Earlier this year, Treasury & Riskreported on the findings of a Deloitte survey which revealedthat corporate treasurers are playing an increasingly strategicrole in their organizations.

The same survey also provided some interesting insights into thecurrent state of companies' treasury systems infrastructure. So wesat down with Niklas Bergentoft, a Deloitte director who heads upDeloitte's treasury technology practice, and Kesavan Thuppil, aDeloitte director who heads up the firm's SAP treasury practice, todiscuss.

Survey respondents were treasury professionals in more than 100of the world's top companies. Forty-seven percent came from theUnited States; 45 percent from Europe, the Middle East, or Africa(EMEA); 4 percent from Latin America; and 4 percent from theAsia-Pacific (APAC) region. Forty-five percent work in companieswith more than US$10 billion in annual revenue, while 55 percentwork in companies with less than $10 billion in revenue.

T&R: What were your keytake-aways from the survey?

Niklas Bergentoft: I see two keytake-aways from this survey. First, more treasurers today are ableto get sufficient budgets to execute projects in the right way.They're approaching these projects as transformations, and they'respending a sufficient amount of money to get them done right.That's happening more frequently than just a few years ago.

And second, they tend to select systems based on the best fit.They're looking at what technology would support their business andprocesses in the best way; that's much more important than the costaround it. Most companies have one or two main treasury technologyplatforms—either a treasury module of an ERP or a treasurymanagement system—but there's still a need for specialized softwareto support peripheral processes such as exposure gathering,integrated forecasting, trade execution, and bank connectivity.

T&R: If companies are now viewingtreasury system implementations as transformation projects, arethey taking a different approach to decisions around which treasurysolutions they are going to deploy?

Kesavan Thuppil: Yes, definitely. Onething I've been observing, this year in particular, is that ourclients are doing a lot of benchmarking, and they're connectingwith a lot of their peers in the industry, trying to get othercorporate treasurers' opinions and perspectives to find out what'sworking, what's not working, and what challenges they shouldanticipate. They're going the extra mile in their decision-makingprocess.

Another trend is that companies have become a lot moreopen-minded in terms of which products they choose. A couple ofyears back, if a company was an SAP customer, they would start froma default position of exploring the SAP Treasury module to see ifit would work. The attitude was 'Why not SAP?' But this year andlast year, quite a few companies I've worked with want to make surethey're being comprehensive in their decisions. They willdefinitely consider SAP Treasury, but they also want to explore theother applications that are out there.

T&R: What's driving thischange?

NB: I think it has to do with the factthat treasurers have been gaining a stronger position within theirorganizations over the past few years. When the approach was 'Whynot SAP?' we tended to see the IT function driving thedecision-making process, and treasury would have to go through thatexercise first. As the treasurer's role has been elevated, thetreasury function has become better able to drive thedecision-making process based on their needs. And they have thestrength to say, 'We're going to pick what's the best fit for ourneeds, whether it's the treasury module of our ERP system oranother treasury management system.'

KT: That's a very good point. From IT'sperspective, system rationalization is a significant objective, sowhen any group in the organization is choosing a software packagethat could be handled by a module in the ERP system, IT tends towant to leverage that main platform which they're already using andare familiar with. But given the changing responsibilities of thetreasury organization, the treasurer is increasingly able to standup and say, 'This is my portfolio; I should make the callhere.'

I was recently dealing with a large consumer productscompany, which is using SAP as its global ERP. They purchasedlicenses for SAP Treasury and were preparing to implement it when anew treasurer was hired into the organization. He was coming froman altogether different industry where he had been using a treasurymanagement system, and he was not sold on SAP Treasury. When hequestioned whether treasury should continue moving forward with theERP module, the first response from IT was that the decision hadalready been made. But the treasurer said, 'No, I am the treasurer.I call the shots now.' So the company re-evaluated its options tomake sure it was selecting the technology that would be the bestfit for the treasury organization under the new treasurer'sleadership. In the end, SAP was able to meet the business case andwas implemented after all.

T&R: One data point from yoursurvey that I found interesting, if not alarming, was that 40percent of respondents said one of their key challenges is aninadequate treasury system infrastructure. Similarly, 40 percentalso said that visibility into global operations, cash, andfinancial risks is a key challenge. Why are so many treasury teamsrelying on problematic technologies?

KT: The main reason is that any systemthat was implemented eight or 10 years ago is unlikely to meet thecompany's needs today as well as it did back then. That isn'tnecessarily a result of poor design or poor implementation. Mostbusinesses have undergone major changes over time, includingmergers, acquisitions, and expansion. They may be operating indifferent geographic regions and different currencies. All of thesechanges can lead to complexities in consolidation and may alsochange the needs of the treasury team, as well as theirexpectations around the technology that supports them. As such,what was implemented 10 years back is just not compatible to thetreasury business as it exists today. So companies are forced toreconsider their options in the marketplace.

NB: Treasury has changed,and systems have evolved too. So if you haven't kept your system upwith the evolution, it's going to be suboptimal today. It is alsotrue that the purpose of treasury technology has changed at manycompanies. In the past, these projects were more about automating,whereas now they're more about transforming and optimizing thetreasury function. A lot of organizations, especially organizationsbetween $2 billion and $20 billion in annual revenue, are currentlyout there looking to replace a treasury systems infrastructure thatthey implemented five or more years ago. And they didn't take thesame approach during the implementation back then—thetransformative view which requires really tailoring the system toyour processes and your processes to the system for the bestpossible set.

T&R: In your survey, 64 percentof respondents said their company's treasury management systemconnects to more than one ERP system. That seems like a goodexample of a suboptimal infrastructure.

NB: Yes, it's a challenge. Most largeorganizations run on more than one instance of their ERP. Onecompany I've seen used 52 ERP systems. I think that's the worstI've seen. That impacts treasury in terms of exposure gathering,balance sheet forecasts, cash flow forecasts, and bankconnectivity. To solve for this challenge, companies areimplementing a standard middleware solution that extracts data fromthe ERPs, transforms the data and imports it into other systemsusing one standard. It's hard for treasury to effect change aroundthis, but the treasurer can partner with controllership and IT tosponsor an ERP centralization effort.

KT: Yes, this is primarily a problem invery large global corporations. Some have gone through aconsolidation effort to reduce the number of ERP systems, but othercompanies are very decentralized in their global operations andtreasury has to integrate with many different ERP instances. Andwhen treasury has to source the data for exposures, for cashforecasting, and for their accounting from different ERP instances,it's very challenging to get data that is consistent and reliable.It's easy to establish interfaces or establish a process whereinpeople send you data on a monthly basis. But that doesn't mean thedata quality is such that you should be relying on this data fordecision-making purposes.

T&R: Is this usually a problemfor companies that have engaged in a lot of M&Atransactions?

NB: To a large extent. An acquisitivecompany that faces pressure to integrate acquired entities in ashort time frame sometimes has to move forward withouttransitioning the acquired company to the corporate ERP. It's alsointentional strategy in some companies. They may want regionalinstances of the ERP system, or different instances for otherparticular purposes. For companies with a smaller ERP system, itmay just be that the technology isn't scalable enough to supportthe whole company on one instance.

T&R: Are there any other reasonsyou see that would explain why such a large percentage [40 percent]of treasury functions believe their treasury systems areinadequate?

NB: I think the most importantreason is that treasury and IT departments alike tend to overlooksupport for the infrastructure they have implemented. These systemswill deteriorate pretty quickly if you don't support them right.The treasury management system requires regular updates to evolvewith changes in business needs. These systems also have interfacesto other financial applications, such as the ERP system, as well asto peripheral systems like the trading portal and investmentportal. Maintenance of this infrastructure requires dedicatedsupport.

Many companies fail to dedicate people in treasury and IT tomonitor business-as-usual activity and to support system andinfrastructure maintenance. They may not have people in eitherfunction who are knowledgeable enough to understand the fullpicture and know how to keep the infrastructure performingoptimally. So over time, these systems stop functioning as well asthey did in the beginning.

T&R: What's thesolution?

NB: Most of the problems tie back toeither not having dedicated people or not having enough trainingfor the people who are there. And that's really a direct effect ofnot identifying who's going to support the systems early enough inthe implementation project. Because the bigger the system, thelonger it takes to learn. Some companies tend to create roles thatare focused only on system support, and those roles may not attracttheir top talent. That can translate into a suboptimal skill setsupporting these systems.

I would suggest companies should be creativeabout how they're defining the system-support roles. They cancreate a position where system support is one piece of the job, butthe job also offers career development opportunities to attract theright people. A solution may be to create a group that'sresponsible for supporting and evolving your treasury managementsystems, but you combine that with strategic initiatives—looking atwhat's the next improvement treasury should undertake. Systemsupport needs are going to ebb and flow, and during downtime fromsystem support, these individuals could be looking at continuousbusiness improvement for treasury, driving the treasury functiontoward leading practices from a process and technology perspective.That would make system support a more interesting role and attractpeople interested in driving change to the organization. If youalso included this position in a job-rotation program withintreasury, then you would position it as an important element inyour overall talent management strategy.

T&R: Your survey also revealedthat a significant proportion of respondents use homegrown systemsin each of the functionality areas you asked about: cash managementand treasury accounting, bank administration and relationshipmanagement, investments and debt management, FX and interest raterisk management, and commodity price risk management. (See Figure 1on page 4.) Does this gibe with what you see among yourclients? Are a lot of companies operating on systems builtin-house?

KT: Yes, but understand that what peopletend to call 'homegrown systems,' for the most part, areExcel-on-steroids types of applications. Typically a homegrowntreasury system started as an Excel spreadsheet, then down the roadit was connected to a database, and maybe they've set up some basicautomated queries to import data and run reports in Excel. I see alot of systems like this. And many of them were developed 20 yearsago, when the type of treasury management systems we have todaywere just not available.

In a lot of cases, the systems are still working reasonablywell, but the maintenance of a homegrown system tends to be veryintense. These systems also create a lot of dependency on internalexperts who can maintain them.

T&R: You would say thatmaintenance requirements are usually higher for a homegrown systemthan for packaged software, in terms of keeping it functioningproperly and connected properly?

NB: Yes, that tends to be true. And, asKesavan pointed out, you have 'key man' risk. I have one clientright now that is working on a divestiture, and they need toreplicate their homegrown treasury system for the divested company.Both the parent and the divested company have staff who know how touse the system day to day, but nobody knows how it works on theback end. The person who developed the system has retired, sothey're stuck.

T&R: So, when you work with aclient that is shopping for a new treasury management solution,what does the selection process look like?

KT: It depends to a large degree on thematurity level of the treasury department, the type of operatingmodel they have, the size of their team, and the volume of activitythey have around all the different treasury functionalities. Cashmanagement is obviously the backbone. They need to know at anypoint in time whether they have enough liquidity to pay theirvendors, and whether they have excess cash to invest out in themarketplace.

Beyond cash management, companies are often looking to havetheir treasury infrastructure automate areas such as bankrelationship management; investment and debt management; and FX,interest rate, and commodity risk management. Because cashmanagement is unavoidable, they start with that but then oftenstretch their system requirements to include other treasuryfunctions.

T&R: What would you advisecompanies to make their key considerations be in selection andimplementation of treasury systems?

NB: First, they need to make sure theirvision for the project is really clear at the outset. And by'vision,' I mean: What are treasury's objectives, and how do thoseserve the organization? This vision needs to be aligned with thefinance leadership, starting with the CFO.

When you've nailed down treasury's objectives, you can startlooking at how you will achieve those objectives in yourorganization. What will your operating model look like when theproject is finished; how do you want the treasury function toevolve from both a process and a technology support perspective?And what skill sets will you need, in your staff or in externalexpertise, to meet those objectives? Then make sure you spend asufficient amount of time evaluating the technology. You don't haveto run a slow selection process, but you need to spend enough timeto validate the capabilities you're looking for.

You also need to take a very structured approach toimplementation. Do not take shortcuts as you're going through thesystem design; the design needs to precisely reflect yourrequirements. Do not take shortcuts, either, in testing and invalidating that the solution will work well and will meet yourneeds. And it's important to be open to tweaking your processes sothat they best fit the system you select. This is not aboutcompromising on how you want to run your business, but it's aboutmaking sure the treasury function works very well without having todo much customizing of the selected technology. If you use thesystems in a way they were not intended to be used, they're notgoing to work well for you.

Finally, as I mentioned earlier, make sure that everyone who'sgoing to use the system is adequately trained, not only on thetechnology, but also on your new processes and on their new roles,if applicable. Likewise, provide adequate training for the supportorganization. Sometimes evolving the support team requires harddecisions, but don't shy away from making sure the right people arein place to ensure that the system continues functioning properlywell into the future.

KT: Over the past few years, I've hadaround a dozen clients that have taken this journey. And I'd say 80percent of them underestimated the effort they needed to put intoevolving skill sets, both in terms of support of the technology andin terms of how the treasury function would operate afterimplementation. Investing in training and hiring is critical for acompany to operate efficiently and to achieve its objectives in atreasury transformation project.