Accurate forecasting is not easy. Here are four measures that can help treasurers get closer to the truth.
As any treasurer knows, the more accurate the cash forecast, the better their business can effectively manage assets and confidently pursue business objectives over both the short and long term. But as treasurers are also aware, achieving that accuracy can be easier said than done.
Here are four strategies, described by Jo Stevens, Product Owner at GTreasury, that corporates may wish to practice to generate more accurate cash forecasting.
Develop a holistic understanding of your enterprise’s cash flows, and how they evolve
Gaining a comprehensive grasp of cash flows requires varying degrees of nuance and effort that largely depend on the size and complexity of your business. As an example, any corporation that has grown through acquisitions and is active within multiple global marketplaces and industries will almost certainly have a complex array of payment terms.
At the same time, dispersed departments may make sizeable and ill-timed expenditures – such as paying a tax bill – without providing corporates much of a heads up. Accurate forecasting also gets significantly more difficult when businesses have high transaction volumes or particularly large value transactions (both variables that continue to increase across the board). Cash flows can also shift rapidly or run up against other challenging barriers.
“Given this growing complexity, it’s critical that all variables and swings in cash flow data are keenly understood to ensure forecasts accurately reflect reality,” says Stevens. “Managers who are responsible for providing data as a forecast source should have knowledge as to how this data will be consumed. Shortcutting will hamper precision and risk business decision-making that depends on those forecasts.”
Collect the right data for your forecasting model
Next, companies need to be sure to use all sources to assemble the requisite cash flow data for an accurate forecasting model. Accounting and finance departments are pivotal providers of this information. Other sources include capital and operating budget and historical cash data. “It’s often the case that established data systems – such as bank feeds, AR, ERP, and other solutions – are already active within the organisation,” Stevens notes. “Transaction data from these systems can be collected and leveraged manually by corporates, or through automated TMS solutions.”
However, she cautions treasurers to watch out for scenarios that can wreak havoc on forecast accuracy. Open communication with providers of the data and stakeholders ensures that the model applied best suits the current needs. “Corporates will want to be especially attuned to any large-value, low-frequency transactions, such as tax payments, capital expenses, and investments that reach maturity,” she suggests.
Indeed, whilst these types of transactions are often reported only when they occur, she believes that the far better strategy for corporates is to gain the insights necessary to recognise and anticipate these events well ahead of time “in order to produce forecasts that are that much closer to the mark”.
Set up a smart data workflow process to support your forecast
“You now need to build processes that will procure the data rapidly and consistently,” says Stevens. Organisational complexity will be the biggest process roadblock many corporates will now need to overcome. “To start, work with department leaders to win commitments and build conduits for the data required to start testing and modeling your forecast,” she advises. “Conversations with these leaders and subject matter experts should also help you glean a better understanding of any unrecognised transaction variables that should be included in the forecasting process.”
Specifically, she advises that developing an optimal data workflow process calls for setting up working data feeds that provide “unmistakable clarity around when data is entered or updated”. It also means that corporates should model some data manually, using spreadsheets to dive into the numbers across different time periods in order to get close to the data and discover deeper insights. “What’s more, corporates will also want to investigate any major anomalous transactions to make sure they are well understood, and predictable, going forward.”
Choose the optimal forecasting method for the task at hand – and then refine it.
Different cash flow categories call for different forecasting logic. Leveraging the appropriate methods will deliver superior accuracy. For example, the cash flow to be forecasted could be:
Operating cash flows – Cash flows for routine, everyday business, such as receipts and disbursements.
Financing cash flows – Capital contributions, long-term notes, and other financial transactions.
Investments – At most organisations, forecasts of investment flows are at a higher-grade importance than operating or financing cash flows.
Corporates should then hone the forecasting model down to the key attributes of the transactions in question, Stevens recommends. This means recognising whether they are, for example, one-time or recurring transactions, variable or consistent, seasonal, dynamic or in-line with predictable trends. It’s then ideal to build, and back-test, models to determine how accurately they perform. Lastly, she says, corporates should iterate for improvements, or switch to more appropriate and effective cash forecasting models to achieve greater accuracy.
“By achieving more comprehensive and insightful visibility into transaction data, and applying it within the correct predictive model, corporates can provide their business with more accurate cash flow forecasting – equipping themselves to make more confident, data- and foresight-driven decisions.”
