Survey Highlights
continued

Cycle Times
The 2009 survey examined the cycle times of seven critical treasury functions: cash flow forecasting, concentrating/pooling cash to establish daily cash position, producing a treasury accounting entry, resolving bank account discrepancies, processing an internal fund transfer, processing a borrowing decision and processing investment elections.

One finding was that for most of these functions, size, as measured by revenue, is not a significant predictor of cycle time.

"But probably the most surprising and instructive outcome was that, with just a handful of exceptions, the responses suggested that automation, in and of itself, does not improve cycle times," Roth says.

The survey compared cycle times for automated treasury organizations with those of manual treasury organizations. For functions such as developing a short-term cash flow forecast, producing a treasury accounting entry and concentrating/pooling cash, the responses showed no difference in cycle times.

The only significant difference in cycle times experienced by the typical automated and manual organizations was in the number of bank accounts they were able to reconcile per cash-managing FTEs (14.5 for automated organizations compared to 9.7 for manual ones).

Organizational Structure
Organizations can structure their treasury operations in a variety of ways. Some choose to decentralize so that each subsidiary or even each location has its own operation. Others consolidate treasury operations into a single location, while in other cases certain operations are outsourced to a third party.

However, according to responses to the AFP Treasury Benchmarking Program survey, the majority of organizations conduct most treasury operations within a single corporate treasury operation. Respondents were asked about how they structure more than a dozen treasury functions, and the percentage of respondents reporting a centralized delivery approach ranged from 58% for "produce treasury accounting entries" to 90% for "manage debt."

Automation
Systems represent about 9% of treasury operating costs, according to the survey. Dollars spent on treasury systems is another area revealing a benchmark gap. On average, the typical organization is spending 4.07 cents per $1,000 in annual revenue, while the benchmark organization is spending 1.29 cents per $1,000 in revenue.

And what are treasury organizations getting for their technology dollars? Not everything they expect, they report. "There were very few spots where respondents' automation expectations were met," Roth says.

While automated organizations reported better outcomes in audit trails, improved payment efficiency, and greater effectiveness for in-house banking and bank fee analysis activities, they largely indicated dissatisfaction with automation's impact on reducing staff and improving cycle times, for instance.

"The most widely cited reasons for automating treasury were reducing staffing and manual errors, improving cash flow forecasting and treasury reporting, consolidating to a single application, and rapidly establishing a cash position," Glenzer says. "And these were some of respondents' biggest areas of disappointment."

When respondents whose treasury organizations rely primarily on manual processes were asked why they had not automated, their top reason was that the cost of automation exceeds the perceived benefit.

"One of the messages of the 2009 survey seems to be: If you think you are going to solve that benchmarking gap by throwing automation at it, or purely by centralizing your operation, these factors alone are not going to drive those performance improvements," Glenzer says.

European Results
The 2009 survey's more robust respondent base included about 9% European respondents.

Although international participation was significantly lower than US participation, across the survey responses it was clear that region was not a significant predictor. In other words, the survey results should be just as relevant to European treasury operations as they are to US treasury operations, the researchers say.

Looking Ahead: 2010 Survey
The 2009 survey results reaffirmed that there are significant gaps between typical and benchmark treasury management performance. As to the reasons behind those gaps, the survey data would appear to eliminate at least two possibilities, namely that the gaps are driven solely by treasury organizations' use of automation and/or their treasury structure decisions.

But if not automation and structure, then what's driving these gaps?

Answering that question will be one of the goals for the 2010 survey, Glenzer says. "We'll be looking at what drives those huge performance gaps in areas such as treasury costs and FTEs between typical treasury organizations and benchmark organizations," Glenzer says. "We'll try to identify factors that are driving up efficiency and driving down costs in treasury while allowing organizations to achieve world-class outputs."

If you would like to be part of the effort to identify these drivers, keep in mind that the next AFP/IBM/Deutsche Bank benchmarking survey will be conducted beginning in early March 2010 and will be distributed by e-mail to AFP members and treasury professionals in the gtnews subscriber database. To ensure that you will have an opportunity to participate and receive a customized benchmark report — which will enable your organization to compare its performance against the overall survey sample and against top performers — contact your Deutsche Bank relationship manager or Global Transaction Banking representative.

Additionally, please let your representative know if there are any areas of treasury and cash management that you would like to see addressed more fully in the 2010 survey.

Previous page