Forecasts are a waste of time

Many companies now issue some form of guidance about the year ahead. It can be simple and vague – such as moderate sales growth – or precise and detailed. Whatever form it takes, I think it is an inherently undesirable practice that can have very negative consequences.

The first problem is obvious. Forecasting is difficult, especially with regards to the future, as Mark Twain is alleged to have said. Managers have access to better data, but they do not possess superior crystal balls. Anyone who has published forecasts will be painfully aware that reality has a habit of making you look myopic.

Many years ago, I was a salesman for an electronics company. Every salesperson had to submit an annual revenue forecast and it was a rare event when it was out by less than 20%. The errors got less when aggregated, but the outcome was still a triumph of wishful thinking over rational analysis. Group-level guidance is usually built on weak foundations.

The second problem is less obvious. Consider the common situation where the third-quarter trading update has been released but the final quarter is going poorly. The CEO will have the unenviable choice between issuing a profit warning or somehow making up the shortfall.

Under pressure

I suspect this is how a lot accounting fraud starts, especially if the CEO is an ambitious bully. It’s easy to visualise the scene at the monthly review. The CEO pressurises the entire management team to make up the shortfall with an implicit promise of reward for success and punishment for failure. The CEO will be unlikely to ask probing questions about how the hole was filled.

The sales team will be wondering whether they can bring forward some revenue. The accountants will consider deferring costs. Anyone with any ability to make a positive impact will be incentivised to do so.

The real problem starts the following year. Most accounting fraud involves moving things around; genuine creation of profits is actually very rare. This new year therefore starts with an in-built shortfall, leaving two unappealing options: the company must either issue cautious guidance or find new, bigger ways to inflate the results. This never ends well.

Lazy approach

My final issue with guidance is quite subtle: it makes analysts lazy and inhibits debate. I once asked a US analyst for the model behind her new earnings per share forecast. There wasn’t one. The company had just issued guidance of $1.50 and her ‘forecast’ was $1.48. There was no forecast for revenues, costs, interest or tax, just EPS. This isn’t research and I’m not sure it even qualifies as journalism.

This was an extreme case, but the underlying problem is widespread. Look at the analysts’ forecasts for any company that issues detailed guidance and they will be closely clustered around that guidance. It’s the same thing with economists. GDP forecasts are tightly clustered and almost always wrong.

In my view, research should be about assessing a wide range of factors, making some subjective judgments and then taking a view. Slavishly following the company line is a dereliction of duty. How can you have an interesting conversation with someone who has no independence of thought?

Issuing guidance is supposed to improve the dialogue between management and shareholders. I think it often does the opposite. Guidance is optional in most countries. In my experience in the industrial sector, the best-managed companies declined to issue guidance because they thought it was a waste of time. If only this sensible approach was more common.