank of the West’s payment operation is 100% front-counter teller image capture, image incoming, image check collection, and image return item processing. Currently, Bank of the West offers merchant RDC, its image ATM is in process, and it is working on a mobile image offering.

Bank of the West made a commitment to moving toward an image-based environment following the enactment of the “Check 21” legislation in October 2004 and had completed the conversion to 100% image processing by July 2010. The bank realized that duplicates had a cost, and that duplicate volumes were increasing.

When Bank of the West began moving to image-based processing, the number of duplicate checks presented ranged between five to seven duplicate items per million payments processed. By the time the bank’s image conversion was complete in 2010, the incidence of duplicates had risen to 100 duplicate items per million payments processed.

As the rate of duplicates increased, the bank began searching for a more robust duplicate detection solution. Three key factors drove the solution selection:

  1. The capability to quickly identify and correct for duplicates prior to posting any entries to customer accounts
  2. The capability to identify duplicates across multiple channels, including check image branch deposits, merchant remote deposit capture, check image cash letters, image ATM, image returns, mobile channels, and ACH
  3. Based on automated business rules both at the file and item levels and had to allow for flexibility in determining how far back to search for duplicates based on the item type

Solving the Duplicates Conundrum

Based on these key requirements, Bank of the West selected a Day 1, multi-channel, duplicate detection platform developed by CONIX Systems, Inc. The Dupe Detective platform it selected enables duplicate identification and review occurring prior to posting, utilizing a common database for all payments to ensure duplicates are identified across payment sources and channels.

Bank of the West reviews:

  • check image deposits,
  • check image cash letters,
  • merchant RDC,
  • merchant image cash letters,
  • image ATM,
  • image loan payments/lockbox,
  • ACH – including ARC, BOC, RCK, XCK codes, and
  • image returns.

This system filters false duplicates based on automated business rules at the file and item level, then links to its outgoing returns system to avoid suspecting re-cleared items.

True duplicates identified by Bank of the West consistently equal 0.01 percent of total prime capture volume, a figure that mirrors the industry. However, this seemingly small number does not equate to a small problem, particularly when you consider that undetected duplicates can cost between $100 and $ 125 to identify and correct.

Bank of the West now encounters fewer than 2,000 dupes per month. Applying the conservative posted-duplicate resolution cost mentioned above, Bank of the West is avoiding a significant expense.

Since a percentage of duplicate payments are fraudulent in nature, duplicate detection can be considered an element in any bank’s fraud detection efforts. According to industry sources, duplicate check fraud costs banks $500 million annually, or half the estimated $1 billion in total check fraud in the U.S.

The balancing act for banks is between minimizing “suspected dupes” and maximizing identification of “true dupes,” since the same business rules that avoid suspecting drafts consequently leave the door open to accepting true duplicates. Duplicate detection software “suspects” items based on matching four elements: routing, account, serial and dollar amount. If a serial number is not read or if a dollar amount is altered, then a true duplicate will not be presented as a suspect for review.

Our recommendation is that banks fine-tune their search history for duplicates over a historical time frame that will allow them to maintain deadlines and quick throughput.

The value of a Day 1 detection solution is that is enables us to pull duplicates out of the process and prevent them from posting. Day 1 interception is the key to managing duplicates.

Teachable Moment

Banks that implement automated duplicate detection strategies in their payment operations are experiencing compelling benefits. And the timing could not be better. Where automated duplicate detection is in place, banks are reducing costs, improving customer experience, and protecting account security. Some have also discovered a connection between the Day 1 identification of payment duplicates and identifying fraud.

Further, there’s evidence that duplicates are on the increase. Our vendor’ findings show an increase from 5 to 7 duplicate items per million payments processed in 2006 to between 40 to 100 duplicate items per million payments processed in 2010, a more than 1100 percent increase.

According to McKinsey and Company analyst Aaron Caraher, it is too early to tell whether the number of duplicate suspects and true duplicates will increase as a result of the adoption of mobile check capture by the broader consumer market.

McKinsey and Company’s 2011 Check Benchmark Study cites duplicate “suspects” as one of the major sources of interruption to check image “straight-through” processing. There is general agreement that duplicates will most likely continue to increase with the push of RDC to broader consumer markets. This is driving banks to invest in sophisticated and expensive systems to provide near real-time views of deposit and clearing activity across channels.

Although duplicate detection software can detect duplicate items within the same bank, no software is available to detect duplicates across banks. There is growing conversation about the need for a shared image archive with the ability to provide duplicate processing across participating banks. In this scenario, the third-party check image exchange vendor might provide duplicate-detection “on-we” services for all of its bank clients.

It is worthy of note that all but one participating bank in the 2011 Check Benchmark Study reports reviewing all items for duplicates, whereas the one outlier only reviews on-us items. The practice of limiting duplicate review to on-us items leads to a dilemma. Should banks take responsibility for “cleansing” their entire payments stream of duplicates as part of a cooperative financial services community? It is a dilemma of both ethics and practicality to selectively allow potential duplicate and fraud items in outgoing items where they can continue to gum up the works and amass costs unnecessarily.

No doubt remains as to the harm and cost of duplicate items staying in the payment stream longer than Day 1. Perhaps the time is right to consider the ethics and practicality of reaching for a universal Day 1 duplicate detection strategy.

Flavia C. Muttera is vice president, Transaction and Call Center Services (TACCS) manager, Bank of the West.