Carillion's Downfall - Not A Complete Surprise, It Seems
Carillion, the collapsed British construction and services company, has been the subject of much media and analyst attention recently. Rightly, much of this attention has focused on what went wrong with Carillion’s business and the decisions of management, still more has sought to question how analysts and large institutional investors failed to notice signs Carillion was in difficulty, and that those difficulties where compounding.
While undoubtedly true that many experienced and seasoned analysts, bankers and asset managers appear to have failed to notice Carillion’s increasingly perilous position, there were compelling indications not everyone was happy with Carillion – and for those listening, the indications go back over a year.
AlphaWeek asked Irithmics, a Bath (U.K.) based fintech using artificial intelligence (AI) technology to analyse the dynamics and behaviour of institutional investors and listed companies, what insight AI might provide as the investment world continues to reflect upon Carillion.
The first striking observation from Irithmics’ AI was that as far back as Q4 2016, longer term institutional investors (the so-called “smart money”) had been paying very close attention to Carillion – Irithmics algorithms categorising long term investor activity as likely to be “Very High” since September 2016 and throughout 2017.
On 10 July 2017, Carillion announced chief executive Richard Howson was stepping down. The firm was also pulling out of three construction markets in the Middle East and that payment difficulties associated with four construction contracts meant it needed to make a GBP 845 million provision. Markets responded drastically and the following day Carillion's shares fell 30 percent as analysts doubted whether the company's efforts to conserve cash would be sufficient. It appears this news was not entirely unexpected by many market participants - Irithmics’ AI showing that throughout Q2 of 2017, long term investor outlook for Carillion had been consistently negative and in the run-up to the announcement, investor sensitivity to news, in particular negative news, had begun to rise – by late June 2017, it appears the “smart-money” knew something was very amiss.
Carillion’s difficulties, and the potential impact on the company’s stock, definitely did not go unnoticed by some and Irithmics’ AI, identifying patterns in the dynamics and behaviours of investors, suggested the firm was attracting “Very High” attention from systematic strategies and short selling investors as far back as mid-May – two full months before Carillion’s profit warning on 10 July.
The question seems not to be how did investors fail to spot Carillion’s difficulties (Irithmics’ AI technology clearly suggests many were increasingly aware sometime before profit warnings, stock price falls and resignations). The more intriguing question is why so few failed to act, but that is possible a question more appropriately directed towards analysts, bankers and asset managers rather than Irithmics.