Conversation as risk management

To plunge right in, do you know that curve with a long tail on one side that plots size of loss on the x-axis and frequency on the y- axis? I want to divide it into three parts. There’s the big pile of business-as-usual outcomes on the left of the graph. There’s the long skinny tail on the right, showing losses that only happen once in a thousand years. And there’s the thick part of the tail near the body, so to speak, where we find once-in-a-generation smash-ups.

Conversation as risk slide as JPegWe compile data from the first section to decide what “expected” losses are, what prices to charge, and generally to understand the behavior of our assets.

We are told to study the far end of the tail to decide how much capital a given business requires. I find the logic of that exercise flawed, but I’ll spare you my rant on the subject.

I believe the thick part of the tail is what banks and their boards should focus on. This is where we find the remote but plausible contingencies stress tests investigate. This is the zone where resilience is feasible. If an asteroid destroys Sydney, the banks will not survive, and there is nothing much we can do about it. But if Australia goes into a severe recession, the caution we exercise in the last years of the boom will matter. The limits we put on commercial property lending will matter. Our capital and liquidity buffers will matter.

Limits and buffers are quantitative matters but setting them requires a certain amount of qualitative analysis. In fact, most of the important decisions we make, whether as board members or in our personal lives, involve a mixture of calculation and gut judgment. Calculation is easy. Banking gets hard, and life gets interesting, when judgment is called for, as it is in the middle part of that curve. Many bankers hope and believe big data will allow them to make an increasing fraction of their decisions with algorithms, but for now, judgment remains the heart of banking.

What role do boards play? The conventional answer is that they oversee management. I don’t like that job description. Creating a body to do nothing but shadow management would be inefficient. Recently, regulators have been encouraging boards to challenge management. This gives oversight more teeth, but I am skeptical about good decisions emerging from a stridently adversarial process.

I believe that in many ways the ideal form of organization for a bank is a general partnership, in which the same group of people are simultaneously management, board and owners. Because partners share risk, they monitor each other and bring collective wisdom to bear on important issues. But partnerships lack permanent capital and have natural size limits. Even Goldman had to turn itself into a corporation.

The division of labor between boards and managements can make them partners in fact if not at law most of the time – sharing information and responsibility, pooling their experience. In that version of proper governance, boards exist to make complex judgments – or to help make them. By “complex,” I mean issues that need to be looked at from multiple perspectives – questions that neither math nor legal analysis nor customer satisfaction surveys alone can answer. Boards can be better at making complex judgments than managements can on their own because boards bring together people with a greater diversity of skills and experiences, and because boards are less hierarchical.

A good management is diverse, of course, regarding skills, gender, ethnicity and hopefully some other things. But they immerse themselves in the life and culture of the institution. They speak a common language. They can probably finish each other’s sentences. In some respects this is a very good thing. On the other hand, non-executive directors who devote, say, four days a month to the bank’s affairs, have the benefit of mental and emotional distance, and will often see things to which the most conscientious executives become blind. In this respect, banks with non-executive boards can be superior to partnerships.

How should boards go about making complex judgments? By doing what actual partners would do – by having conversations. In my view, that’s the best way to put disparate perspectives to work.

The great values of talking are naturalness and suppleness. In contrast, at least for many people, writing seems to be a process that clogs the brain. Remembered rules slow you down. Don’t use contractions. The passive voice is weak. If you are sitting at a desk within an organization, you may feel the need to indicate your adherence to its core values or adopt your boss’s characteristic vocabulary, even though those genuflections aren’t germane. Pretty soon you’ve produced a document that doesn’t sound like you. That’s not a good sign. Just talk.

One of the things we have to decide over and over again – in life and in business – is what moral and political attitude to adopt with regard to situations we find ourselves in, what distance to deal with a problem from. These can be questions, to which logic does not provide answers. They are just as complex as the strategic and risk appetite judgments boards are designed to make. Talking about them seems to work better than writing. When you have to deliver written English on ethical issues, see how close you can come to spoken English. It is a test of authenticity.

Two contradictory adjectives could be applied to productive conversations, whether they occur in the board room or walking on the beach with a friend: “unstructured” and “disciplined.”

Lack of structure delivers a subconscious message that we’re all in this together, that hierarchy has been partially suspended, that people are empowered to speak up, to admit to confusion. One of the astonishing things that can happen when there is no clear agenda is that the frame of reference can shift. You thought you were talking about x but we seem to be on y, which is actually more important. The elephant in the room becomes visible. Issues that were being avoided go on the table. You know the right sort of conversations are happening when people in the group make statements like, “I’m so glad you said that,” or “I’ve thought that for years but I’ve never dared to say it.”

Brain-storming is a particularly open-ended form of unstructured conversation. Its animating premise – that there are no stupid suggestions – means that unexpected comparisons and connections get made. Some people find this atmosphere makes them creative. But it is not necessarily a good way to make important decisions.

Whatever the purpose of the conversation, a degree of discipline is necessary. You want talk to flow freely, but no one should be allowed to dominate the discussion. It is hard for a conversation to develop a life of its own in less than 45 minutes, but you can only carve so much time out of board and committee meetings. Like a session with a psychotherapist, an unstructured conversation at board level needs to end on time.

Having said that, the time set aside must be protected, even when the rest of the agenda is quite full. Conversation is talking but it is also listening. You have to attend to each other’s worries, precisely because they are different from yours. You have to let people mention things they have noticed recently and recount “stories that may not mean anything.” Those are the dots you hope someone connects. That’s how you glimpse the future.

Conversation as risk management 14-9-15

One thought on “Conversation as risk management

  • Yes, and… Disciplined and unstructured conversations, particularly about risks (seen and unforeseen), can come about very powerfully through scenario-building, at least as I have learned to practice it. This brings together the conversational authenticity of which you speak, and frees the mind from boundaries that are sometimes imposed, however subconsciously, by business culture, industry practice, and the power of herds. You were always one of the best I’d met at resisting these constraints, and if there were one voice raised at the perfect moment in 1997 to save Morgan Stanley from what cost Merrill $1.2 billion worth of a bad acquisition in Japan, it was yours.

Leave a Reply

Your email address will not be published. Required fields are marked *