More and more organisations are looking into investing in a forecasting system to support the forecasting process. The credit crunch has put more impetus behind this idea that had been on many lips already. The technology is constantly improving and is now at such a level that it can meet most corporates' requirements. A controlled purchase process is very important, in order to ensure that you don't end up with a solution that cannot meet your needs in the long run. Remember that improving any business process with a new system is as much - or even more - about improving the work process as it is about choosing a system.

Basic Needs

First, one must evaluate the corporate's needs and how they may change over time. When the forecasting system is first investigated and implemented, the needs might be simple and straightforward. But these requirements will grow over time with the organisation's awareness of the availability of forecasting information. Therefore, it is important for the system to be flexible and for its output to be adaptable to a growing need for forecast information.

Stakeholders

Many corporates make the mistake of seeing the forecasting system as a treasurer's project. The treasurer is certainly a key stakeholder but surely not the only one. To get a full picture of what a forecasting system can achieve and to foresee future needs, it is vital to get other stakeholders and interested parties involved. Figure 1 shows an example of stakeholders for forecasting and liquidity information.

Figure 1: Stakeholders in Forecasting and Liquidity Information

The Forecasting Process

Improving the process with system support is related more to the business process than to the choice of system. Therefore, it is vital to outline how you see a best-practice forecasting process. This process description should determine the requirements for the system's functionality when you are preparing your requirement specification.

It is important that the system supports the forecasting process in terms of how information can be input and how this input process can be followed by the treasurer or another person responsible for the forecasting process. Figure 2 illustrates one option of how a forecasting process could look. Each forecast - short- and long-term, for example - will need its own process.

Figure 2: Cash Forecast Process

System Input

The most important element for the cash forecast is input data. Without good input, you can never get a forecast with good output data. Input data is usually partly manmade, as it is difficult for a system to fully automate the forecast; however, the input can be simplified via integration. Therefore, it is important that data can be checked and adjusted by the appropriate personnel with key knowledge. The extent to which data can be fetched through integration differs between companies. For example, if most of the sales are cash-based, there is not much benefit in integration with an accounts receivable (AR) system, but you might be able to fetch data from a sales system. If it is not possible to fetch relevant data from a source system, the forecasting system might support different calculation rules used in forecasts, such as moving average and seasonal variation.

In the input of forecast data, there are some key problems to take into account. One of these concerns internal transactions. If two subsidiaries have accounts within the same cash pools but forecast different cash flows, the treasurer will see irregularity in the forecast for the pool that will not be realised. Here a netting system could be the solution. Some forecasting systems allow counter-transactions to be created according to a netting principle, meaning that only one party registers the cash flows. This rule might also be applied to data fetched from AR/accounts payable (AP) systems, with internal transactions only fetched from AR or AP.

Figure 3: System Input

Adaptability and Ease of Maintenance

As mentioned at the outset, it is important for the system to be flexible in meeting new requirements. This includes ability to adapt the system to new forecasting routines and adding of additional forecasts. For instance, a company might start by making only a short range forecast but later wish to add a separate long forecast. Another requirement might be additions at more detailed levels, such as calculation formulae in current forecasts. Also, easy addition and removal of accounts is essential in the system.

For companies performing acquisitions, it is important that the administrator be able to add new companies and remove those that have been sold. In this situation, it is usually crucial to quickly obtain high-quality information on liquidity in the acquired company.

Reports and System Output

The requirements of the various stakeholders will, of course, differ in terms of both the detail of information and the timeframe of the information. Therefore, it is vital that the system be flexible in its output of information. Requirements concerning who is able to access the information will also differ; therefore, how user rights can be set up is important.

By flexibility in reports, we do not refer only to flexibility for naming different type of cash flows according to your specific needs; it is vital for one to be able to set up the forecasts in multiple ways with the same data, simply looking at things from a different perspective. That is, you should be able to filter information and choose what data to have on row or column level, and thereby build reports according to your needs.

It is also important to be able to create different types of summaries as your needs dictate, otherwise you might need to export the data from the forecasting system to Excel and do manipulations there. In terms of access, export information can be vital for management presentations etc, but for the daily work, exports would be time-consuming and complex. Therefore, it is vital to be able to view the information needed in the short term directly in the system without any amendments or additions, including calculations. Four main restrictions in terms of flexibility can sometimes be found if the system uses fixed reports:

1. A report structure might not be found that matches the needs. For example, it might be possible to see cash flows company by company, but not business area by business area.
2. Data might be visible only in fixed periods. For instance, a day-by-day forecast might not be convertible to week view.
3. It might not be possible to recalculate currency reports to use other currency units.
4. Freely definable summary calculations separate from the report structure may be missing - for example, with no way to summarise operational cash flows in a report when the structure of the report does not show operational cash flows separately.

Figure 4 illustrates how a forecasting system may use a 'reporting builder' (i.e. OLAP cube) to apply different dimensions for transactions to set up reports flexibly.

Figure 4: Forecasting System as Reporting Builder

Alternative Systems

There are different types of systems to investigate when buying a forecasting system. First, there are different types of vendors and also different types of techniques are available, as covered briefly in a previous article 'It's Liquidity Management, Not Cash Forecasting [www.gtnews.com/article/7364.cfm].

Treasury management systems (TMSs) vary greatly in how they work and what functionality they contain, but normally they focus more on the need of the treasurer and financial risk management than on needs related to the cash manager and the operational cash flow. This limits the ability of the TMS to handle the cash forecast in an efficient way. TMSs focus on real cash flows rather than forecast cash flows, which, of course, is a problem in adapting the system for cash forecasts. But TMSs are improving more and more in their support for the forecasting process and we will most certainly see continual improvement in this area.

Enterprise resource planning (ERP) systems do sometimes support cash forecasting. Often the main problem is that the company uses many different ERP systems. When companies are looking into moving to one common ERP system, this can be a sensible option. However, the flexibility is not that great. If a company needs a special report, these normally need to be built as bespoke solutions. Unlike TMS solutions, ERP systems are not designed to handle the financial cash flows. This limits the system's ability to provide the full picture in a cash forecast.

Specialised cash forecasting systems are built to give the full picture of all forecast cash flows - both the operational and the financial cash flows. Cash forecast systems vary in how standardised they are and thereby how much you can tailor the system to your needs. Abilities to automatically fetch data from different data sources, such as bank, TMS, and ERP systems, also vary.

Evaluation Process

First, you need to find potential suppliers that you think may be able to meet your requirements. When a short list of possible suppliers has been created, many companies write a request for proposal (RFP) to uncover the details, finding out which supplier meets the requirements best and the cost for each system. This evaluation process is vital when one is buying an IT system. It is important for the treasury function to stay in the driver's seat during this process and not let this be run as an IT project. As mentioned above, the business process should form the basis for the final choice of system.

Implementation Process

When the choice of vendor and system is determined, the implementation process should begin. Naturally, to gain the full benefit from the system, you need to commence an implementation project with relevant stakeholders. How well the implementation process is run will directly influence the final benefit of the system. Therefore, the implementation needs to be well prepared and executed.

The implementation process might look like the example below:

• Specifications: getting all details settled with the vendor.
• How the forecast process should be set up.
• Cash flow models.
• Company structure.
• User information and user rights.
• Bank information.
• Type of forecasts needed.
• Calculations in forecasts.
• Other elements. 

Planning: The timetable for the project and planning of internal and external resources needed to implement the system.

Delivery: System implementation, including installation and tailoring of the system.

Pilot: Running the system with pilot entities, and making updates on the basis of input from the pilot.

Rollout: Starting the forecasting in the new system.

Summary

Buying a forecasting system is a task that should not be underestimated, but with the right people involved and a carefully planned process for implementing the forecast system and its support, the money will most certainly be well invested, especially in these days of the 'credit crunch' with its shortage of liquid resources. One of many important elements to consider in making the investment is to ensure that forecast information can be developed within the system, as demands for liquidity information usually increase when high-quality information is available. Flexibility of the system is a core aspect of enabling this. Also for companies expecting one or several acquisitions or other organisational changes, flexibility is a key word. To ensure that the investment and implementation process meet the company's requirements in the best possible way, it is vital to see this not as a treasurers' project but a process involving relevant stakeholders while the treasurer or cash manager is in the driver's seat.

Fler artiklar