Un article de fond sur la vision anglo saxonne du "Cash forcasting" avec un interview de Xavier AUDIBERT, Directeur Produit chez Kyriba et de Benoist MULSANT, Responsable du Cash Management chez PSA PEUGEOT CITROEN.
1Executive Summary
·MNCs have no confidence in their own cash flows.
·Lack of influence over forecasting still apparent amongst treasurers.
·Cash flow forecasting identified as the top priority for treasurers in 2005 - but no real progress made at many companies over the last two years.
·Spreadsheets still dominate even in blue chip companies.
·Systems integration still a problem.
·A business case still to bemade for cash flow forecasting as many companies seem unaware of the average 30bp (basis points) improvement on investment.
·Companies are still in the "first phase" of forecasting with many of them wanting more accurate numbers for today, tomorrow and the quarter.
·Companies have yet to use the information held within their own systems to its full potential.
2gtnews predicts
·As systems integration becomes better, cash flow forecasting will become more accurate and reliable. It will be a natural by-product.
·It is likely to be a problem area for several years to come
·There will always be companies who do not operate treasury on a centralised basis and who do not cash flow forecast on any more than the basic level.
·Benchmarking is underlining those companies who are forecasting accurately - and unlocking extra return on their cash for the group. There will be an increasing gap between those who forecast well and those who do not.
·There will be more development of forecasting systems that can cope with companies which have multiple ERP systems.
·A larger involvement of banks in supporting corporate cash forecasting - perhaps the development of bank-based forecasting tools.
·There will be a continued trend towards companies integrating their payment factories, ebilling processes or shared services centres with their cash forecasting to continue.
·There will be increased take up of Internet applications for cash forecasting information collection.
·US banking "Regulation Q" will drive demand for cash forecasting in America - as banks currently earn no interest on deposit accounts.
3Introduction
Forewarned is forearmed they say. While no one can see into the future making forecasts is perhaps the next best thing - certainly in the world of treasury. Cash flow forecasting has been an area of hot interest for companies and their treasurers for several years - but in 2005 little progress has been made as the majority of companies still use disparate and manual spreadsheet systems to compile them. Companies are usually dissatisfied with the level of forecast accuracy and as a consequence are not making full use of the amount of money they hold. Bad cash flow forecasting is making companies less efficient than their peers.
On a global basis this topic is top of treasurers' agendas for 2005 because of the desire for improved returns on investment and the increased need to have more transparency, thanks to regulations such as Sarbanes Oxley.
And why wouldn't a company want to have better, more reliable information that allows them to invest or borrow funds at better rates, and have a direct impact on the bottom line?
Of course treasury departments the world over are under-staffed and overworked but there is a sense that perhaps the business case for cash flow forecasting accuracy has still to be demonstrated to some senior level management. With REL Consultancy estimating that there is over €580bn1 worth of excess cash waiting to be freed up within the top 1000 European companies alone, it is surprising that less progress has been made.
But there is no doubt that the interest in this area is high - banks, consultancies and software vendors have seen a rise in customers coming to them for help. Countless surveys have found it to be at the top of treasurers' priorities over the last two years - but systems integration and consolidation data are throwing up real and practical problems to implementation. Even if the will and resources are available - cash forecasting can often be overshadowed by larger treasury management implementation projects. But 2005 could be the year to see progress.
Also driving increased interest in cash flow forecasting are new regulations such as Sarbanes Oxley. These demand more control over business processes - and a better demonstration of why a certain course of action was taken. Cash flow forecasting has an important role to play in this and the trail of data for it is a natural reason for demands for better forecast accuracy.
As John Nicholas, at HSBC, explains: "The reason companies are looking at improving cash forecasting is because regulations are pushing to remove spreadsheets from the treasury function - auditors are no longer happy to sign off financial audits that are spreadsheet based. There needs to be a lot more robustness around the cash flow forecasting process."
Can Supportive Banking Structures Help your Cash Flow? - John Nicholas, HSBC Global Payments and Cash Management
Mike Gallanis is a consultant with US-based treasury consultancy Treasury Strategies. He believes that there are really two reasons driving treasurers' interest in cash flow forecasting: regulation and promised returns. "The first (reason) is the need of an organization to have a good handle on their cash flows, from a control perspective, with the introduction of Sarbanes-Oxley. With the focus on maintaining controls, forecasts can really serve as a very effective tool to really give people an indication of what cash levels should look like. It can also serve as a mechanism to detect any variations that might be inappropriate. The tightened control environment has had some impact on improved forecasting pursuits."
"The second thing driving the need for accuracy in forecasts, probably more so than any other issue, is the material impact of accurate forecasting. Companies are interested in capturing increasing interest returns in a rising interest rate market place. As it becomes possible to invest funds at higher rates, there is a much greater potential for reward," he continues.
Bruce Lynn of the Financial Executive Consulting Group agrees that cash forecasting is high on companies' agendas. "If there was ever any doubt about the importance of forecasting the future, the recent survey on "Treasury Issues - 2005" conducted by the Financial Executive Consulting Group should eliminate any doubt. Among the 200 companies that participated in the survey forecasting remains the top issue for 2005."
Forecasting: If it is so Important Why Can't Companies Get it Right? - Bruce Lynn, The Financial Executives Consulting Group
It is the need for more clarity of information and accuracy that is causing this demand according to Rob de Gidlow at JPMorgan. But one of the issues is that getting the right information from the right people at the right time can be extremely difficult. "The people who are asked to provide the information, for example at the subsidiary level, sometimes do not see the benefit of doing it. When it becomes an ad hoc task rather than a specific strict and structured process, the information provided a) may not be timely and b) may not be accurate enough," says de Gidlow.
"There is no doubt about the whole point of cash flow forecasting - if you have clarity, accuracy and confidence in the information you are receiving you can know with certainty what you can invest or need to borrow. That is where the benefit of cash flow forecasting lies," de Gidlow adds.
And the business case for cash flow forecasting seems so strong. Mike Huntstad of Treasury Strategies says, "From the results of the Treasury Strategies Corporate Liquidity Survey (2004) we saw that those companies that did forecast cash out more than three months were able to pick up on average 30bp extra return on their cash portfolio. That is a very significant return given the interest rate environment we are in today." So why aren't more businesses going after this reward?
Gallanis speculated that one of the reasons that there hasn't been more progress made on forecasting over the last few years may be due to a company's concerns over the real value of improving accuracy. "If a company has a current forecasting process that might be 50 per cent accurate, they may be concerned about marginal enhancement of the results through a significant investment in time and resources," he says.
If for example a company puts X amount of resources and time into trying to improve it are they going to attain a Y per cent improvement in the forecast results, and how is that really going to translate into savings for the organisation? Does it really offset the time they put into it? It is apparent that some companies are not really clear of the benefits available through forecasting and that may explain hesitancy in implementing forecasting.
According to HSBC's Nicholas forecasting classically splits into four key areas - today, tomorrow, medium and long term. "It tends to be today, tomorrow or the next day that is the hard stuff," he says explaining that it is the unexpected receivables, that is the real challenge to accurate forecasting. The payables side is relatively easy of course because the obligations are usually fairly well known in advance. "If a company has a salary run to make through ACH payments it knows the day of the month that it makes it, can put them into the system two days ahead - and can reflect it in the cash forecast. But it is the unknown receivables side that presents the biggest headache," he believes.
The fact that many companies are still using spreadsheets and stymied in their progress may be due to the difficulties in knowing where to start on automating the process. Huntstad of Treasury Strategies explains, "For the most part companies are using spreadsheets. A lot of companies are forecasting at the business unit level, and are trying to aggregate those forecasts at the corporate level and are having a very hard time doing that." And it may be hard to believe but many companies do have a cash forecasting functionality on their TMS but simply do not use it. According to Huntstad: "For the most part even those companies that do have the treasury software or the treasury workstations are not using the forecasting functionality of their software, simply because it is not robust enough to do the kind of forecasts that they want to do."
4The Pitfalls: Systems Integration and Data Collection
One of the main problems in improving accuracy is that so much work needs to be done to clean up the data. Essentially if you put rubbish in, you will get rubbish out. At best a cashflow forecast report is only as good as the quality of the underlying input. Improving the quality of the input is pivotal to improving the overall adequacy of the cashflow forecasting.
According to Bas Rebel, Zanders, "improving the quality of input does not necessarily require complex or lengthy projects." He says: "simple tools on a treasury intranet site can create instant results. Basic Internet technology allows firms to easily share information stored in Excel files, for example, across the intranet in standard report formats. Sharing information with the originator is often the first step to improving the quality." Bas Rebel - Improving Cash Flow Forecasting - Part 2: Data Collection
Meanwhile Gallanis, at Treasury Strategies, agrees that good data is one of the most important aspects to accurate forecasting. "Up front, solid, accurate data collection and research is one of the keys to accurate forecasting. A lot of people don't envisage going through and collecting the appropriate amount of data, and they don't go through the process of really ensuring that data is reflective of historical trends. They don't prepare the data and build the framework of the cash-forecasting model effectively. So that is one of the downfalls - they are starting with a weak foundation and ultimately they end in failure," says Gallanis. Building the right foundation and building customised solutions around the forecast is really what he sees as being the right approach.
Systems integration is without doubt another major barrier to accurate forecasting. Thanks to poor integration the collection of cash flow forecast data is often the weakest treasury process. This is understandable when data is often scattered across several different systems and may even be managed by different parts of the organisation.
Rob de Gidlow, of JPMorgan, spells out the problem. "The key to cash flow forecasting is systems integration - if you don't have integrated systems and integrated processes what you find is that the cash flow forecasting information that the treasurer needs in order to get those better returns is not available," he says.
He continues: "I have never really seen a cash forecasting tool, other than a spreadsheet, that really works. There are modules of both ERPs and treasury management systems that give the ability to forecast - but if the information that drives those reports isn't updated or is manual - it might be out of date by the time it gets into the system. It's all down to accuracy of information," he believes.
And HSBC's Nicholas is in agreement on this, but also points out that treasurers still lack enough influence over their subsidiaries. This he says can be a problem in getting the clean data in the first place. "If a company appoints a central or regional treasury function to operate on behalf of its subsidiaries in a region then it ought to have the power to influence those subsidiaries to submit regular and accurate cash forecasts. This cash is the life blood of the company." Nicholas believes that the onus may be on treasurers themselves to demonstrate clearly to the subsidiaries how important accurate information is. "If the treasurer can demonstrate to the subsidiaries a value and a benefit in forecasting in the first place then they are more likely to do it," he adds.
At the next level of systems integration for cash flow forecasting a company should consider adding sales and payments information to the forecasting equation. As Nadeem Shamim of Trema outlines in his article How Technology is Making the Treasurer King of the Counting House - Nadeem Shamim, Trema Group integrating billing and invoice systems with forecasting is going to bring real benefits. "From a cash flow perspective, the best data may actually reside in sales forecasting tools but access to these and other key data sources is limited. Nor is there any guarantee, for example, that a global energy firm's French petrol retailing business uses the same tools or reporting procedures as its wholesale gas distribution operations." It is up to treasuers to make this integration happen, he says.
But this level of integration is far from normal. According to Xavier Audibert at Kyriba, companies are more interested in putting a basic forecasating system in place in the first instance. According to Audibert "We see companies saying that they need to first put the cash flow forecasting system in place - and use the information that is already available to them in the ERP." The plan, he explains, would be to later expand to calculating theoretical values and presumably linking to transactional systems. "But at the moment the wealth of information that companies have is not being used to its full potential," says Audibert. Companies are still in the first phase of forecasting - they want accurate numbers for today, tomorrow, next week.
Echoing Shamim's comments, Bhupi Singh, chief financial officer, Avolent examines the impact of eBilling on forecasting in his feature, Linking Forecasts to Action with eBilling - Bhupi Singh, Avolent
5Best Practice
So just what should companies do to address the difficulties of cleaning up data and integrating and automating their cash forecasting information?
Tony Vadasz, practice director at Parson Consulting believes that in order to approach forecasting best practice some serious consideration has to be given to the process. In his article Accurate Forecasting As A Fundamental Part of Financial Management Vadasz outlines that forecasting is not an afterthought to be tacked onto the end of the period-end reporting process, but is a critical process in its own right. In turn, business units must be given time to consider and submit high quality forecasts.
In reality one size will not fit all when it comes to software solutions to a company's forecasting needs. Gallanis, of Treasury Strategies, believes instead it is necessary to develop a tool to suit your company. "The only way to accurately develop an effective forecasting approach for an organisation is to customise a solution that meets the specific needs and nuances of a business." In terms of best practice he says: "It is quite possible that you are going to have to apply a number of methodologies to produce the optimal aggregate fit for your forecast results. You frequently have to apply multiple techniques and build it into a tool or a model that is really able to address the specific nuances of your business," he explains.
The actual methodology of cash forecasting, whether it is the model or the tool that is used, is probably less than 20 per cent of the answer, says Hunstad. Typically it is not something that is wrong with the model or with the technique that they are using. It is that the information that they are putting into the model is bad. "Typically what happens is that companies don't have a firm grasp on the true volatility of their cash flows. When variances arise and when forecasts are off they don't have a good method or communication channel to discover why the forecast is off. So it is really a matter of research - a ‘deep dive' (in-depth look) into the cash flows and mapping all those cash flows out, that is a very demanding task," he cautions. But he says that this is a very effective way to bringing companies a good portion of the way towards getting the kind of cash flow forecasting accuracy they want.
6The Technology
According to Audibert, at Kyriba, the most demand from companies is for the short- to medium-term forecasting. "Having accurate forecasts for the next 10 - 13 weeks is a key demand," says Audibert and adds that subsidiaries are more or less doing cash flow forecasting with spreadsheets at the moment. As a consequence he says companies are demanding help with consolidating inconsistent data from these spreadsheets and want a tool that can optimize variance analysis. Audibert sees differences in demand based on geography thanks to the presence of value dating in southern European countries.
"Value dating is a way for banks to apply back value date to payments. Typically in France and in Italy if you pay an amount today your account will be debited two days in the past. In other words if you want to know what your account balance will be today you need to be able to anticipate the payments you will make in two days from now," Audibert explains.
He comments that using a centralised software or common application would allow weaknesses in subsidiary information to be unearthed more quickly - throwing light on some of the subidiary info that you may have thought was OK - but when a rigorous analysis is carried out it may be weaker than previously thought.
The Kyriba cash forecasting model can link to a client's multiple banks. Clients can either enter data for their cash forecasts manually or import it from source files automatically. The idea is to allocate each forecast to a specific bank account, a specific date or a budget card, etc.
Audibert explains that spotting the discrepancies at an earlier stage is a key aspect of more accurate forecasting - as it can get the reconciliation process moving faster. "This is where we see the Internet helping companies to change their organization. Someone from payables and someone else from receivables can have access through the Internet to these positions and they can make decisions by themselves." Theoretically it should free up the cash manager to only deal with the consequences of the action (money to pay or money to invest) - rather than initiate all of the actions themselves. "In a large company it can give the cash manager the opportunity to share this information, and to decide who should be nominated to make a specific decision," says Audibert. For example, the cash manager can decide who within the payables department needs to make the decision on whether to carry a specific invoice forward or not.
Patrick Coleman, sales and marketing director, SimCorp, defines cash forecasting as both treasury forecasts and operational forecasts. For treasury forecasts SimCorp includes those future payments and receipts associated with treasury deals. Operational forecasts on the other hand are non-deal-related cash flows like single known items (asset sales or purchase) or a series of recurring items where the approximate amounts are known (e.g. payroll or rent), or even part of a multi-currency forecast over time e.g. a subsidiary's rolling 12- monthly sales forecast.
According to Coleman, many of SimCorp's recent clients are focused on improving liquidity management and this is their main driver for cash flow forecasting. "They need tools to collect forecast cash information from their business units in a controlled environment and to consolidate these forecasts into a singe position from which decisions can be made," he says.
Web-based applications (like SimCorp's IT/2 Net) can provide tools to collect forecasts from a number of business units. Coleman explains that the IT/2 system can, if appropriate, allow senior finance staff at the business unit to approve the forecast for submission to treasury. "Then, treasury can accept the forecast. At all points in the approval process the forecast is available in real time in the application database enabling the treasurer to have an instant view of the forecasted position," says Coleman.
Coleman adds that for treasurers who have an ERP system with the required cash forecast information, such a web-based system could integrate with the ERP to collect the information and present a single consolidated group forecast to the treasurer. It can be that an ERP will have detailed operations payments and receipts for the next 30 days and not longer, in which case, he says, a combination of integration to an ERP and IT/2 NET for longer-term forecasts would deliver better results.
And banks too are developing liquidity management tools. The Northern European bank SEB has developed an internet-based cash forecasting information collating tool - Webforecast. Erik Zingmark, global head of cash management sales at SEB, believes the difference between traditional cash flow forecasting and web forecasting is the ability to get all the information via the Internet in a manner where it is updated. Zingmark says: "Traditionally what happens is the treasurers get some kind of Excel report from their subsidiaries, and in some cases they have made a macro which aggregates that information into the parent company's treasury system." But there can be different kinds of problems with getting information in this way. Zingmark says: "One problem is that the Excel spreadsheet itself, sent by the subsidiary, is manually compiled meaning that you have aggregated a lot of errors in your reports."
"Companies are really trying to move from this Excel world into a more efficient, automated way of gathering cash flow information and are trying to make it as efficient as possible so that they can provide a timely and accurate liquidity position to senior management," according to Zingmark. He adds that cash flow forecasting is becoming an increasingly pressing matter for companies - but he believes that not all companies have the strength or resources to implement an efficient system.
Using the Internet to Maximise Liquidity - Erik Zingmark, SEB Merchant Banking
7The Reality
We present several case studies on what corporate treasurers in the market are actually doing.
8PSA Peugeot Citroen
Benoist Mulsant, head of cash management, PSA Peugeot Citroen.
Cash flow forecasting is a key element for PSA Peugeot Citroen liquidity management. The central treasury in Paris manages two businesses within PSA Peugeot Citroen. The first business is the industrial and commercial side - the car manufacturing and selling arm. From this the company sees cash flows of hundreds of billion euros per annum and very important financial assets. The second business within PSA Peugeot Citroen is a bank, which finances its retail customers and the distribution subsidiaries network. According to Mulsant this bank total borrowing is about €20bn. "From this it is clear that it is important to get clear cash flows forecast," says Mulsant.
"We have two kinds of forecasts. We have medium-term forecasts, which are from one to four months ahead. These are cash forecasts/ treasury forecasts. The second type of forecasts look at a rolling 10 working days period. Both are quite important to conduct our business," he says. The treasury department does not use longer-term cash forecasts.
The information from the subsidiaries comes to Paris in two ways for PSA Peugeot Citroën. The medium-term forecasts are compiled by the financial controllers within a central IT system managing the group consolidated accounting on both the bank and commercial sides of the business. Mulsant has got access to this system to retrieve the data.
For the 10-day forecasts the subsidiaries send their rolling 10-day forecasts on a daily basis, as Excel spreadsheets that are emailed to Paris. "Then we implement a Kyriba solution in Paris at the headquarters, and in Spain too," says Mulsant.
Every cash flow forecast comes from:-
In and out flows from operational activities (through accounting ERP)
In and out flows from treasury activities (through the back office system); and
Marginally from banks (through the account statements).
These are automatically sent to Kyriba and are managed within Kyriba to get these 10-day forecasts for both France and Spain says Mulsant.
"At the moment we are working with Kyriba so that the Spanish forecasts can be automatically consolidated (centralised in Paris) within the system," he adds. The company plans to put this Kyriba system in all of its main subsidiaries so that it can get the 10-day forecast directly, says Mulsant.
"The Kyriba system helps us with all our cash pooling system and the treasury operation of the main French and Spanish subsidiary." Mulsant says that the forecasts are now more reliable because they are conducted through one system. "We can see our money in a more transparent way," he adds.
"Most of our cash is invested on a quite short-term basis because we are an industrial group and not a financial group. Naturally we have got a certain amount of cash that we invest on a longer term, but this is not subject to change every month. That is independent of these cash flow forecasts," he explains.
But Mulsant says that the fact that the company now has more accurate forecasts over the 10-day period means that PSA Peugeot Citroën can optimise the maturity of its short- term investments. Regulation has not influenced PSA Peugeot Citroën's cash flow forecasting.
9Ingersoll Rand
Colin Evans, manager, treasury operations - Europe, Ingersoll-Rand European Sales Ltd.
"As free cash flow is one of the primary measures of our business it is naturally very important to us. The difficulty is in defining a method that is accurate beyond a few months," he says. He continues: "Of more importance is to try and forecast cash flow by currency. As a multinational business with multiple currency flows we want to be able to understand the size and denomination of our flows by currency in order to cover our foreign exchange exposures against accurate timelines." Evans says that cash collection and cash flow will continue to maintain a high level of priority for Ingersoll-Rand.
Due to the cycles of the Ingersoll Rand business Evans says that the company is able to produce reasonably accurate forecasts out to months by obtaining individual forecasts from the finance head in each of the company's locations. Beyond that he says it ceases to be an exact science - but he still aims to be within 10-15 per cent tolerance up to 12 months.
Although Ingersoll-Rand has looked at third party software for forecasting it is currently still using spreadsheets says Evans. The company operates multiple ERP systems - which adds an extra layer of complexity to the integration process and makes it difficult, he says, to standardise the format and quality of forecast date from the internal systems. Currently the forecasts come in by email on Excel spreadsheets.
Evans says the company takes monthly forecasts and several day-by-day forecasts (very short-term) towards the end of the month as they adjust the original numbers to greater accuracy. It is in the short- to medium-term that the company is most interested in.
Evans says that the drivers to improve forecasting is not so much regulations, as the need to have better foreign exchange risk management and better re-positioning of funds in key locations depending on where Ingersoll-Rand expects to have bank debt. The company also wants to minimise surplus liquidity in subsidiaries and sees cash flow as a measure of company performance.
Evans believes that fragmented ERP systems and the lack of suitable third party software (in Ingersoll-Rand's case) to improve information flow make it difficult for some companies to make real progress on improving cash flow forecasting accuracy.
According to Evans obtaining accurate flows by currency is actually more important to Ingersoll-Rand than the ability to know the consolidated flows. "We see this as an important tool in managing market risk," he adds.
10Fluke Europe B.V.
Martien van Dommelen, subsidiary accounting manager, Fluke Europe B.V.
Danaher Holding is a company that acquires other companies on a regular basis with its own financial funds. An accurate cash flow forecast is thus very important says Martien van Dommelen, subsidiary accounting manager, Fluke Europe B.V. The company is giving cash flow forecasting an equal amount of priority as they have done in previous years he says. "Fluke Europe is part of Fluke Corporation in the US. Fluke Corporation is part of Danaher Holding, also in the US." Van Dommelen says that all companies within Danaher Holding have their bank accounts in a cash-pool with Bank of America, HSBC or Deutsche Bank. For Fluke in Europe, for each entity, all the balances of all the bank accounts in the several countries are passed on automatically by the end of the day to a ''mirror account'' in London. For example Fluke France has an account with Bank of America in Paris and a mirror account as part of the cash pool in London, by the end of the day the balances are passed onto the London account and the account in Paris is always zero. The company, does not use any specific forecasting software, and forecasts are not integrated to the ERP explains van Dommelen. "We make the forecast for all our European entities once a month and send them to Danaher Holding. Variances that have a great impact on European forecasts need to be reported to Danaher directly, he says. Forecasts are made in an Excel spreadsheet and sent to Danaher ,he says. "Danaher Holding acquires other companies on a regular base, so the driver for us to improve the forecast is on one hand of course SOX and cost savings, but on the other hand as being a part of Danaher Holding the need to have accurate forecasts because of our acquisitions," he explains.
11France Telecom
Raffi Basmadjian, deputy treasurer at France Telecom.
Forecasts are the key to how France Telecom's treasury operates, according to Basmadjian. "Every day we have a rolling five-day forecast, per bank account and per entity. This is a very detailed forecast we issue every day. On top of that, every Friday we also receive a six-week forecast from the subsidiaries." In this way he says the five-day rolling forecasts should match with the first week of the middle-term forecast - providing a check on the numbers coming from the subsidiaries. The final layer of this forecasting comes from the 12-month forecasts received on the last working day of each month."
De Gidlow from JPMorgan says that he has seen many statistical techniques for cash flow forecasting but has rarely come across any companies that use them to any great degree of success. "I think cash flow forecasting should be kept simple - there is no rocket science behind this in my opinion."
And so agrees Glen Solimine of XRT: "More detail in a plan does not necessarily lead to a better forecast. For example, forecasting receipts by a line of products may yield more accurate results than forecasting receipts by specific models. Take the time to figure out what level of detail you will need. You want enough information to produce the results that effectively let you manage cash flows, yet don't bog you down with massive amounts of data that aren't really necessary to accomplish the goal."
But technical analysis of historical data either from a regression perspective or a statistical perspective can be a strong tool for treasurers in improving predictability. With sufficient forecast history available, regression techniques could uncover structural deviations between the forecast and the actuals. For example one subsidiary may consistently over estimate the forecasts while others could underestimate them. Spotting these types of trends can only happen if a rigorous analysis approach is adopted.
Statistical analysis too can boost the predictability of the cash forecast as submitted by operating units in a way that can be verified and is independent of human interpretation. According to Bas Rebel, Zanders: "Given the quality of input data, an initial analysis could provide clues to the extent of the enhanced predictability. The hard dollar benefit of such 'statistically enhanced' cash flow forecasting can be calculated. With the benefits quantified, a more insightful decision can be made on setting suitable priorities for a cash-forecasting improvement project and, if appropriate, allocating the required resources."
John Nicholas at HSBC believes that banks have an important supporting role to play within corporate cash forecasting. While he admits that the data that the company holds itself - within its ERP system or its treasury management system etc is rich data - banks like HSBC, he says, can help to make the information about payments and receipts to the company as transparent as possible.
Nicholas explains: "If we can extract the information we know about - because there has been a transaction somewhere in our system - and present that to the client there is some value in that. But if we can play a role in consolidating that data with other third parties, presenting it to the corporate treasurer in a way that is consistent, timely and of value there should be more value in that to all parties concerned." It is something that HSBC is currently looking at says Nicholas.
One way in which Nicholas believes that banks could be of more use on cash flow forecasting is in the area of maintaining transaction references. "The important thing is that we carry the references that come with that transaction throughout the system without truncating them so that when they do find their way to the company the company can reconcile that receipt in respect of that sale," he says.
Martien van Dommelen, Fluke Europe, says banks have an important support role to play for his company's cash forecasting. "We expect our banks to help us to set up cash pools for several European currencies and help us to have a limitation of the currency impact. Also the transfer from one account to another account should go on the same day with a minimum of charges. A good relationship with the bank is very important," he claims.
One of the problems with forecasting is that it becomes very difficult to forecast cash flows when you are looking at bank cash or daily cash and trying to forecasts flows that are coming out of accounts that are multiple-use accounts. Huntstad at Treasury Strategies gives the example of an account being used for both receipts and disbursements - trying to strip out which is which from those accounts and forecasting those individually can be difficult as it operates on a net basis - the total balance of transactions. It doesn't indicate the level of transactions - how much came in, how much went out - so those types of banking structures make it a little bit difficult. "Unless you have a very clean banking structure you have to do a lot of work cleansing the data - to get it up to a point where you can actually feed it into the forecasting model and get good results," he says.
Echoing these comments is Rob de Gidlow of JPMorgan. "Having clearly defined account structures will enable companies to see exactly what is going in and coming out of their accounts and so allows them to be able to take advantage of products like sweeping and pooling." These, says de Gidlow, can really help on the back-end of the cash flow forecasting process. And the result of that support would be treasury payments like sweeping or payment factories with an inter-company impact.
Perhaps controversially Audibert of Kyriba sees banks' interest in cash flow forecasting as a reaction to the potential threat of more open banking markets in the future. "Banks know that in some years from now there will no longer be any obstacles for a company to easily consolidate its position from any bank or send treasury payments to any bank," he argues. If an intermediary platform came into the market which would allow access to any bank there is a danger that customers may no longer visit banks Internet offerings or portals, he says. Naturally banks are keen not to lose the investments they have made here - so more interactive support on cash flow forecasting for example may be a way of providing the customers with what they want, he believes.
14The Future
So what does the future hold for cash forecasting? Are we set to see progress on it after years of talking about it?
Colin Evans, of Ingersoll-Rand, believes that future developments in forecasting focusing on the availability of software either from banks or vendors that will make it easy for a multiple ERP company to collect more accurate cash flow data in a central repository such as an intranet site and will allow better report formats. "I see the ability to use the company intranet for the collection of data and software that is easy to implement and operate in the current environment as being the "next generation" for forecasting," says Evans.
There is a sense too that cash flow forecasting will naturally fall into place with further integration - and that this should start to happen over the next few years. JPMorgan's de Gidlow believes that better forecasting will come as a natural by-product to better integration. "I believe that over the next few years as systems become more integrated and companies become more centralised the cash flow forecasting will naturally become easier and more reliable." He points to the trend towards payment factories and says:"If payment factory functionality is integrated with your cash flow forecasting model then the outgoing info becomes more accurate and complete."
Looking at regional drivers, Huntstad, of Treasury Strategies, believes that cash flow forecasting is more of a priority for American companies because of US banking regulations. Regulation Q prohibits the payment of interest on commercial demand deposit accounts. "Companies must accurately move excess cash out of bank accounts and into investment vehicles. Accurate cash flow forecasting is needed in order to do this properly," he says.
HSBC's Nicholas predicts the future of cash flow forecasting depends on the following:
buy-in at the top level of the organisation thus empowering the treasury department to dictate to operating companies and subsidiaries the need to forecast the company cash flows with performance reward schemes linked to the process.
continuous improvement in the accuracy of the forecasts enhancements to technology and improvements in systems integration acceptance that spreadsheets must be replaced.
less focus on responding to regulation/governance as these aspects are satisfied.
the banks playing their full role in the process, and, provision of the correct analysis which demonstrates the value in cash flow forecasting.
15Conclusion
As systems integration becomes better cash flow forecasting will become more accurate and reliable, but there will always be companies who are more focused on treasury optimisation - including cash forecasting - than others. Some companies will stay decentralised - for their own reasons - and will continue to do the bare minimum on forecasting. These could lose out as benchmarking performance with peer companies becomes more popular and exposes operational inefficiences. Progress on forecasting will be slow over the next few years, but the development of bank or third party software and Internet forecasting solutions should help to navigate the choppy waters of systems integration from multiple and dispersed sources. The future holds the possibilities of linking A/R, A/P, ebilling, payment factories and so on, right down through the supply chain. With this much information available to a company large sums of working capital can be freed up. Cash forecasting is just one piece of the liquidity pie - but an extremely important one at that.
As Nicholas of HSBC says: "If you could get 50bps improvement in your investment yield through better cash forecasting isn't that a good thing?" It would seem to be a hard argument to disagree with.
