Thinking about banking

Some years ago, I was introduced to a well-known writer at a party in Melbourne. “Harrison’s a banker,” my host said helpfully.

“Gee,” responded the writer without missing a beat, “I’ve never met anyone in organized crime before.”

I spent half a second searching for a clever response. But the Global Financial Crisis was raging, and bankers attract a certain amount of abuse even at the best of times, so I just laughed. I’d like to offer a better account of our profession today – and I thank the Financial Review for the opportunity to do so.

Few banks around the world emerged from the GFC with honor. There are actually two ways of looking at that fact. It may be evidence, as my new friend implied, that bankers are scoundrels. Or it may tell us that banking is hard. I endorse the latter interpretation. Some portion of the public inclines to the other view – even here in Australia, where the crisis was managed well.

In the next twenty minutes I want to give you three ideas to argue about at lunch. That was the first one: that banking is not as simple as it looks. Everyone in the industry kind of knows this, but we often fail to acknowledge it.

The second idea is that banks should not be profit-maximizing institutions. They have duties to the community that oblige them to forego a certain amount of upside.

The third idea is that banks cannot be profit-maximizing institutions. Or to put it another way, the path to that destination has a twist in it. At the top of every economic cycle, banks will tend to underestimate the risks they are taking. The economically rational response is to take an amount of risk that feels irrationally conservative.

Why these three ideas? They illuminate the technical, ethical and psychological challenges of our calling. Taken together, they tell us what it means to be a banker.

Please do not think – to go back to idea number one – that I am saying greed and corner-cutting had no role in the GFC. There were reasons for that writer’s

cynicism. Nor do I wish to make excuses for bad behavior. What I am saying is that it does not require criminal misconduct to undo a bank. Our business model makes failure disconcertingly easy.

Most people here will know the model’s essential features: maturity mismatch, opaque assets with far more downside than upside, markets capable of whimsical violence, quite a bit of leverage. Banking is a watchful profession. We’d really like to see around the corner.

Human beings cope with future perils by making mental models of the past and searching the present for clues. We’ve been doing it forever. What time of day do the lions hunt? Are there fresh antelope tracks? It doesn’t always work, but that’s what we do. Banks manage risk the same way.

Correction. Risk means known odds and unknown outcomes. What banks and humans contend with is uncertainty, which means unknown odds. But never mind. Banks infer the odds, and then act as though they were certain, which usually works. If you make a few million residential mortgage loans, you can get pretty good at predicting default and loss rates.

There will always be nasty surprises. A bank needs capital and liquidity buffers to ride out recessions, rate shocks and other remote but plausible contingencies. This is where the business starts getting hard. Rare events are difficult to model. You may not have much data. There turn out to be unexpected correlations and non-linearities in remote-but-plausible-contingency-land. Building a correct estimate of average losses into your pricing is important, but you can make adjustments as events unfold. Deciding what buffers to build into your balance sheet is harder. By the time you get market feedback, it may be too late to make changes. But we all know that.

One of the principles imbedded in the brains of wise bankers is that new means risky. My favorite illustration is “out-of-territory loans.” For many years, U.S. banks could only have branches in a single state. These restrictions did not apply to lending, so ambitious banks sometimes chased growth by deploying loan production officers to states where demand was strong. The resulting out-of-territory loans so often turned out to be troubled that analysts regarded growth in that category as a red flag. Interloping bankers didn’t hear the gossip, couldn’t recognize signs of potential distress – with the result that they made the loans the local banks wouldn’t.

The same concerns apply to rapid growth at home, which requires new customers. In fact, anything new calls for caution – new markets, new products, new IT systems, new competitive threats. Your radar is likely to be deficient. And you won’t know what you don’t know.

Being suspicious of anything new is a drag. You feel like a wimp. You miss out on the excitement. People call you a curmudgeon. And in fact, you mustn’t shut the door on change, no matter how wary you train yourself to be. The community expects you to support the economy. Your shareholders want growth. And since change is going to happen whether you like it or not, it might be smarter to anticipate it. So you are pulled in different directions, sometimes quite abruptly. Steering the right middle course can be difficult.

And then there’s what we might call the Minsky phenomenon – in honor of the economist who drew attention to it. As time passes since the last crisis, risk aversion leaks out of the system. What Keynes called “animal spirits” return – which is not a wholly bad thing. People become less frightened, and then less cautious. They decide out-of-territory loans aren’t that risky. They hire more staff. They borrow to expand capacity or to innovate. And some people borrow too much.

Minsky’s insight means that every equilibrium contains the seeds of its own destruction. A banker must never forget that what gets called “prosperity” may turn out to be progressive madness. William McChesney Martin, who served as chairman of the U.S. Fed for nineteen years, is remembered for saying that the job of a central banker is “to take away the punch bowl just as the party gets going.” A commercial banker has a similar responsibility, but has to drink the punch herself while deciding when to act.

Second idea. Because of the enormous damage a failing bank can do to the community it serves, and to the broad economy, a bank should pursue a minimax strategy. That’s my view, at least.

A minimax strategy means playing it safe. It seeks to minimize the maximum loss that could occur. Carrying an umbrella on a sunny day is a minimax strategy. You do that if you only have one suit and an important job interview tomorrow.

Suppose a bank faced the prospective payoffs shown in this matrix:

Strategy A

Prospect 90% chance of 50,

10% chance of -200 Expected value 25

Strategy B
95% chance of 100, 5% chance of -1000


The probability of a positive outcome is higher under strategy B, the probability of a bad outcome is lower, and the expected value of the strategy is much higher than under strategy A.

Externalities not reflected on the chart mean a bad outcome under strategy B is horrendous, however. The bank fails, but the damage extends far beyond the thousand of loss absorbed by holders of its shares. There is a financial crisis and a serious recession. Unemployment surges. Blameless citizens lose their homes. A board that believes it has a responsibility to the community would choose Strategy A, even though, to a portfolio investor, Strategy B would be the economically rational choice.

Similar decisions not to chase down the last dollar get made every day at every level of a bank. Do we restructure or foreclose? What is our policy for hardship cases? Does this quite insistent customer have a legitimate complaint? A bank has to protect itself. There are plenty of scoundrels out there. But in most of its dealings with the public, banks have preponderant power. Concern for the interests of the shareholders – which, please remember, are legitimate too – must be balanced by a sense of fiduciary obligation. You cannot call yourself a banker if you do not feel that obligation’s tug.

For any neo-classical economists in the audience, I recognize that upside forgone can be regarded as a license fee, meaning that banks that pay it are profit- maximizers after all. But doing the right thing doesn’t feel like paying a fee. A sensible board will consult its collective conscience. Except as an academic amusement, I would argue against building a model to decide how virtuous to be.

The third idea has to do with the fundamental opacity of risk – and the necessity of financial conservatism. A potential message people might take from the stream of scholarly papers flowing out of Basel is that a competent bank can measure the risk the enterprise as a whole is taking. In my view, it cannot. If you are looking at a single line of business, and you have good data, it is possible to build a model that tells you the probability distribution of outcomes. But risks interact. Credit losses kill a bank because of their impact on liquidity. Operational failures damage reputation. Building a model that accurately reflects the probability of such chain reactions among multiple businesses is impossible.

To be clear, banks can and do, through an ad hoc mixture of quantitative analysis and common sense, get their arms around the risks they are running. Stress tests and scenario exercises help a board and senior management explore hidden linkages and transmission mechanisms. Most of all, they are a vehicle for discussion, which is the best way to pool experience and refine judgment. But when it comes to answering the questions, “How bad could it be?” and “How likely is that?” – or in quant speaks, “How fat is the tail?” – decision-makers will rely heavily on the group’s collective gut instinct. They also rely on instinct when exploring unfamiliar territory. And instinct varies.

This chart illustrates what can happen.

Actual Risk


horizontal axis represents perceived risk. The vertical axis represents actual risk.

The diagonal dotted line connects the point where perception equals reality. The

wiggly solid line is the path of a given bank over time.



It’s a map, not a graph of a function – and a conceptual map at that.

The Perceived Risk

What we care about is actual risk, but all we can have to go on is our perception of risk, which is an amalgam of raw data, model output and gut feel – the pattern of the past and current clues. We identify a level of perceived risk we don’t want to exceed – the vertical line marked “A” – and hope that keeps us below the ceiling – marked “B” – that we wish to put on actual risk.

A bank can take a variety of paths, but this one is typical. It starts in the lower left, frightened by the last crash, not taking much risk, and if anything overestimating the risk it is taking. It weaves back and forth a bit, but so long as it sticks to products it understands and customers it knows, it tracks the diagonal dotted line fairly well.

At first, it moves slowly. Nothing bad happens, which increases everyone’s confidence. Then one day – at the point on the wiggly line marked “C” – the CEO decides he wants to be more aggressive. “We’re businessmen,” he reminds his colleagues. “When we spot opportunities, we’re supposed to pursue them.” To the great misfortune of everyone involved, the other directors like his attitude.

Soon enough, the bank’s path moves significantly above the dotted line. Pursuit of new opportunities means that actual risk is increasing rapidly. But it doesn’t feel that way – for two reasons. The bank is engaging in the functional equivalent of out-of-territory lending, so it misses important danger signals. And as predicted by Minsky, good fortune has dulled the group’s general sensitivity to risk. As you bask in the sunshine of past success, increase position limits, make larger loans to weaker credits at fatter spreads, grow faster, offer new services, enter unfamiliar markets, trim capital and liquidity buffers, undertake more IT upgrades and get less sleep, at some point, losing control becomes easy. By the time the bank reaches its perceived-risk boundary it will have gone way past the boundary that matters.

Risk is fundamentally opaque. We know that. Bankers aren’t supposed to be heroes. Both common sense and the minimax ethic demand a margin of safety. We must call a halt well before we reach the perceived risk boundary – which is to say, before we think we have to. That’s what I meant when I said successful banks could not aim to be profit maximizers. Survival periodically requires them to forgo what look like marvelous opportunities. That’s idea number three.

Someone famously asked one of the Rothschilds how they’d gotten so rich. “Selling too soon,” the Rothschild replied. The most important thing a bank board can talk about is whether it is fooling itself about risk, and how soon – or if you will, how early – to stop taking on more. One reason for having a board with a diversity of perspectives is the hope that, as the champagne is being called for, at least some of the directors will feel a chill.

The Rothschild’s answer combined shrewdness and humility. Bankers need both. Deciding when to be cautious and when to be bold calls for nitty-gritty understanding of cash flows and markets and relevant political dynamics, but it also hinges on remembering the perils of the new and the distorted perceptions that flow from a run of good luck. Selling early is a test of character.

If there are six hundred people on a ferry and a school of porpoises appears and everyone rushes to the starboard side to see them, the ferry may capsize. Some people, knowing there are porpoises in the harbor, having gotten a sense of how the ferry handles the waves, and observing the size of the crowd on deck, will get off at the next opportunity. Such individuals make good bankers. There are not enough of them.

Banks being necessary, governments have responded to the scarcity of natural clairvoyance by issuing regulations and appointing supervisors. This has helped. Banks do keep finding ways to fail, however. Non-bankers attending this conference – or reading the Financial Review on a regular basis – might conclude that perfecting the rules is the answer to it all. They would be wrong.

Recently, there has been increasing emphasis on governance, which an American friend of mine – himself a former regulator – describes as “outsourcing supervision to boards.” There is more to the focus on governance than buck- passing, but it is not a panacea.

There is permanent debate about what supervisory duties can be outsourced to the market. Some observers take the view that capsizing ferries are part of nature’s way. “Pity about the passengers,” they say, “but you have to admit they brought it on themselves.” Personally, I do not find this approach appealing.

Another concept getting increased attention is risk culture. Bankers have always known it matters. Regulators, who traditionally focused on rules, are now committed fans. Paragraph 13b of APRA’s CPS 220 makes stewardship of a bank’s risk culture an explicit part of a board’s job description.

Without question, a strong risk culture is an asset. It helps us see risks and gives us a language for discussing them. It sets the tone. It fills the space where neither regulations nor internal policies provide explicit guidance. It embeds behaviors that defend the bank – a tradition of speaking up, for example, which helps guard against rogue trading and fraud, or something as simple as a habit of informally consulting colleagues before making significant decisions. Risk culture is like an invisible friend who tugs at your sleeve when you are about to be stupid. Some of the institutions made notorious by the GFC seem never to have acquired such a companion.

I’ve been employed by half a dozen banks and advised quite a few others. The differences among them were profound. Their weaknesses were persistent. Many people believe that repairing a weak risk culture is the work of a decade. I disagree. Risk culture can be changed for the better much faster than that, provided the project gets priority and the right approach is taken. This subject deserves fuller treatment than I can give it here.

Let me conclude with a personal perspective. What our stakeholders want is systemic stability, an adequate supply of credit, fair treatment and a reliable dividend that increases over time. While regulation and supervision, strong governance and risk culture all help us achieve those outcomes; I believe our profession has an important role to play. By “profession” I mean the worldwide company of bankers and the body of knowledge they possess.

Delivering for our stakeholders requires our acceptance of the self-governing principles discussed earlier: fiduciary responsibility, a minimax strategy and caution when we feel most confident. Successful risk management depends on our recognizing the difficulty of modeling extreme stress, the riskiness – as well as the importance – of new initiatives, and the Minsky phenomenon. These are all mildly sophisticated concepts, which a self-conscious profession is best equipped to articulate and sponsor.

Every decent bank has a sprinkling of individuals who are respected as much for their character and judgment as for their place in the hierarchy. Smart banks treasure them. They are guardians of professional standards.

I like to focus on our profession – or craft or discipline – because doing so gives our thinking an active bias and makes us all accountable. You practice a profession. You master a craft. But you are influenced by culture. There is luck as well as merit in how high you rise in an organization, but anyone prepared to work hard can achieve professional excellence. And while a bank’s risk culture is important, it has whatever quality it has because of the behavior of the individuals who work there.

The French have a phrase I’m fond of: “deformation professional.” I cannot speak that lovely language, but I believe what this refers to is the way a person’s daily tasks and role in society stamp her or him with recognizable mannerisms. You can identify good bankers by their quirks. Let me give you a few snapshots.

There is the sixty-year-old chief credit officer in his corner office. (I’ve made these people up, you understand, but they are not unrepresentative.) He’s a natural contrarian who wears a bowtie because most men won’t. He spots mistakes in financial forecasts without apparent effort and is unforgiving if he’s lied to. Presenting to him is a rite of passage for young bankers. Senior colleagues seek his advice.

Then there’s the youthful head of market risk. She runs five kilometers every morning and has a doctorate in math. Her office white boards are covered with graphs and Greek letters. On the wall she has hung a framed sampler with her motto: “Simplicity.” Her feisty grandmother embroidered it for her.

And not to be forgotten, there are neighborhood bankers all over the globe finding ways to finance their customers’ dreams, but when possible persuading them to save a bit more before launching their business or buying their first home.

Ask any of these people how they learned their trade. “Making mistakes,” will often be the answer. They might also mention a boss or colleague who taught them new techniques and traditional values. Viewed as a discipline, banking is a permanent apprenticeship.

I was explaining all this to a friend in another line of work a few weeks ago and he stopped me. “You have a romantic view of banking, don’t you?” he said. I confess that I do.

Banking is a demanding profession. Being good at it means mastering yourself – a capacity that is both a virtue and a skill. You must balance prudence and ambition – your customers’ and your own. Banks have a public purpose, which gives dignity to the work. If profit is your only guide, the markets will destroy you. Good bankers have a characteristic style – a way of going about things that combines calculation, skepticism and a sense of obligation. We honor our craft, embrace our responsibilities and aspire to grow in wisdom. We should take pride in that.

Thank you for listening.

Delivered 28 April 2015 at the Australian Financial Review Banking & Wealth Summit in Sydney.

Thinking About Banking – Harrison Young.

Leave a Reply

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