Why the Art of Risk Management Prevails over the Science

EXPERIENCE AND THE HUMAN DIMENSION OF RISK

One could reasonably ask why so much emphasis is placed on experience in risk estimation, modeling and management. While experience benefits people in any line of work, there is a good reason why the introduction of new technology, new sets of regulatory rules, or shinier forms of process automation do not render the experienced risk practitioner obsolete. Most of risk, especially financial risk, is dominated by the human dimension.

Hurricanes happen, and fires and floods, but most of the damage inside and outside financial institutions is dominantly affected by the human dimension. Experience is the greatest teacher of how to incorporate that in one’s approach to risk.

To take the example of boundary conditions: there are many financial situations in which the rules create a boundary condition between acceptable and unacceptable, whether it is credit score or bond rating or a covenant ratio in a loan agreement or a rule for crediting a transaction as sales revenue. Humanly speaking, you can count on any situation where 6.01% is acceptable—but 5.95% is not—to attract attention. Some people will bend the input, create adjustments, do whatever to push the ratio to the acceptable side of the boundary. The tracks they leave, in terms of distorted measurements, and therefore distorted views of the risk involved, generate all kinds of incremental risk and eventual loss.

Similarly, wherever there are strong incentives to do something financial, such as sales bonuses, etc., there are people who will figure out how to get the deal done, worthy or not. The pattern of absurd mortgage loan approvals that led up the financial crises of 2008-2009 is a Matterhorn-sized case in point.

In a related phenomenon, observed human nature demonstrates over and over again that the speculative among us do not self-regulate. The profits of chasing a fashionable investment drive people and markets to unsustainable price levels—bubbles—then they pop and the bad things that risk analysis tries to estimate all happen in quantity.

A risk professional learns to understand that our shared humanity affects how regulators regulate, how outside accountants audit, how employees respond to an indifferent managerial climate.

On the regulatory side, for example, there is an understandable desire to have a good picture of what a bank’s risk is. Historical cost figures on longer-term loans and assets say very little about what they are worth this month. That has led to decades of push toward mark-to-market of balance sheet assets, and where assets are just too illiquid for a market quote, mark-to-model. There have been whole decades of pushing banks toward more mark-to-market.

But the same regulators find there is twin trouble in mark-to-market and mark-to-model in the banks they regulate. First, market prices don’t hold still, they sometimes fluctuate strongly. Probabilities of default for wholesale obligors fluctuate even stronger still.

The same BBB borrower may have an honest one-year default probability of under 6 basis points (0.06%) in good times, and an equally honest probability of 49 basis points (0.49%) in really bad times, implying risk of loss eight times as big. Most financial ratio tests and capital requirements baked into banking law and regulation are fixed numbers.

If a risk-conscious capital measure goes from 100 to 800 in the course of five or six years, what does that do to a by-the-numbers regulator? It is a bad situation in truth, and appears even worse to onlookers and critics.

So what happens over time?

• Rigid, risk-insensitive capital measures drive profitable, riskier business off banks’ books to non-bank lenders, and excesses erupt in unregulated parts of the economy. So, regulators move to more risk-sensitive measures to keep regulated institutions viable and central.
• Then after a while, the risk environment worsens and risk-sensitive capital measures soar, and an industry suddenly looks terribly under-capitalized. Healthy, solvent institutions suddenly show huge, unrealized mark-to-market and mark-to-model losses, and public confidence is shaken. In the worst case, financial crisis erupts and institutions begin to fail in numbers.
• To prevent another embarrassment, regulators (and legislators) swing back to conservative, risk-insensitive measures, then the cycle repeats.

Accounting and historical measures are objective and verifiable; risk measures are subjective and analytical—reasonable experts can differ. You can be criticized for taking risk-insensitive accounting numbers, but risk measures demand a level of good faith and confidence in subjective analysis. And oh, the criticism and second-guessing that assail a regulator who is so trusting!

The human condition lands bank executives, regulators, and the public in an unforgiving quandary. To quote T.S. Elliott, “Human kind cannot bear very much reality.”

An experienced risk professional understands all this, which is why humility is such a prized virtue.

Contributed by: Phil Chamberlain, 35+ years experience leading risk at global financial institutions

For more of his insights on Youtube:

https://www.youtube.com/channel/UCGjpiuf0rRtnIyRAdhUtr1w/featured