The importance of forecasting working capital

Imagine your business is forecast to grow. If things were going spectacularly well and you expected the business to double in say five years’ time, by that point you’d be doing twice as much business with twice the customers. The amount of money owned to you by your customers would double (“receivables” or “debtors” in the UK).?

Although we’re imagining business going really well for you (more business, more profits) just the fact that you have to wait for new customers to pay is a drain on your own cash flow. That’s working capital. Growth is not all good – it costs you in cash.

Applying a simple pattern into the forecast can mislead you about likely future working capital 'wobbles'

A common forecasting mechanism or ‘trick’ would be to look at the historic pattern for receivables and forecast that forward.

The problem is real life doesn’t always play out like a model says it should. Maybe a key customer goes bankrupt. Maybe a customer doesn’t want to pay because cash is tight for them or they don’t like the work undertaken for them. Maybe we take on a new customer who is big and juicy but just makes us wait longer.

Look closely at the history

Rather than concentrating on wiring up a beautiful set of forward-looking forecasts perhaps the thing to do here would be to look at the history. So maybe:

• Look at the monthly position as per the management accounts, perhaps even going back 2-3 years pre-dating the point where somebody decided that now’s the time to do a business-transforming transaction and forecasting or possibly even massaging working capital is important
• On the assumption that a company, being aware that the end of the month was coming up, and could (shock horror) choose to manipulate its receivables just a little bit e.g. by issuing a flurry of invoices - you could take a close look at how the receivables position changes within the month
• On the basis that it’s pretty hard for a company to fudge the cash that’s entering its bank account, you could look at the amounts of cash real customers are depositing in the company’s bank account each day and see whether those peaks and troughs come anywhere close to the equivalent receivables position. If you can see that the company has been on top of collecting cash from customers as it approaches the end of the month, you might want to know about the precipitous drop that occurs at the beginning of the next month when all the cash drains out of the company’s own bank account e.g. as its is forced to pay its suppliers but can’t immediately find any more customers who will pay because it strong-armed them all at the end of last month.

The real-life swing (accounting for the occasional difficult or newly-bankrupt customer) is likely to give you a much more accurate picture of working capital swings compared to forecasting the nice smooth trend from the historic overall picture.?

It's OK for your final estimate of forward working capital requirements to end up?a bit rough

If, in your forecast, you thought the business might double then you might feel safe doubling your estimate of the forward working capital swing. Taking such an obviously ‘rough’ approach to working capital and effectively building a working capital ‘fudge factor’ into your model might be safer than kidding yourself that your working capital forecast really is accurate. The forecast is inevitably going to be built on top of averages. It’s likely to end up smooth. If you ended up with one box in your model marked “working capital” that one box should remind anyone looking at the model that the number you’ve got there is likely a pretty-rough estimate (and they should be pretty careful about it and maybe allow a comfortable margin for error).

Don’t kid yourself that life will play out exactly as it does when you roll forward average working capital ratios calculated to three decimal places (“customers will pay in 29.494 days”) in your Excel spreadsheet.

What about liabilities?

With some of your liabilities (e.g. tax, VAT) you could easily make an equivalent mistake – making use of historic balances and ratios and inadvertently smoothing your view of working capital requirements into the future.

It might seem sensible to look at a liability balance at the end of a past year. Projecting forward, if the business is forecast to double over time you could decide to double the liability balance as well. The change will increase your cash flow. Delaying paying the people you owe money to will preserve your cash.

But if you stop to think about how something like tax or VAT works, and how the amount we owe in tax changes within a year, you could be making a mistake. The tax authorities tend to asks you to make payments over the course of a year. At the point you make those payments your liability will be cleared down, maybe to close to zero (no matter what the size of your business) and it will cost you cash to do so.

By (it seems sensible at first) linking certain liability items to the size of the business you could inadvertently be smoothing your view of the cash flow movements that arise from working capital items. An alternative approach would have you thinking about when the tax man will want to be paid during the year and how much he will want to be paid. That way you'd be getting closer to modelling ‘real life’ cash payments and liability reductions and looking at how those payments will impact cash balances during the year. Trying to model the likely future cash payments could give you a very different view of working capital than just linking those balances to overall business growth.?

By simplifying your modelling you could easily end up mistaken about likely future working capital requirements.

A monthly model vs an annual model?

From the discussion above it should seem clear that working capital is going to ‘bite’ you within the year, particularly when there’s some seasonality involved (imagine you’re involved in a business that grows Christmas trees or makes fireworks). An annual model might be fine as a quick first look at an opportunity. But, if you think that working capital could ever be an issue in the business you are forecasting for, your model is going to need to become a monthly really quickly.

It's all a guess but make it the smartest possible guess

In real life looking at forward working capital requires a bit of guesswork. Don’t let your or anyone else’s model lull you into a false sense of security.

Note: this post is part of the series "7 things to be wary of in financial forecasts"