We are currently contemplating a 3 year rolling forecast. I have raised some issues as to the value of the latter part of that forecast, given that as you get further out, the numbers become more sketchy and subject to significant change with the next update. What are your thoughts and what are other companies doing?
What is the value of rolling forecasts as duration increases? (Webinar Attendee Question)
Answers
In my opinion, the value of rolling forecasts is that they keep
Benefits of a rolling forecast include calculating loan covenants, employee bonus levels, capital investment capacity, and dividend options.
Best of luck with your forecasts.
We run an 18 month forecast and the numbers in the out quarters are certainly more subject to change but I agree with Henry's comment. It keeps your managers thinking out beyond the current year. We also periodically produce a five year plan but this is a bigger effort and happens less frequently than the once a quarter forecast.
We have used 18 month rolling forecasts and of course we put more emphasis on the current year, but it is helpful to see where we see the business moving. As shared above, it is helpful to get managers to think about the bigger picture and ensure the Company is addressing issues before events are imminent. We spend the majority of our time reviewing current month and then we review the quarter, then the current year. Beyond this, we make sure that we are on track to make sure we won't trip covenants, modify our cost structure and get a sense of where the business is heading.
As with anything it depends on the nature and circumstances of the company and the sector it operates in as to how far you go out. I worked for as FP&A manager of a semi-conductor manufacturing company and each month we'd do a quick close, review results and then compile a rolling 18 month forecast. This created a great amount of focus and discipline throughout the company worldwide. We also did a five year plan but typically after the first quarter of the new year that plan did not have much relevance other than for some more high level strategic issues. The 18 month rolling forecast drove the operating plans. In my opinion it was an excellent management tool and process.
My experience in financial services was like Marc's above. In periods of great change (most of the time), the effort was considerable, but we kept working process improvements to truncate the timeline. We also continued to benchmark key performance indicators (throughout the industry) to simplify the process.
Even in a very stable industry, all forecasts have the negative of being seen as a likely future (as in "expected value"). Their value to me is in modelling the impacts of trends. Tying this to KPIs (per Linda) helps you to understand what to look for. Some of the impacts are longer term, and a 3-year forecast* can be very helpful. For the companies I advise they are standard. Nobody believes them, but everyone wants to know how the business scales or fails given the known unknowns.
*We don't call them "forecasts", incidentally, as that suggests we have an opinion about the likelihood of the outcomes. We discriminate between shorter-term forecasting (think weather) and longer-term business-modelling.
With an ever changing economic climate, I can almost see a rolling 18 forecast working but don't believe there is any value in going out much further. This exercise would certainly keep the finance folks busy but to what end? I would rather see greater effort and quality put into a rolling 12 month forecast that management would deem more accurate and actionable. For those who have gone out further, have you gone back to earlier forceasts to see how they correlated to actual performance over time? I believe that you would be very surprised.
Rolling forecasts are a vital tool in managing results and expectations for the future. You have to be aware of the need to tweak the model for changes in assumptions. Definitely 18 months, but 3 years is not out of the question, especially if there is a long-term goal. (If something goes awry early on, you need to know how to regroup.)
I believe that the question being asked actually has two parts and is easier the solve when you split it that way. The first is the use of rolling forecasts. Over the last five years finance has definitely shift to agreeing with the usefulness of this tool. Virtually all comments above agree that yes you should use rolling forecasts.
The second part of the question ask what is the appropriate time horizon for the rolling forecast. The questioner is concerned about the uncertainty of the later half of three years. Uncertainty is a problem in some degree with all forecasts - it just magnifies the further out you go. In our book Future Ready: How to Master Business Forecasting, (Wiley: 2010) Steve Morlidge and I provide a lengthy discussion of time horizons in Chapter 3. Basically you should pick the time horizon that matches the decisions you are trying to support.. Most organizations are finding 18 months as a good horizon but we have seen as short as 6 months and as long as 10 years. The longer horizons tend to get into capital requirements (i. e. a semiconductor manufacturer might use a five or ten year horizon because they know it takes four years to build each new fabrication plant).
Be cautious when you check on forecast accuracy after the fact. A forecast tells you where you think you are headed. If that is not where you want to go, management actively works to change direction. If they are successful, then management actions invalidate the prior forecast.
Personally I believe that financial models are an essential part of your financial
The timeline depends not only on your company and industry but also your audience. Banks and credit rating agencies will want to see some forward looking data to assess your credit-worthiness and you wouldn't want to share something with them that the BoD hasn't also seen. In this case, how far out you look will depend on the drivers of your business and the tenor of your debt.
Working in heavy industry where investments can take years to come on line and decades to give positive NPV I have produced models that show timelines from 3 months to 30 years. In terms of accuracy don't get bogged down with the $$$ - we know we can't be right predicting 30 years in to the future - focus more on the drivers and sensitivities. Omitting or misunderstanding one of those is much more serious than getting the absolute figures correct.
I think it depends on the length of your business cycles. Companies providing long term projects lasting more than a year need a long planning horizon. If it takes a year to acquire resources then you better forecast out that far. Otherwise, its not really necessary.
For most clients, I get a lot of value from a 6/3 rolling forecast. That's six months actual history and three months of projections. This kind of short ran planning helps to maneuver and plan resource levels fairly well.
Also trailing 12 month numbers are great finding sales trends when you have volatile performance issues.
Forecasting accuracy improves business performance. If aligned with the financial plan for 1 year, rolling forecasts over the next 90 days, with updates at periodic intervals, can improve liquidity management. After all, cash is king. At most businesses finance folks at company operating units are certain of cash flows in the next 1-2 weeks, after which changes can be substantial. If you are having multi-currency transactions, accuracy in your cash flow forecasts, can translate into better FX exposure management and minimize the need for larger cash buffers. Accuracy over a 15 day period in organization's cash flows can translate into efficiency in the management of working capital. A lot can happen beyond a month's horizon. Given the rapid changes in the environments affecting the business, any increase in the accuracy of cash flow forecasts, will surely enable
A rolling forecast process is an excellent way to get your sales and operating groups to think strategically with regular frequency. I recommend a simple process that is not overly complicated or administratively burdensome. The term of the forecast depends on your product life cycle and market dynamics. If you operate in a dynamic market where product life cycles are short, then you will get limitted value trying to envision how markets and customers will change and what products they will require. If product life cycles are long and the market demand and customer needs are consistent over time, you will get great benefit out of the process. As an example, we used an 18 month rolling forecast process in a fashion-oriented business as the typical product shelf life was between 12 and 18 months, but the customers core needs were consistent. We created a rolling 18 month forecast on a monthly basis because it was important to constantly monitor customer demand so that we had plans to replace existing product with downtrending demand with new product. High fixed cost businesses with long product development lead times (e.g. automotive, telecommunications equipment) also benefit from a rolling forecast process because it requires the regular review of market demand to ensure that today's R&D investments get good returns or that resources are adjusted accordingly.