|
|
|
 |
|
How much did the risk management calamity really cost SocGen shareholders? |
Nick Gibson
|
|
||||||||||||||||
We are told that the sub-prime losses were €2bn, the unrealised loss on Kerviel’s positions €1.4bn on discovery, and the realised loss €4.9bn by the time the position was closed out in a sliding market. A large hit by any standards. That is not, however, the full story, and there is an interesting argument relating to the massive further negative cost of poor risk management. Let’s start with a controversial statement. Thesis Over and above the crystallised losses from Kerviel and sub-prime, the crystallised reputational cost to SocGen shareholders of the failure to have remotely adequate risk management and controls was at least €800mn.This also ignores the costs yet to be incurred by having management and PWC investigating how things were supposed to work, who did what and when, who didn’t do what and when, how to make things better for the future, and in supporting the regulatory investigations and interventions by the Banque de France, AMF, and the French government. €800mn. Yes, that’s €800,000,000. A tenth of that would buy a lot of control resource in one bank. Argument – SocGen value I’ll take 14th January as the start point, because it was then that rumours started swirling in the market that Societe Generale was going to take massive provisions against subprime holdings, despite their November announcement of the €230mn “worst-case forward-looking scenario” provision.On 14th January, SocGen’s share price closed at €98.10, giving a market capitalisation of €46bn. I’ll take 28th January as the endpoint, because on the following day a very precise rumour hit the market that BNP Paribas was about to bid €92 per share for SocGen and the price started to recover. On 28th January, SocGen’s share price closed at €71.05, giving a market capitalisation of €33.3bn (a fall in value since 14th January of 27.6%). Argument - the market comparison On the same basis (to take into account the market slide that started on 21st January) between 14th and 28th January, the CAC-40 index fell by 10.3%, the Credit Agricole share price fell by 9.2%, and the BNP Paribas share price fell by 10.9%. Societe Generale, BNP Paribas and Credit Agricole are CAC-40 constituents.The worst performing of the three comparitors, BNP Paribas, provides a reasonable and conservative cipher for where SocGen’s price and market capitalisation would have been on 28th January in the absence of bad news: SocGen 10.9% down at €87.41 and €41bn. The difference between the notional conservative clean value for SocGen on 28th January and the actual market value of €33.3bn is therefore €7.7bn.
We know that markets are neither perfect nor instantly efficient but SocGen’s market capitalisation had therefore fallen by a minimum of €800mn (over and above the actual diminution of real assets over a 14 day period) measured against relevant external comparitors. That value evaporation would be as large as €1.6bn if we use Credit Agricole instead of BNP Paribas as the control. Counter arguments There is a – by no means unanimous - body of academic opinion that says that corporate value diminution arising from a serious regulatory event only exists in the short term, and that over a three- to six-month period following publication of the event the value recovers - to where it would have been had the event not occurred.This, to my mind, ignores predictable future value increases based on the financial institution’s current trajectory – it is worth bearing in mind that Societe Generale was becoming the perennial winner of various Equity Derivatives House of the Year awards, principally for its structuring activities. It seems unlikely that the associated upward value momentum will be sustainable post-Kerviel, despite the fact that he was not trading those products. The issue is now about trust. Competence Whether you are a professional or a retail client, you want to do business with people you trust. Trust is about more than honesty – it’s about respect and competence. If I think my bank has been massively incompetent I lose respect for them and am likely to change my banking relationships.For SocGen, the competence issue is not just about failure to enforce basic controls around the Delta One business. It’s also about sub-prime. We know that no-one in the industry has been calling sub-prime right in terms of provisioning, but some seem much worse than others. In November, SocGen provided on a “worst-case, forward-looking scenario” a €230mn write-down in respect of its sub-prime-related losses. Credit Agricole, at about the same time, announced sub-prime write-downs for the year of €850mn, subsequently revised just before Christmas to €2.5bn On 24th January 2008 SocGen announced a further write-down, over and above the worst-case scenario from only two months previously, of €1.65bn, with a further provision of €400mn against the sub-prime losses worsening. Just 18 days later, on 11th February, the SocGen figure was revised to €2.6bn, with a further €550mn of bad or doubtful sub-prime related assets suddenly recognised. Which of these two banks – with broadly similar write-downs - do you feel has more day-to-day control and understanding of the financial impacts of the sub-prime crisis? Which do you feel is more on top of its risk management? The one that under-estimated its potential losses by a factor of 3, or the one that under-estimated its potential losses by a factor of 11 whilst announcing that it was being prudent? The question is relevant, with the simplest example of an effect being that a bank that doesn’t inspire confidence around its risk management abilities will have to pay a higher rate to attract deposits and credit. Poor risk management is also a cost of capital issue. Conclusion There is a serious message here for risk management and control functions – including compliance – to get across to the Board as a matter of urgency.Failing to invest properly in good risk management is equivalent to taking a huge punt on the company’s future value and good name, and whilst the odds on getting caught out look relatively low, the economic costs of being a bank that loses that bet will go significantly beyond the actual loss incurred by crystallisation of the risk event itself.
| |||||||||||||||||||
© Chase Cooper 2008 |
