☰ Menu eFinancialCareers

Every analyst’s nightmare

It’s enough to make the average analyst feel faint: Terra Firma is being sued over a flaw in one of its models.

According to the Financial Times, and numerous other sources, Terra Firma’s being put on the spot over a fault related to cash flow calculations for Boxclever.

The dodgy model allegedly served to inflate calculations of the net present value of Boxclever’s TV rentals business. The company was duly refinanced by Natixis in 2002, only to default one year later. Now Natixis is suing Terra Firma.

How pervasive are erroneous models? Very, very, according to those who live and breathe them. “Every model has its mistake,” says one associate. “Most of the time they’re minor – someone will have linked the wrong cells on an Excel spreadsheet. Whether or not it’s spotted is purely a question of how much time is spent checking it. Personally, I’ve never seen a mistake make it through to the pricing stage.”

Guy Eastman, European investment director at fund of private equity funds SVG Capital, agrees that mistakes rarely see the light of day: “So many people on a deal are involved in the numbers that it’s very unlikely for a deal to go to completion with an error in the modelling.”

And if it does? “If a mistake went through to the pricing stage, the MD would take a bad hit,” says the associate. “If the error’s more mundane, the associate will get all the blame and the analyst will get a black mark.”

One error won’t end a career, but a flood of them will. “If you consistently make mistakes, you’ll become known as someone who can’t be trusted with numbers,” says the associate. “It’s the kind of reputation that takes around a year to establish, and if it happens then this isn’t the job for you – you have 12 more years of numbers to come.”

Comments (6)

  1. Any model is only as good as its assumptions – you have to accept that a model is only a model and is built by mathematicians in the belief that the real world can be confined within fixed parameters, which isn’t the case at all. Sub-prime models illustrate the the root of the problem – everyone thought they were accurate, but that has turned out to be far from the case.

  2. It is unfortunate that the model should prove te be defective later on. As one of my partners in PwC once admonished, if you don’t understand the business model first, then you can never really be a good auditor.

    Masters in finance Reply
  3. All models are wrong, but some are useful. – George Box

  4. “mathematicians” in M&A huh?

  5. It’s easier for Natixis to blame someone else than to admit to themselves that they are plain stupid to have suffered the losses. Why would they have relied entirely on someone else’s calculations?

  6. Yes, and I’d take CC’s point even further: Don’t all software products (and any other off-the-shelf product, whether a textbook or whatever) that are designed to aid investment decisions, contain an explicit disclaimer on the package and/or within the license agreement — stating that the manufacturer can under no circumstances be held responsible for any errors that result from use of the product? Of course an aggrieved customer might argue that provision wasn’t valid; but I find it impossible to imagine they’d prevail (in a US court, at least).

The comment is under moderation. It will appear shortly.


Screen Name


Consult our community guidelines here