Such losses are so big they can seem almost unreal. That can serve a dual purpose. First, the deficit can be dismissed as the product of “mere accounting”, the result of pedantic number-crunching. Assets have been written down in value, but that is not the same as a cash loss.

Second, when new managers are appointed, a huge loss can be blamed on the previous regime.

All the bad news can be revealed at once, a phenomenon known as “kitchen-sinking”. From that point on, the only direction for profits must be up.

Analysts tend to be very supportive of such arguments. They typically dismiss big write-offs, even when they failed to forecast them. They argue that it makes more sense to focus on operating profits, which reflect the health of the underlying business. These, it is said, are a better guide to the future direction of the firm.
.

The same arguments crop up at the aggregate level. Robert Shiller of Yale averages profits over ten years to get his cyclically-adjusted price-earnings ratio. So this measure includes numbers from 2008, when firms in the S&P 500 index reported a plunge in profits thanks to huge write-downs at banks (see chart). This, argue critics such as Jeremy Siegel of the Wharton School, makes the stockmarket look more expensive than it truly is.

This optimistic view, dubbed “earnings without the bad stuff”, has deep roots. Investors place a high value on predictability. They like companies at which profits grow, quarter after quarter, year after year. They want the corporate sector as a whole to produce steadily higher profits—and that is just what analysts tend to forecast. Predictions of recession are rare.

Indeed, back in February 2008, Mr Siegel said there would be no recession that year and that profits, the stockmarket and the financial industry would all do well.

Companies are well aware of investors’ biases and strive to gratify them. But this creates behavioural problems. A short-term focus on the next quarter’s earnings means that companies may reject long-term projects that could boost the value of the company but would hurt profits in the coming year or two. Managers will also resist changing a failing strategy (and thus admitting assets are worth less than before) until the last minute.

Take the recent catalogue of disasters at Tesco, a British supermarket chain. The company unveiled one set of write-downs in 2013, reflecting poor results in its foreign operations. But that was merely the prelude to the even bigger £6.4 billion ($10.4 billion) loss announced last month, which reflected a write-down of the value of its British property. Tesco had been slow to realise that its customers were losing enthusiasm for huge out-of-town stores and were switching to online shopping and to discount stores such as Aldi and Lidl.

Seen through this prism, a write-off of assets is not merely an accounting nicety. Ultimately, equity holders have a stake in the assets as well as the profits of a company; if those assets are worth less, shareholders are poorer. To the extent that the business paid too much to acquire those assets, that is a loss as real as selling baked beans below cost. Indeed, investors’ calculations of future profits were probably based on the company’s ability to earn a decent return on such assets. Furthermore, such write-downs tell investors something about the management’s ability to understand its own industry.

In the case of the financial industry, the massive write-downs during the crisis were a necessary correction of the overstatement of profits in previous years. Banks had generated profits by issuing, underwriting or lending against mortgage securities. Those profits turned out to be illusory. If you were to ignore the ensuing write-downs, you should also adjust down profits for previous years.

The simplest answer, which lies at the heart of Mr Shiller’s analysis (and was proposed by Ben Graham, the doyen of security analysts, before him), is to average profits over several years.

The mistake is to believe rosy forecasts for future profits, and to dismiss spikes in losses as exceptional. Write-offs are usually right.