ACCA - The global body for professional accountants

Do rolling forecasts offer businesses a better way of making tricky decisions in a difficult climate? We talk to IBM financial performance expert Mark Rolfe FCCA

This article was first published in the April 2012 UK edition of Accounting and Business magazine.

Ask a finance director to list what a business needs to survive in these uncertain times, and they will talk about adaptability, responsiveness and planning for different scenarios. They’ll also tell you about having a firm and accurate grip on costs and a realistic assessment of revenue streams.

For most businesses, the boom days when they could comfortably expect annual growth are now a distant memory. Yet many have not changed the way they budget to adapt to the new, more challenging conditions.

Mark Rolfe FCCA, an IBM financial performance management expert who has been giving a series of talks to ACCA members on the theme of ‘Rolling forecasts in turbulent times’, says that a surprising number of companies still take the ‘traditional’ approach. In the good old days before the financial crisis this usually involved adding a few percentage points to revenue and costs at the start of the budgeting year, then tweaking it with a bit of mild quarterly flexing.

‘Very often they are just budgeting up to the end of the financial year, so you end up with what we describe as “the wall”,’ he says. ‘If you take a calendar year and you get to, say, March and you want to make a decision, you have nine months of forward-looking data. But if you’re making that decision at the end of September, you only have three months of data.’

Rolfe says that while most businesses, quite sensibly, run their budgeting and forecasting processes together, a distinction should still be made. A budget is a view of where a business wants to get to, while a forecast is a view of where it believes it is going, either in current circumstances or as a result of changes it can make.

‘They both need to be honest, but if a forecast is not going to be a brutally honest view of the future, then it’s going to be challenging to make it the basis of good decision-making.’

The financial crisis, Rolfe says, has elevated the importance of forecasting in people’s minds and made it very obvious that a traditional budget can be outdated very quickly.

Under a typical rolling forecast system, instead of just flexing the annual budget every quarter until the end of the year when the whole process begins again, the business looks forward an entire year each quarter (or whatever timeframe is appropriate to the business), so that there is a more consistent time horizon. This eliminates the wall, and the company has business planning in place for at least the next nine months.

Other advantages of rolling forecasts, says Rolfe, include a reduced reliance on outdated budgets and more objectivity by reducing the otherwise huge focus on year-end. He also points to studies – for example, by the Hackett Group – that suggest rolling forecasts speed up the budgeting process.

Practical challenges

In general, Rolfe says, although there is huge interest in making the switch to rolling forecasts, only a relatively small number of businesses have so far implemented them to any real degree.

But there are more challenges to improving budgeting and forecasting than just extending the time horizon.

‘Producing assumptions for a budget is much more difficult in the current economic climate,’ he says. ‘And that’s been made much more significant by increasing globalisation. An event on the other side of the world can actually have a pretty significant impact on a company in the UK or the US.’

This, he says, along with a growing variety of data sources, is making life even more complex for finance professionals. ‘At the same time you’ve got governments and regulatory authorities adding more and more requirements, whether that be reporting, filing, corporate social responsibility reporting.’

On top of all this, businesses have an increasing need to plan for different scenarios. He cites the IBM 2010 CFO Study in which 74% said the difficult economy was generating a need for faster decision-making.

Event-driven

‘We’re increasingly seeing organisations looking at event-driven forecasts – getting themselves into a position where they can produce a forecast in response to a particular situation, or opportunistically where they see a particular opening.’ This, he says, produces two challenges.

The first is practical – the huge reliance many finance departments still place on spreadsheets. Many, says Rolfe, rely pretty well exclusively on spreadsheets, so limiting their ability to introduce any element of what he calls ‘multidimensionality’ into their processes. This means using a set of data entered into a spreadsheet for a purpose other than its original purpose, which, says Rolfe, can become very problematic, especially for those who often need to pull forecasts together speedily.

The other challenge of event-driven forecasting is that it requires a much better grip of what drives a business than does traditional budgeting.

‘Finance professionals need to be starting to think in the language of the business,’ Rolfe says, ‘looking at the real business drivers.’

He gives the example of a call centre manager being asked for a revenue forecast. ‘They would probably give you an estimate based on previous periods, but the reality is that they are going to be thinking more in terms of opportunities created. They’ll be thinking about the number of staff they have, the number of calls they can make, about cross-sales, and the relationship between all these.

‘So, for instance, the number of calls will drive opportunities, which, via an average deal size, will create a pipeline, and a conversion rate assumption will take you to a revenue figure. It’s about trying to forecast in a business language.’

So how do you start making these sort of changes, especially with cost control now so tight? Rolfe recites the mantra: ‘Think big, start small, deliver quickly. If you start thinking about where you would ultimately like to get to in your forecast process, you could take a particular area, or start with a small change to your processes. If you get some changes delivered, so the business can actually see some benefit, you’ll get a lot more buy-in.’

One of his suggestions is to extend time horizons on forecasts for certain items – perhaps looking at them quarterly rather than monthly when forecasting further ahead, allowing greater time to be focused on key business drivers.

Take a step back

‘There are a lot of things that you could do that are not necessarily expensive, but just involve taking a step back from the business and saying, “What benefit do I get from my current process? Is it actually helping to make decisions?”’

He also points to the evolution of many corporate accountants into finance business partners, increasing their understanding of a business and thus able to provide better information on which decisions can be based. ‘Back when I was working for Nintendo [as controller of financial planning], I moved out of the finance department into sales and had a finance role there. That was relatively unusual back then, but I think we’ve gradually started to move out of traditional finance roles and become business partners.’

Other secrets of success, he says, include fostering a forecasting culture. And, of course, technological upgrades may ultimately be required.

Not, he warns, that any of this will result in a business being magically able to predict the future. Quoting Peter Schwartz in The Art of the Long View, he says: ‘The end result is not an accurate picture of tomorrow, but better decisions about the future.’ If the process does not enable better decision-making, then its value has to be questioned.

SIX STEPS TO SUCCESSFUL ROLLING FORECASTs

TIME HORIZON

Think about what is appropriate to your business, says Rolfe. ‘From a business perspective, a retailer would have a very different timeline to, say, a pharmaceutical organisation.’ He also suggests using different time horizons for different items. ‘You would use a different horizon for, say, advertising than you might for capital purchases.’

PARTICIPATION

You need to build in a means of engaging the business with the forecast process so business colleagues are contributing to it and taking ownership of their contribution. ‘Getting some accountability and transparency into your process is important,’ Rolfe says.

MODELLING

‘Talk in the language of the business and try to include real business drivers,’ says Rolfe. ‘Start to talk in terms of what the business can understand itself.’ He adds that many businesses tend to forecast in too much detail, so it’s better to put more effort into areas that are particularly volatile or material, and take a lighter approach to other areas.

RISK AND UNCERTAINTY

This is a big area and it can be approached with varying degrees of complexity, says Rolfe. ‘You could start by introducing upside and downside elements into the process, or go all the way to quite complex modelling and stress-testing.’

CULTURE

‘You need to have an acceptance by the organisation that it wants to forecast – and accept – an honest view of the future,’ says Rolfe. ‘There needs to be a culture of wanting to discover how to change the business – not the forecast – if you see a result that you don’t like.’

He highlights common behaviour such as management using forecasts solely as a target-setting process, and ‘sandbagging’, in which budgets or forecasts are understated so an individual, department or whole business can enjoy the benefits of exceeding expectations or even to mislead a rival into over-reaching.

MEASUREMENT

‘A lot of finance departments have traditionally spent a lot of time explaining variances on actual versus budget, which would be better spent looking forward,’ he says. ‘A part of the way you can improve forecasting, however, is to start measuring your forecast accuracy. There may be a modelling error or bias in certain parts of the organisation, and measurement is critical to improving accuracy and effectiveness of forecasting.’

Chris Quick, editor

Last updated: 20 Mar 2014