Print this article | Return to Article | Return to CFO.com
A Nokia treasurer urges his colleagues to work with each other and their financial institutions to usher in cost-reducing standards. And from this corner, a defense of the banks.
David Blair and Nick Franck, CFO Asia
November 1, 2003
The treasurer's lot is not an easy one. More often than not our employers see us as practitioners of a non-core activity. If we're lucky enough to have a financially savvy management, our business may be considered necessary hygiene rather than a generic service that can be outsourced with all the others. In any case, in these challenging times we come under increasing pressure to do more with less.
In such circumstances, a casual observer might expect treasurers to be a radically innovative bunch, always trying to find new and smarter ways to work. Unfortunately, we seem to be deeply conservative, hiding behind the sanctity of money. This is ironic: money, being just a collection of (hopefully secure) bytes on a network, is highly conducive to technological reinvention.
At Nokia, the search for ever-improved processes in our effort to survive as a cost and technology leader pervades treasury as well as other operations. I often find myself sitting in meetings with my logistics colleagues. After they have explained how they can move physical parts around the world for pennies, it is deeply embarrassing — not to say career threatening — to try to explain why it costs me US$10 to move a few bytes of payment around.
Nor can I shirk my responsibility for this. My logistics colleagues did not wait for the logistic service providers to bring them ready-made solutions on a silver platter. They had to go out and find suitable partners and push them to think and act out of the box. And if I whine that money is highly regulated, they laugh me out of the room. These people have dealt with some of the most recalcitrant customs authorities on the planet. And they have succeeded — to the benefit of both their commercial and governmental counterparts.
Being just a simple treasurer, I am unlikely to succeed in wiping clean hundreds of years of banking legacy so that we can enjoy payment services at email prices. With some frustration, I have decided to accept the legacy clearing systems as entrenched and unchangeable. Consequently I focus my efforts on what I can change — our intra-and inter-corporate processes.
Back in May 2001, The Economist wrote about "a 14th-century banking system". Since then, various new approaches to clearing have been proposed and some have even gone live. Interestingly, the banks themselves seem the most enthusiastic about avoiding using the traditional clearing systems — witness interbank solutions such as Visa and MasterCard, CLS, and STEP2. (See "A Treasurer's Alphabet," at the end of this article, for explanation of the arcana of treasury system acronyms.) If the banks are so eager to avoid traditional clearing, surely that might be a sign that corporates should also be looking for cheaper alternatives?
Some industries like credit card service companies, airlines, and telecoms have developed vertical solutions for off clearing settlement. Visa and MasterCard offer horizontal solutions like VisaCommerce, which are genuine alternatives to traditional clearing systems. But most corporate efforts try to work around legacy clearing systems rather than to effect any fundamental change.
The RosettaNet Payment Milestone Program falls into this category. We are not trying to redesign any clearing systems, and in fact we have explicitly stayed away from the banking space per se (no new clearing, no new payment standards, no new bank processes). We are focusing only on the part we can change — our own corporate systems and processes.
The RosettaNet Payment Milestone Program sponsors have determined that we can save time, reduce costs, and improve customer satisfaction by automating accounts receivable reconciliation. The current manual reconciliation process is labor intensive and the resulting delays frustrate customers waiting to see their account clear. In close cooperation with banks and SWIFT, with solution providers, and with other standards bodies like TWIST, we have developed a process that automates accounts reconciliation using existing legacy banking systems.
Furthermore the sponsors have designed the process to work across any banking system with any flavor of corporate-to-corporate communication be it RosettaNet, other XML, EDI, or iDoc. It's completely agnostic when it comes to any banking system or standard. All ERP systems support automated posting so the process does not require reworking of internal processes within a corporation. We hope that by making the process totally generic, we can encourage adoption across industries and regions, and kickstart a virtuous cycle of network benefits for everyone.
For this to happen we need corporations first to start using the process and second, to pressure their banks into supporting it. While the process works with legacy banking systems, some banks may have to adjust the way they work to ensure that they follow best practice in the matter of preserving customer reference data through the banking systems.
The problem is that, over the past 50 years, many banks have got into the habit of overwriting customer reference data with their own internal reference data. To be fair, this was a reasonable compromise in the past when corporate customers did not seem to use the customer reference data capabilities of the banking systems.
Now that corporates want to start to use the customer reference data, we are facing decades of legacy behavior that has to be cleaned up. Whilst this is not about new systems or even material revisions, even a seemingly small clean-up in legacy banking systems has a cost. So the banks need a clear message from their corporate customers that we want this clean-up to be done.
A Dominated Market
OUnfortunately corporate treasurers have a very poor record in getting their wishes heard as a customer group by their principal service providers, the banks.
A case in point is the evolution of online FX trading. In the late 1990s, a couple of technology companies saw in the FX markets a huge volume of homogenous high-value transactions being executed in a very inefficient and non-transparent way. They saw an opportunity to create technical solutions by moving the front- and back-office processes onto the internet. The prime example is Currenex.
The global FX market is dominated by banks, which provide a valuable service in providing liquidity but also profit handsomely from their position as prime intermediaries (or from a hugely asymmetric market). As Currenex floated onto their radar screens they panicked and sunk reportedly hundreds of millions of dollars into two rival platforms Atriax and FxAll, each backed by separate and competing groups of banks representing some 30 percent of the market.
Banks have a history of investing in technology supposedly to benefit their customers (often without asking customers what they really want). Most of these are not profitable and end up getting sold to technology companies. In early 2002, Atriax duly folded. Commendably, two of its three shareholders — JPMorgan and Deutsche — saw the light and began supporting both Currenex and FxAll.
Many banks still unilaterally support FxAll only. They seem to be trying to squeeze liquidity out of Currenex and to kill it off. Why? No doubt because of a lot of invested money and ego, but mainly because it is nice (for the banks) to be trading on a platform designed, owned, and run by banks.
Just how nice? Consider the moment of closing a deal itself. On Currenex, when a customer clicks to deal, both sides are locked in. In FxAll, when a customer clicks to deal, the customer is in fact passing to the bank a free option to deal at that price. The bank may choose to refuse to deal! Given this situation, treasurers can be forgiven a rueful smile when FxAll banks boast — as they sometimes do — about how few deals they refuse. In Currenex they would not have the luxury of deciding whether they deign to deal with their customer.
Another example is the presentation of quotes. In Currenex, quotes are ordered in real-time with the best at the top of the list. (A first-year computer science student would be able to figure out that this is what a customer would want.) In FxAll, quotes are listed in historical order in a blatant attempt to make selection harder for the customer. In Currenex, a customer can deal with an unlimited number of banks. There are good reasons to limit the number of banks per deal, but it is left to the customer to decide that. In FxAll, there is a fixed limit of five banks.
Of course with a large number of powerful banks exclusively supporting FxAll with a view to killing off Currenex, it is not unreasonable to expect liquidity to accrue to FxAll, and that many treasurers prefer FxAll because of the issue of liquidity. But what if we decided what was best for us from a technical and structural perspective — and then insist that our banks trade on the platform of our choice? Isn't chasing the liquidity putting the cart before the horse? The liquidity will surely go where the customers are.
Some treasurers are pushing their banks to join Currenex. And the fact that Currenex continues to thrive in the face of such powerful opposition is testament to its attractiveness to discriminating customers.
My point is that treasurers hold a stronger position to influence their service providers' direction than they have exploited so far. And that we need to use this power more effectively. It would be a sad state of affairs if, fast-forwarding a few years, treasurers' only recourse was to complain at conferences about how the banks have cornered us with a bank-owned monopoly trading platform.
Embracing automated accounts receivable reconciliation as recommended by the RosettaNet Payment Milestone Program is one way we can avoid this depressing scenario. The Milestone Program's process design is generic and allows implementation based on very minimal criteria. The bottom-line requirement of the banks is that they preserve customer reference data through their systems.
This does not seem like a very demanding thing to ask. As for corporate treasurers' skittishness about this proposal, so far I have been unable to imagine how encouraging global banks into a commitment to preserve 16 characters of customer reference data through their systems would harm any corporation. Corporate treasurers should rally around this idea and let their banks know clearly that they want this to happen.
David Blair is Nokia's corporate treasurer for Asia. The views expressed here are his own and not those of Nokia or RosettaNet.
From This Corner — A Defense of the Banks
Let's summarize David's article: banks are reluctant to assist corporate customers in supporting industry standards, a fact that creates problems for treasurers. But does that mean the banks are engaged in a cartel against corporates? Of course not! Bankers have very good reasons to be unenthusiastic. These standards create problems that most treasurers have not even tried to address.
Consider David's own example of the logistics industry. True, technological improvements and economies of scale have allowed logistics providers to "move physical parts around the world for pennies". But what good has it done for the providers themselves? According to a report by Mckinsey Quarterly ("Growing Pains for Logistics Outsourcers," 2003, Number 2): "Even as customers ask more of their [logistics providers], they are beating down prices through canny contract negotiations that shrink profit margins and make returns on invested capital inadequate."
Common systems in the internet-based foreign exchange market result in a similar bind for bankers. One bank treasury sales manager told me recently that the spread on major currencies was already so narrow that it barely made a difference. The banker added: "In the second tier currencies it has taken two to three pips [price interest points, or the smallest increment that a particular currency pair can move] off our margins." For several second tier currencies a spread of two to three pips was already the only real difference between the buying and selling price.
Consequence: there is little profit to be made in dealing small amounts any more, only large amounts. And banks trade far bigger amounts, far more frequently with each other than with corporates. Therefore, there is now even less advantage in dealing with corporate clients. Now, treasurers may well say, "So what? If the banks want the rest of our [more lucrative] business, they should provide us with these services at cost or as loss leaders." The problem is, they're not small, these loss leaders. In fact, they're very expensive. In an overbanked environment, bank margins are being squeezed. One consequence is that banks are being forced to consolidate worldwide. Another consequence — more relevant to this debate — is that capital expense budgets are limited.
Now take the external environment. The worlds' central banks have issued new regulations with the admirable intention of reducing the risk of one bank's failure leading to multiple bank failures through a domino effect. These regulations, however, mean that it's no longer possible for tens or hundreds of thousands of payments to be reduced down to just a few through netting. This alone makes the cost of settling payments very high. Combined with the fact that many of the world's bigger banks had legacy, batch-processing systems (i.e., not real-time enough to process the payments) there is an ongoing major investment to be made just to comply with these new regulations. This is happening domestically (real time gross settlement systems), cross-border (continuous linked settlement) and with securities such as equities and bonds (delivery vs. payment systems).
Result: For the foreign exchange market — including the internet based foreign exchange trading platforms mentioned above — the cost of transacting has risen dramatically, but not the trading margins paid by the customers. It's the same for the payment transacting banks. Electronic or paper-based payment prices have — mainly — come down for large customers, not risen. And there is no sign of that trend reversing itself either.
The same central bankers have issued instructions for banks to manage their internal risks better, to know their customers better, and the list goes on. More capital expenditure, and still no better margins.
So what is the real disposable budget the transaction banks have? Not much. Bankers have to make a decent return on the remainder to make these changes affordable — and make a profit for their shareholders.
In view of this, let's look at RosettaNet. A request for industry-wide standards, on the face of it, looks reasonable. It would please many customers. So why not? The short answer is that the return on investment may not be there for the banks (and could even amount to a negative return).
The electronics industry, from which the RosettaNet sponsors come, is a fairly straightforward one. IT is mainly US-dollar-based and payments between major companies or between them and their distributors/agents are mainly electronic payments and suitable for transmission with additional electronic data. Therefore their requirements should be reasonably easy to satisfy.
But now take a look at the oil industry. It's served by the same banks and has an identical structure as well — US-dollar based with large value electronic payments. It would clearly also benefit from the RosettaNet standards. The difference is it also has a lot of cash, paper and credit card-based payment methods. Now what do most treasurers want? Simplicity. "If you can do this for my electronic payments, why can't you do it for my paper and cash payments as well?" More development necessary, more cost. "Oh, and we'd like you to handle this in all locations we have in every country, in the capital city, and in all of the small locations where we have customers or suppliers, in every currency. You can't do it yourself? Persuade your partner banks in-country!" Banks have to serve many industries using the same product sets.The difficulties and the costs mount up, and the margins are no better.
An example springs to mind. In the past (and to this day) certain key firms overwhelmingly dominated the German banking market. With their dominant position they were not going to accept a bank-specific payments platform. They therefore created a bank-independent one, Multicash. This allowed the transmission of payments to any bank at the flick of a switch—which is exactly what some corporates did. Having negotiated prices with their banks on the basis of total forecast volumes of payments, some of these companies (including one large chemicals company in particular) "punished" banks at a moment's notice by switching payments away from them. Result: as industry "standards" came in, down went the income of the banks (and up went its volatility).
Bankers are not stupid. Although they know that their cash-management products are (or will end up) being judged on price and price only, they can be forgiven for not wanting to accelerate the process, and for not wanting to end up in the logistics industry's current position either (if they are not already there.) They want to make a fair return on investment just as much as any other company or line of business, be it in the electronics, oil or chemical industry.
But treasurers are often downright unhelpful when it comes to explaining themselves to banks. Not one Association of Corporate Treasurers in Asia (with the exception of Australia) allows bankers to be members, thereby depriving them of a forum where the needs of the corporates can be listened to without having to pay (and therefore having to make a return on attendance.) No group of treasurers has explained to the banks how the commoditization of their business will be offset by newer, less variable business. To be fair to the RosettaNet participants though, they have indeed spoken to and explained in great detail what the benefits of standardization are for the banks and other suppliers. But their challenge now is to convince them that most corporates — once the barriers to entry and exit have been reduced by standardization — will not jump from one supplier to another at a moment's notice.
Until that kind of trust is built, industry-wide standards will keep having long and painful births. It happened with EDIFACT, BOLERO, and other common standards between banks and corporates. There is a danger of this happening to RosettaNet.
Work with your banks, don't pressure them.
Nick Franck is the founder and director of CFO Solutions, a consultancy to banks and corporate customers on treasury management. His opinions are entirely his own and do not represent those of any particular financial institution.
Better with Rosetta?
RosettaNet is a self-funded consortium set up by technology and telecommunications companies to create and put in place open standards that would cover e-business transactions. The goal of this non-profit venture is to align processes between supply chain partners around the world. The lingua franca of RosettaNet is its Partner Integration Processes (PIPs), a uniform code that applies to design, product information, volume planning, ordering, shipping and logistics, invoicing, and settlement.
In late 2002, several RosettaNet members saw the need to tie the remittance advice (PIP3C6) with the credit advice to allow for automated reconciliation of accounts receivable. To make this happen, a number of companies — Cisco, Intel, National Semiconductor, Nokia, and Texas Instruments — decided to sponsor a Payment Milestone Program within RosettaNet.
The sponsors recognized that reconciling accounts receivable by hand costs time and money and causes customer dissatisfaction. An April 2002 report from Killen and Associates, a US-based consultancy, estimates that working capital inefficiencies can cost US$27 million per annum for a typical company with sales of US$1 billion.
The sponsor companies worked intensively with their partner banks, which included ABN, Bank of America, Citibank, Deutsche, HSBC, JPMorgan, Nordea, Standard Chartered, and with SWIFT, TWIST, and IFX to develop a globally applicable standard. Solution providers such as FormFill, IBM, MicroSoft, Tibco, and WebMethods also joined the collaborative effort, which culminated in a grueling week of design work hosted by SWIFT in Brussels at the end of June 2003.
The design developed by the Payment Milestone Program allows the remittance advice to travel separately from the money flow, so as to work with legacy banking systems. The only requirement made on banks and clearing systems is that they preserve a 16 character Unique Remittance Identifier (URI) through the banking system. According to studies by the Payment Milestone Program, this is feasible in most clearing systems.
A Treasurer's Alphabet
CLS stands for Continuous Linked Settlement. This is a new settlement system for foreign exchange designed to reduce "Herstadt Risk" (the risk that one weak bank could trigger a domino effect of liquidity crises throughout the banking system) but which also minimizes the number and value of cashflows passing through bank clearing systems.
EDI stands for Electronic Document Interchange. This is a set of legacy standards for communicating business messages usually over private networks. iDoc is SAP's proprietary messaging standard. It can be considered functionally equivalent to XML and EDI.
IFX stands for Interactive Financial eXchange. IFX is a gathering of business and technology professionals dedicated to developing a robust XML framework for the electronic business-to-business exchange of data among financial service institutions around the world.
PIP3C6 is the RosettaNet remittance advice message. RNIF stands for RosettaNet Internet Framework. This is the standard transport protocol for sending XML messages between parties over RosettaNet.
STEP2 is the new euro low-value, pan-European clearing system that essentially operates as a netting system between banks to minimize the number and value of cashflows passing through the older bank clearing systems.
SWIFT is the Society for Worldwide Interbank Financial Telecommunications. SWIFT sets standards and runs a secure network for interbank cashflow messaging.
TWIST is the Treasury Workstation Integration Standards Team. TWIST started out working on standards for financial trading in foreign exchange and money markets, and has since expanded into standards for commercial payments.
XML stands for eXtensible Markup Language. This is the de facto standard for machine-to-machine communication across the internet, and is deployed by RosettaNet.