f the cost of processing the "next transaction" is in the $.005 to $.012 range (see Part II) and if interchange costs are equal for all merchant acquirers (see Part I), then why do merchant processing costs vary so much from one ISO to another?

To answer this, we need to revisit the subject of transaction-based costs. We have not yet discussed the fixed costs of running a processing center. We also need to discuss the fixed and variable monthly costs of customer service and back-end processing. To round out the discussion, we need to factor in things like capital costs of interest, depreciation and amortization, and income taxes and profits.

Depending on which set of figures you use, there are approximately 4 million U.S. merchant locations processing about 28 billion transactions at the point-of-sale. If true, this means the average merchant location processes 7,000 transactions per year.

The aggregate U.S. bankcard volume processed for 2002 will be about $1.2 trillion for an average ticket of about $39 and an average process volume of $275,000 per location annually.

But we all know averages can be misleading. Sixty percent of the transactions processed in the country, according to Marc Abbey's article in the January 2003 issue of Card Management magazine, are processed by the top 230 merchants. If we assume that the top 230 merchants accept cards in an average of 1,000 locations, that means the rest of the merchant locations process an average of a little less than 3,000 transactions per year.

Let's try to make an educated guess about what the largest merchants are paying for transaction processing. What, for example, do you think Wal-Mart is paying for merchant processing?

I know for a fact that one of the smallest transaction platforms recently bid 3 cents over interchange for one of the largest merchants in the country and that that bid was the highest bid out of 14. Thirteen platforms bid less than 3 cents. I am told by a usually reliable source that Wal-Mart pays less than $.003 per transaction. Most of the largest merchants in the country, I am told, pay less than one penny a transaction plus interchange.

Although interesting, this isn't really relevant to ISOs because few of us do the kind of volume that Wal-Mart processes, and each of our merchant locations needs and receives service levels above and beyond what Wal-Mart asks from its merchant acquirer at the store level.

Another important factor is that ISO volumes are spread across hundreds of different kinds of terminal and software platforms deployed in thousands of different kinds of environments using multiple communication protocols and lots of different people calling in asking all sorts of questions.

But if ISOs can't get Wal-Mart's pricing from their processing partner, how do they get the best deal? Not surprisingly, the deal the ISO gets boils down to the negotiating power of the ISO. The power to get the best deal is based on existing and projected volumes of transactions and service levels that are acceptable to the ISO.

As I indicated in Part II, most ISOs are paying in the range of 5 to 12 cents per transaction. In addition, I believe that most ISOs are paying another $2 to $5 per month per merchant location for settlement, back-end and call center services plus another per-call fee of $1 to $2.50 per incoming customer service call. These fees often are bundled into the transaction fee. For back-end processing services there are a few extra variable costs of sending a merchant statement (postage, envelope and paper) and there are minuscule variable costs of storing transactions and masterfile data and granting Internet access. But all of these combined are less than $.50 per merchant location per month. We previously discussed telecom costs, which also apply to the variable cost of taking service calls for 800 numbers.

How does all of this get reconciled to the transaction costs for the "next transaction" of $.005 to $.012? Part of the answer is the additional cost of providing banking services. All non-bank acquirers must process through a bank and therefore pay their sponsor bank for their services.

These services include accepting the primary level of risk of the portfolio with Visa and MasterCard, sending and collecting merchant monies via ACH, and often providing funding to cover the cost of the cash advanced for interchange each day while the ISO provides monthly or weekly billing of discount fees to the merchants.

What do these services cost? The last time I checked, the highest variable cost of an ACH transaction to a bank was $.007 and the lowest variable cost was $.00. (Yes, the Federal Reserve also gives banks compensating balance credits against their fees!)

If prime rate is in the 4% range, then it costs the bank another $.00275 or so to cover the prime rate interest expense for ISOs who collect discounts monthly. Of course, daily discounting and weekly discounting cost considerably less to the bank because the advances for interchange funds are repaid either daily or weekly.

Of course, ISOs like to offload the risk whenever possible, and the banks must provide for the cost of their losses as well. Losses in our industry run as low as .00005% (half of a basis point) of process volume to more than 1%.

My understanding is that the large ISO banks average losses in the 2-3 basis point range. Of course, a couple of ISO banks have failed recently because of their ISO business, so their costs must have been a lot higher - but that's another story.

The bank has to mark up all of these costs and often charges a per-transaction fee, sometimes called a BIN fee or sponsorship fee. For alliance banks such as Chase or Wells Fargo or Key or Wachovia, etc., this is often part of the structure of the alliance arrangement, and these costs are covered out of the alliance bank's share of the spread. For others, bank transaction fees range from $.001 to $.02 depending on the number of transactions and whether the fee includes the MasterCard INET fees billed weekly to the bank (see Part II).

The key driver of costs (no surprise) is the allocation of the fixed costs to the number of transactions processed on a specific platform. It costs millions of dollars to operate a processing center, and the variable transaction-processing costs are loaded with costs and expenses such as equipment purchases, software development and maintenance costs, disaster site operations, personnel, facilities and utilities.

It isn't difficult math to divide the fixed costs by the number of transactions. But don't forget to factor in an EBITDA (Earnings Before Interest, Taxes, Depreciation and Amortization) margin as well. A large shop is going to need 40% of its gross margin to drop to the EBITDA line to cover capital costs and income taxes and earn an acceptable bottom line.

As an example, if it costs $1 million per month to operate a processing center and the price of the transaction is marked up to 5 cents over the variable costs, then the center needs to process 33,333,333 transactions per month to generate a 40% EBITDA. At 10 cents, the number of transactions is obviously cut in half.

One can make the argument that it doesn't have to cost a $1 million per month to operate a processing center, and that is true. But do you think there are very many existing platforms available to ISOs that cost much less than this? I think the answer is no for any platform that an ISO could use.

Almost every ISO seems to need a little bit of this and a little bit of that. There just isn't product focus on much of anything from most ISOs except to process a transaction however the merchant wants to send it.

This hodge-podge of service offerings demands the continual tweaking and maintenance of hardware and ongoing development and maintenance of an expensive array of software platforms and packages. Some ISOs talk a good game about being niche players, but there are few truly niche processing packages in use by the merchants most ISOs solicit.

The products of the typical ISO are the products developed by the processors for other organizations. In other words, most ISOs are selling the same processing products being sold by their ISO brethren. There are very few ISOs focused enough in today's world to deliver a consistent, standardized product to a broad market niche. This drives the fixed costs of software development, call center management and product development through the roof.

Also, every ISO wants up times of 99.99+%, and that requires a redundant processing center located in a different geography with different hardware, software licenses, communication lines and VAPs and MIPs (direct Visa and MasterCard links). This all adds to the fixed costs of the processing platform.

In future articles I will address the role of processing scale, the available ISO platforms and talk more about the benefits of starting a new processing platform.

Finally, I want to report an important development in the Visa-FDC litigation. On Dec. 22, 2002, Visa filed a Motion to Dismiss FDC's countersuit. This motion is scheduled to be heard on March 12, 2003 in San Francisco. (For some reason, the new legal filings are not available electronically and must be fetched in hard copy from the courthouse.)

Visa's motion argues that the VisaNet system, which First Data Net is attempting to circumvent (along with the related fees and 8.4 basis points of assessments - see Part II), is an internal system of Visa and states (page 16) that "First Data is attempting to use the antitrust laws to force Visa to create 'competition' by outsourcing its own central processing function simply because First Data has created the ability to do so. ... Of course, that would be contrary to the antitrust principle that 'no firm has a general duty to injure itself in order to benefit a rival.' "

To this industry observer, Visa makes strong arguments against FDC's attempt to jumpstart First Data Net. It seems to me that Visa has a good chance of getting the countersuit dismissed. If the schedule holds, March 12 is going to be a big day for all acquirers and ISOs. If Visa does not prevail on the dismissal of FDC's countersuit, then a call to action by the acquiring and ISO community will be mandatory, in my view.

Visa is doing a lot of heavy lifting right now to preserve its brand and financial integrity. First Data is working hard to gain a competitive advantage that it alone could have as both a large issuing processor and a large acquiring processor. While these titans fight it out, Visa is coincidentally fighting for the preservation of the level playing field all acquirers and ISOs require to survive.

This brazen attempt by FDC to push aside Visa is the kind of thing that can happen when you outsource your core business, as the vast majority of Visa members have done with their acquiring and issuing processing. The chickens have come home to roost.

While I am optimistic that eventually this litigation will get resolved favorably for the rest of us, there are other lessons to learn from this case. Look to this series of articles for ongoing monitoring of this litigation as well as more commentary on these topics.

Bob Carr is the Founder, CEO and Chairman of Heartland Payment Systems, the nation's largest privately owned merchant acquirer and ninth largest overall, with annual revenues exceeding $300,000,000. Heartland was recognized by INC Magazine as the 57th fastest-growing private company in America and is one of the 10 largest INC 500 companies. Bob was a Founder and Vice President from 1988 to '90 of the Bankcard Services Association, which has since become the ETA.

Before entering the bankcard industry in 1986, he developed computer software systems for unattended fuel pumps and created the first integrated accounting applications for PCs. He also started the computer department at the Bank of Illinois and served as the Director of the Computer Center and as a mathematics instructor for Parkland College. He earned degrees in mathematics and computer science from the University of Illinois in 1966 and 1967.

To learn more about Heartland, visit www.hpsteammates.com or www.heartlandpaymentsystems.com, or e-mail Bob at Bob.Carr@e-hps.com