For any business, knowing how much money will be coming in and going out gives a better understanding of how much will be needed to keep operating and to enable growth. The optimisation and most effective use of cash, in essence, is what cash flow forecasting is all about.
Whilst the fundamental principles and purposes of cash flow forecasting remain largely unchanged, the needs of the modern market have placed greater emphasis on accuracy and cash visibility. The full 100% minimum liquidity coverage ratio of Basel III (to be enforced in 2015) will change the liquidity landscape as banks – including Asian banks – place different values on the deposits sitting on their balance sheets. As corporates revisit their investment policies and risk appetites in this environment, and as structures such as in-house banks, payments-on-behalf of, collections-on-behalf of and payments factories continue to emerge as favoured models, visibility over cash becomes ever more vital. The aim is to be able to segment that cash into the most appropriate operational, reserve and strategic liquidity pools and, particularly in Asia, be able to navigate the regulatory diversity that characterises regional markets.
In achieving this goal, one of the key requirements now is for treasurers to better understand the needs of their business colleagues, and for those colleagues in other business functions to understand the needs of treasury and the function of liquidity. Even corporates that, by the nature of their business or sector, have not been best placed to deliver accurate forecasts should now be looking for a clearer all-round view. For an organisation, being sufficiently agile to get the best from its financial data is important; technology can play a part, but open communication is essential.
The treasurer is typically charged with looking after the organisation’s cash, ensuring it has appropriate liquidity and mitigating any risks and changes in business flow that may arise. But often the information related to these tasks resides in other business units and all too often these functions sit in siloes and communication is minimal. The calls from treasury for yet more financial data may not always be seen in a positive light if there is little understanding of why it is needed. This really becomes a problem when the other functions fail to recognise that the information they provide has to be current and accurate; rough approximations will lead to inaccurate forecasts and a subsequent incorrect assessment of cash needs by treasury. This is a major risk for the entire company.
If business units are held accountable for the determination of their own cash and liquidity needs and feeding that back into the cash flow forecasting process, it will encourage understanding. It will be beneficial, in terms of the accountability of the business units, for those units to be measured and monitored using KPIs to ensure that the task is carried out in a timely and accurate fashion. It will also be helpful for the treasurer to clearly explain to the business unit heads that the information provided will be used not just for liquidity planning but also possibly for interest rate planning and FX hedging, and that getting it significantly wrong is a serious risk event. Indeed, if the forecast is significantly wrong because the information provided was inaccurate or because the modelling was poor, the treasurer may be hedging a position that does not exist, exposing the business to trading and open positions that are not backed up by real business flows. Sizeable and unexpected losses may ensue.
Cash flow forecasting in practice: TRILUX Group
International lighting solutions provider, TRILUX Group, has more than 20 production and sales sites in Europe and Asia (including China, Hong Kong, Philippines and India). Its Group Management division operates a multi-layer five-year integrated budget and liquidity forecasting plan which includes an annual view of the Group’s P&L, balance sheet and cash flow across all business units, covering some ten currencies in total. This, explains Oliver Thissen, Member of the Management Board, Finance/Legal, enables Group Management to draw up the budget and liquidity plans for the following financial year, which it does every September, the figures being based on refinement at multiple points in the year with reference to actuals.
To enable this, Thissen explains that the Group has a cash forecast and cash management plan running on a daily basis, using information generated by the import of SWIFT MT940s into its single SAP treasury management system (up to 90% of bank statement data from more than 150 accounts is loaded automatically, says Thissen). Each business unit is required to create a rolling four-month liquidity plan which is revised using actuals on a weekly basis. On a monthly basis the annual budget and liquidity plans can then be revised for each company within the Group. A comparison is made between monthly target and actual account statements with explanations sought for any variance. The revised results are fed back into the annual budget and liquidity plans as part of the continuing cycle of updating and improving.
The rolling data is in fact a series of five liquidity planning spreadsheets, notes Thissen. These cover the following data points: an overview of liquidity allocation; assumptions of tax (including VAT) and personnel expenses (such as social security payments); an estimation of financial status (external loans and deposits and those managed via TRILUX’s in-house bank) and a consideration of payment terms (such as the posting of sales revenue versus customer payment data); estimated and actual opening balances (each business unit has the opportunity to enter and explain any manual adjustments). The fifth spreadsheet is the output of all companies, uploaded into the TMS and consolidated into one view which is used to produce annual budget and liquidity forecasting plans.
Collection of figures starts with actual budget figures from the previous month, output from SAP; the rolling budget forecasts for the next month are then added. Using what Thissen refers to as an “Excel bridge”, data is transferred into the rolling forecast for cash flow. The entire process to this stage is carried out on a per company basis so the results uploaded into the TMS generate around 35 different cash flow forecasts within the system. With the help of TMS he explains that it is then possible then to consolidate all forecasts into a single view. This view is used to produce annual budget and liquidity forecasting plans – and the cycle continues.
The frequency of forecasting depends on industry sector and, to an extent, the size of the business. A treasurer will often run a forecast off the back of a cash positioning process, used to determine elements such as current available balances, cash concentration and funding. The period depends on need: short-term cash forecasting is for overnights and investments into the market, whereas the longer forecast ensures cash is in the right place at the right time and in the right currency.
As with any process of prediction, the further out you go, accuracy will diminish. This is why frequent updates using actuals is essential and why known future financial events, such as debt repayments, must be incorporated. In considering the latter, it should be apparent that forecasting not only recognises cash events over a certain time horizon but also non-cash events such as repayments, covenants, interest rate and FX risk exposures and even hedging policies.
Good quality information and a means of managing accuracy are vital elements of forecasting. This can be achieved with a ‘look-back’ process – sometimes known as variance analysis – to analyse and update previous forecasts using actual figures.
Attaining 100% accuracy in a forecast is unlikely, especially for the longer-term view. In fact, the Treasury Today 2014 Asia Pacific Corporate Treasury Benchmarking Study reveals that just 8% of treasurers in the region reach an accuracy level of more than 90% when comparing forecast versus actuals on volume of transactions; it slips to just 5% when comparing for value of transactions.
However, it is important to understand cause and effect to eke out incremental improvements in the managing, monitoring and provision of data. If it is understood why something happened it may be possible to prevent or replicate that scenario as required. The more the treasurer is in charge of what flows are going to look like, the better the forecast will be.
Methods and models
When generating a forecast, it is important to have access to an historical database of actuals that show how the business needs and uses cash over time. But data is just data until it is cross-referenced and interpreted; only then does it become usable information. By bringing it all together, it is possible to apply relatively simple statistical analysis to produce a forecast (it is also possible to apply highly complex mathematical models, explanations of which can be found elsewhere). Obviously it is important to consider whether the business has changed in any significant way (a major change of sales model or product line, for example) before using previous forecasts and actuals to influence results, but by looking at the database of cash actuals relative to projections, any variance will be apparent.
There are two basic types of variance analysis: ‘common cause variation’, or the ebbs and flows found in the normal course of business; and ‘special cause variation’ which reflects abnormal events likely to result in a negative or positive ‘spike’ in forecast performance. It should be possible in the analysis of ‘actual versus expectation’ to place any variance either in the realm of common or special cause. If the spike was ‘special cause’ and thus not predictable, it will be beneficial for future forecasting to try to put in place an early indicator of such events occurring again; learning from the past is a key part of forecasting.
The role of technology
According to a 2013 survey of APAC treasurers carried out by SunGard and Bank of America Merrill Lynch, 67% of respondents do not use any form of forecasting tool for cash and of those that do, 69% rely on spreadsheets. The survey also showed that of those respondents that did not use any cash flow forecasting tools, just 10% were ‘very satisfied’ with their processes, the vast majority (52%) claiming to be just ‘moderately satisfied’. The Treasury Today 2014 Asia Pacific Corporate Treasury Benchmarking Study further reveals that implementing a cash flow forecasting solution is the top technology issue and that cash flow forecasting is the number one area in need of improvement for those with an existing TMS.
It is clear that connected systems on a widely distributed network are the ideal when collating forecasting data. A TMS or ERP that is accessible via the web by the various business units, regardless of geography or time zone will afford direct input of their element of the cash flow forecasting data, this either being used as a continuous update or allowing the treasurer to collate, analyse and present the data relatively easily.
But business units that sit outside of an otherwise connected business can present data for manual entry, even if that means emailing a spreadsheet to the treasurer or picking up the phone. Despite the ‘shock, horror’ headlines of the APAC survey, a formally dedicated spreadsheet is a suitably flexible tool for the job and can be used with an ERP or TMS (see case study, opposite). However, as with all spreadsheets, care must be taken around version control, incorrect formulae and input error.
With the centralisation of treasury, the rise of in-house banking structures and the increased use of connected technologies and analytical tools, there is an opportunity to have better visibility over cash and cash flows. For those that choose to take advantage of these technologies, it means the ability to forecast is improving too. The models used by proprietary solutions may be far more complex than common spreadsheet formulae, but they do signify a reduction in reliance on the ‘best effort’ of an individual and a shift of emphasis towards more accurate and timely information and data support.
Linked with the ability to electronically move cash around the globe to optimise balances and make rapid transfers (using a pooling structure, for example), technology could play an increasingly important part in cash flow forecasting over the next few years. However, the value of significant relevant forecasting experience and an in-depth understanding of the company and the sector in which it operates should never be underestimated. It is thus advisable to consider proper ‘succession planning’, fully documenting all processes; if only one person has the knowledge, it will disappear if they leave the company.
As banks increasingly play a role around cash application, they are in a position to provide more data flows to clients related to the receipt of cash payments and the application of those payments to account balances. Some 25% of the APAC treasury survey respondents said improving cash visibility was top of their agenda.
But as bank products are commoditised, the banks themselves are seeking more of an advisory role to cement their relationships with their corporate clients. APAC is a complex region for treasurers not least because the diverse tax and regulatory environment makes cash optimisation challenging, especially where currency controls exist. If understanding the rules is difficult and cash visibility is an issue, in passing on their accumulated experience and understanding of integration, processes, systems and organisational structures, the larger banks with regional presence are in a position to genuinely add value to the cash forecasting process.
That said, banks continue to have a more traditional role to fulfil. To ensure an organisation has the right level of cash and liquidity it may set up either an uncommitted or a committed facility with a bank. If, on the basis of solid forecasting, the business is reasonably confident of its cash positions going forward it may feel able to meet its liabilities in a given period. In such a case it may not need to put in place or call upon its facilities at all. However, with poor quality forecasting or in an unpredictable or volatile financial environment, there may be a need to seek bank cooperation.
A committed facility will cost more than an uncommitted facility, but with an uncommitted facility, if the need for cash arises simultaneously with many other businesses, bank funding may not be available. Of course, financial derivatives exist to hedge against unfavourable movements but there is a cost here too, depending on the kind of programme used. Hedging using FX forwards is cheaper than using options and can help predictability by smoothing the curve of FX variability through the neutralising effect of any gains and losses (options are better suited for those seeking the upside whilst mitigating the downside, but they attract a premium). It is all a matter of confidence in the forecast.