Ensuring liquidity at all times is the major task of any treasury and finance group. This not only requires an accurate and daily cash position over the whole group but also a cash flow forecast that shows future liquidity. It enables timely and appropriate funding decisions to be made.
The following criteria need discussing before a cash flow forecast is implemented:
Planning categories that make up your forecasting structure
The time buckets of your forecast.
Together, they ensure that your cash flow forecast fulfils
its purpose and is recognized as a meaningful tool.
There are, of course, a number of categories that are always the same, such as wages, insurances, tax and interest. The more challenging categories are usually those that concern your individual operative business.
Operative cash flows: two approaches
Here we usually see two types of approaches, the first high-level and the second very detailed. The first one is where you simply ask for external, internal and other operative inflows on a high-level. This works for some companies and is the best solution for them. However, before deciding on that approach make sure that the units involved in the planning can relate those categories easily to their business model. In some cases it might be more appropriate to break down those categories to a more detailed level.
Examples are companies where external inflows are planned individually for the five biggest customers and all other external inflows are planned as one sum. Companies that work on a project basis need to reflect each project in the forecast structure as such. For utilities, telecoms and online retailers it might make sense to break down the inflows by type of payment (direct debit, check, cash, credit cast) as this will influence cash flows.
For some companies the high-level approach will suffice; others will need the more detailed one. You need to find the proper balance to show each entity a different forecast structure and consolidate the numbers centrally.
How to handle internal cash flows
Another relevant issue is the handling of internal cash flows. Do you want to be able to reconcile internal cash flows between companies in your forecast? After all, they should be zero over the whole group. This requires specifying the internal counterparty while planning the internal numbers. It will lead to additional work and maybe involve heated discussions between the involved entities during the reconciliation process.
On the other hand, can you accept that a production and a sales entity differ in their opinion about future flows? It is a tough call to make and is strongly influenced by your business setup and culture. If most internal flows are between a strong central production entity to a larger number of smaller sales entities, you could automatically mirror the plan of the production entity to the sales entities and discuss afterwards. If the flows are between many different local entities dispersed globally, the workload in handling this might become almost unfeasible for the results gained. It is sometimes more advisable to establish the forecast without planning the internal categories per counterparty and to introduce this at a later date once the forecast is established and accepted.
The time horizon, frequency, and the time buckets of the forecast
The possibilities for choosing the time horizon, the frequency and the time buckets are virtually endless. You could go for the classic 12-month forecast on a monthly basis, or plan the first month in days and only the remaining year in months.
You could go for a 13-week forecast done once a week, or a quarterly one where the next quarter is planned in months, the rest in quarters and after that a yearly number and so on. Before deciding on the horizon, the frequency and the time buckets you will need to answer the following questions:
Which time horizon makes sense for my type of business?
If you are set in a short-paced business; for example a trading company for agricultural goods, you can probably only predict the next few weeks with acceptable accuracy. For retailers, utilities and telecoms, with a relatively linear business and maybe only seasonal variations, a monthly forecast over 12 months should be realistic. For companies undertaking project-based business (such as machine builders and construction companies) the time horizon might become completely irrelevant.
As each project is planned separately anyway, the time horizon is determined by the run time of the project. If your company needs to balance the utilization of credit lines carefully you will go for a shorter time horizon—maybe even on a daily basis. If you are set in a cash-rich business this becomes irrelevant. Look to your business model, consider what purpose you want the forecast to fulfil and the time horizon will became evident.
Which time buckets make sense for my business?
The time buckets are usually determined by the time horizon. For a 13-week time horizon weekly time buckets probably make sense; a 30-day forecast usually has daily time buckets. Things become more challenging when you move to longer time horizons that are not only planned in months, but also where the beginning of the time horizon should be planned in smaller buckets. This is possible but contains a number of pitfalls that need to be addressed.
For example, suppose you want to plan the first month in weeks and the remaining 11 months in months and do this once a month. How are you going to handle weeks that are partly in the first month and in the second month? This is a simple example, but highly likely to confuse your end-users and introduce unnecessary problems into the forecast. The best advice is to keep it as simple as possible. Once you deviate from regular time buckets such as days, weeks and months and start to introduce split-weeks, buckets of four weeks and similar you start confusing people. If two different time buckets are needed, it can even be better to introduce a short-term and a long-term forecast as they might consist of different planning categories anyway.
Which frequency can my end users handle?
To complete a high-quality cash flow forecast takes time. If
you ask for a daily or weekly forecast you need to allocate enough time to your
users to complete that forecast. Otherwise, you will trigger cut-and-paste
exercises that will lead to a forecast that cannot fulfil its purpose. Keep the
frequency realistic and only ask for daily and weekly updates if you have a
specific business reason and are able to process the results yourself.
Hubert Rappold is CEO of TIPCO Treasury & Technology, Vienna, Austria.