Cash flow forecasting is the bedrock of treasury. Treasury is fundamentally forward looking, and basically concerned with future cash flows which generate liquidity, foreign exchange, interest rate, credit and operational risks.
Without cash flow forecasts, treasury cannot operate. Without cash flow forecasts, the risks are unknown and risk management becomes impossible. Worse, practicing risk management without forecasts risks at best inflated costs and at worst added losses. For example, entering into foreign exchange forward contracts without properly understanding the underlying foreign exchange exposures risks increasing rather than reducing foreign exchange risk.
What, how, when
As with many business processes – only more so – cash flow forecasting processes must address what, how and when. What are we trying to achieve? How will we succeed? When should we do it (with what frequency)?
The what element requires clarity about the purpose for the cash flow forecast. The how element relates to the process design and the tools used. The when element is concerned primarily with frequency.
What means objectives
Simplifying, most businesses use three distinct cash flow forecasts:
- FX exposure forecast serves to capture FX exposures (however defined) across the group, so that they can be effectively hedged (however defined). This is often for a rolling 12 month period by legal entity by currency.
- Rolling 12 month cash flow forecast serves to determine subsidiary and group level funding needs, so that these can be met cost and tax effectively. This is typically a rolling 12 month period by legal entity in functional currency.
- Cash positioning serves to ensure that individual bank accounts have just enough balance to cover upcoming payments and to minimise idle balances. This is typically daily for five to 30 days by bank account and currency.
(For simplicity, I am ignoring the long-term or strategic forecast often for three to five years at group level.)
The FX exposure forecast (1 above) could logically be a subset of the rolling 12 month cash flow forecast (2 above). Many businesses plan and manage in functional currency, which means that the FX exposure forecast has to be a separate exercise.
Cash flow forecasts can also serve for reporting and control purposes, but the above are the key intrinsic treasury functional purposes which relate to capital structure and funding, risk management, and short-term cash management respectively.
In the latter instance, cash positioning can allow businesses to progress from ‘just-in-case’ cash balances to ‘just-in-time’ balances. This can produce material reductions in idle cash.
A key issue with granularity is avoiding excessive demands on reporting entities. Too many treasuries seem to espouse the notion that more detail is better. They seem to ignore how much work this means for colleagues across the organisation.
In general, the guiding principle for granularity must be accountability. Granularity should be sufficient to enable treasury to identify whom to contact in case of anomalous variations – but not more. A simple example is to split internal vs external flows, since these are often handled by different teams. Another is to split direct and indirect procurement because often they are managed by different teams.
Another critical factor is what actuals are available. Cash flow forecasts more granular that what is available from actual (historical) reporting are basically unverifiable.
When means frequency
Frequency of cash flow forecasting also has to be modulated according to the utility of the information. Again it is a question of applying minimum work to generate maximum actionable information.
For example, in cash positioning, the 20% of bank accounts that generate 80% of flows should maybe report daily or twice a week. For most accounts, weekly forecasting should be sufficient. For less active accounts, monthly may be sufficient.
Most of the other two forecasts are most commonly monthly. In some businesses more regular updates might make sense.
Many businesses specify a materiality threshold for ad hoc updates ie when material new information arrives within the organisation. Ad hoc inputs are disruptive to the planning cycle but enable closer management of balances and risks. What constitutes material will vary between different businesses.
Although many geeks will want to use the latest gee whizz technology, in most cases technology choices should probably be predicated on more prosaic criteria.
Since cash flow forecasting in its various guises often requires input from teams across the business, not just treasury, the starting point should be what is already widely used at least within the finance community of the business.
To many this will mean Excel. However Excel, for all its flexibility and widespread use, is not a platform for operationally critical processes. Excel is wonderful for experimentation and exploration, but it lacks the sustainability and resilience of ‘proper’ systems based on distributed databases with robust access control and availability. More importantly, Excel solutions often rely on extensive copy pasting which is a minefield for unnoticed errors. Excel’s very flexibility is its Achilles heel because there is no data validation and no error correction.
For treasury, TMS will seem the logical choice but not all teams outside of treasury will be familiar with TMS, and using a system once a month is a recipe for frustration. ERPs – where usable for forecasting – are generally well supported at least in the finance function. BI platforms are similarly broadly used, and may thus be a good choice.
In general, the technology platform for cash flow forecasting should probably be based more on user accessibility than technical merit.
AI – or more prosaically machine learning (ML) – is being used for forecasting with some success and much fanfare by some businesses. Digging deeper almost always reveals that ML is only trusted with extensive human sanity checking.
In current times when COVID-19 has disrupted whatever was normal before the pandemic even for businesses who have weathered the crisis well, the limits of ML have been exposed. ML is based on extrapolating from historical data patterns. To the extent that historical data patterns have been invalidated by the pandemic, ML has limited utility – ML is basically a pattern matching game, when the patterns are no longer valid, ML’s magic evaporates. In true disruptions, wetware (aka humans) are the best predictors.
In times of great uncertainty, best practice is to produce forecasts for different scenarios – normally optimistic, expected and pessimistic aka minimax scenarios.
Scenarios help businesses prepare for different possible outcomes, which is critical in times of disruption. Scenarios can help elucidate the true value of real options.
McKinsey recently wrote a pertinent piece on scenario based cash planning.
Sustainability and resilience
Although more and more treasuries aim to contribute to growth and profitability, ultimately treasury is a risk management function whose role is to enhance the sustainability and resilience of the business.
Management sets strategy and tactics, and treasury ensures the business survives to deliver, by providing adequate but not excessive risk capital at optimal cost to support the business.
Cash flow forecasting in its various forms is critical to this role of treasury. Without good forecasts, treasury has to provide ‘just-in-case’ capital rather than ‘just-in-time’ capital, which is inefficient and costly.
Another aspect of resilience is to ensure that treasury can continue to support the business adequately even when processes such as cash flow forecasting are disrupted. BCDR (business continuity and disaster recovery) is a critical aspect of cash flow forecasting, since it is critical to the functioning of the business.
Cash flow forecasting has always been foundational to treasury, because treasury cannot manage risk without forward looking information, however imperfect it may be. The disruption brought by COVID-19 has made this priority even more clear. Pragmatic and functional approaches are essential to ensure treasury enables the business to thrive in troubled times without draining internal resources.