cowboys and germans

more thoughts on credit suisse

Some additional cybernetic thoughts on corporate culture and banking…

Over the weekend, I wrote a piece for Starling, loosely linked to the publication of “Meltdown”, Duncan Mavin’s new book about the rise and fall of Credit Suisse.  I tried to make the case that the culture Duncan describes (scrappy, aggressive, rule-bending risk takers) isn’t necessarily always all bad in a banking context, and that one piece of evidence for this is that CS lived and died without ever taking taxpayer money, and that it was one of the few banks to come out of the 2008-9 crisis with a better franchise than it went in.

The reason for this is that CS had an almost perfect risk culture for a disaster like the Great Financial Crisis.  It was a company of cowboys, who knew they were cowboys.  The top executives knew that they were managing a rodeo, and that their job was to prevent it from turning into a goat rodeo.  And all the individual cliques and gangs within the firm knew that they had to keep an eye on each other.

When the most important risks are external, and relate to the inconvenient interaction between current publicly available facts (house prices are falling) and past decisions about the issuance and valuation of traded securities, a team of cowboys is what you want.  They will cut positions ruthlessly, aim to preserve capital, look for spots of opportunity and get back into the market when it’s safe to do so.  Risk management systems set up to manage cowboys are usually very strong in terms of having a handle on the size of positions, the amount of implicit leverage and the existence of risk concentrations.

The trouble was that the crisis was (in my view, incorrectly) blamed on “excessive risk taking”, rather than “excessive attempt to avoid risk, by advertising securities as riskless when they weren’t”.  And so something like the musical “Oklahoma” took place, where the cowboys were encouraged to settle down and quit their lawless ways.  Credit Suisse adapted very badly to this, because risk management in the new world was all about detecting bad apples like Lex Greensill and Bill Huang. Systems that are basically adapted from the cowboy days weren’t set up to do due diligence, model client behaviour or deal with problems that couldn’t be exited at market price in a short period of time.

I could have told a story about Deutsche Bank, if I’d wanted to – it also had a great business model pre-crisis and also came through the crisis without taking taxpayer money directly.  But Deutsche’s key franchises were dependent on its status as a “national champion” in a different way – as a too-big-to-fail client of the most solvent big state in Europe, it had unquestionable credit risk and liquidity, and this allowed it to write a lot of complicated long-dated interest rate hedging contracts. 

These “swaps” contracts were very low risk, and correctly treated as such for regulatory purposes, but they were consequently low margin and so Deutsche had to write a huge amount of them.  The swaps hung around on their balance sheet and made the accounting assets very big compared to the risk-adjusted version against which their capital position was judged.

After the crisis, people decided that they didn’t trust “risk weighted” assets calculations any more, and so Deutsche was suddenly regulated based on its accounting balance sheet.  At the same time, the EU changed the rules on state guarantees and bank bailouts, so the national champion status was much less useful as a guarantee of funding costs.  The swaps book transformed from a healthy source of profits that were set to drip-feed out over twenty years to an albatross; what I referred to as a “fatberg” of assets that couldn’t be sold because they were too big and too complicated for anyone else to buy them.

In many ways, Deutsche had the much worse problem – its risk management systems for the fatberg weren’t always able to even provide daily reports, because it had never got into the habit of being questioned about them.  And in my understanding, the bank came a lot closer to losing its New York banking licence than anyone wants to admit.  So why did one bank survive and the other not?

I genuinely think that the answer is in Stafford Beer.  In both cases, we are looking at a threat that comes at the highest level of organisation, and tests viability in the most fundamental way – the environment has changed in such a way that the existing arrangement of the operations (“System 3”) is no longer tenable.  In that situation, survival depends on the ability of the intelligence system (“System 4”) to understand and transmit information about the new world, but even more so on the capacity of the highest level organisational subsystem to translate the new intelligence into decisions about what kind of change is needed.  System 5 is what Stafford Beer called “philosophy” or “identity creation and re-creation”.

On this analysis, Deutsche Bank survived because it had a stronger sense of itself – it had the name of a G7 state above the door, and knew that potentially everything had to be sacrificed to the core mission of “providing financing for German industry worldwide”.  Credit Suisse also had the country name, but it was never the dominant Swiss bank, and had always been split between domestic and American factions (both of which had separate and not always overlapping fascinations with Asia).  Credit Suisse basically died because (and plenty of people said this, even during the good times) it just never made sense.