My brief task today is to introduce our guest blogger Brian Ranson. Brian has been a respected credit risk management expert and pioneer in the use of quantitative default probabilities instead of legacy agency ratings. After a long career at the Bank of Montreal, he then spent many years as advisor to sophisticated financial institutions at one of the major rating agencies. Read on, and enjoy!
Donald R. van Deventer
March 17, 2010
Brian Ranson on “If you don’t know where you are going……………”
As Lewis Carroll wrote, if you don’t know where you are going, any road will get you there. As any risk manager should readily admit, there is no guarantee of where risk is going, but that does not mean that there is no way to plan for risk. Indeed most banks plan risk changes every year, however they do so with blissful ignorance and call it a business plan. There is a way to change this and it can be done without having expensive software or technical experts. That does not mean that expensive software is an unnecessary expense, rather it means that there is an easier way to get to the point where that software (if you have it) can be used to its fullest extent.
The business plan
Toward the end of each financial year, a battle begins in virtually every bank between the CEO and the operating units. The CEO demands a substantial profit increase and the operating units try to reduce their targets so that they have the best chance of reaching or exceeding the plan, thereby gaining credit as well as bonus opportunities. Very often the CEO holds firm to his number, which means that the business plan becomes a negotiation between the operating units to try and transfer part of your target to another unit. With many banks having a host of cross charges between departments as well as “head office” expenses the debate rages hotly about nothing more than rearranging internal costs. Not too productive.
What about risk?
At one point I recall asking the CEO in the height of the business plan negotiations, “how much more risk would you like to take on?” The answer was fairly predictable, “as little as possible to reach the target” which deflected the issue back to the operating units without any real guidance. Based on that experience I would suggest that asking the CEO for a risk target is unlikely to be a successful approach. Instead every risk manager should take on the responsibility of showing the potential risk outcomes as the business plan is developed. This may well be possible without too much stress and perhaps even with the full cooperation of the line managers/sales staff.
Corporate banking and the loan pricing cycle
Taking corporate banking as the test case (by which I mean the portfolio of larger corporate and government clients), each relationship manager (“RM”) (or the equivalent) should have a limited number of clients with loan facilities. Each of these facilities has a maturity date, pricing, a risk rating and other data. Some of this is well recorded centrally, some is artfully hidden in desk drawers, but the relationship manager should have all these facts at his or her disposal.
Once a list is prepared the manager should know which facilities will mature in the following year. Most will be renewed; however there is a high probability that the price will change. It may change because the risk has changed, it is more likely to change because market pricing has changed, but any change means that the revenue from the facility will change. If the credit cycle is such that the market prices have gone up then the manager may get near his business plan target by doing very little (a great way to get a bonus). If the credit cycle is such that prices are falling then he or she is faced with a big problem for to achieve a revenue increase a lot of new business needs to be found. The problem is especially acute as approximately one third of a corporate loan portfolio will typically reprice in any one year.
Assuming that the RM can use the current data to estimate repricing levels, he or she will have an estimate of revenue for the following year from the existing portfolio. At this point I am assuming that the usage of facilities will remain constant but that too can be factored in based on current market expectations. At that point the business plan demands can be considered in the light of the revenue demands and the market.
If pricing is static then a 10% increase in revenue will require an increase in lending facilities of 10% if and only if the current risk mix remains constant. That is a low probability for the mix is generally not within the manager’s control when seeking new business. If a bonus is the RM’s carrot for reaching the revenue targets then the least difficult way to increase the revenue will be to find higher risk clients or to lend more to the higher risk clients in the portfolio, for they pay more. In other words the path of least resistance to meeting a business plan is usually to increase the average risk.
This is not necessarily a bad thing, however it is a real thing, and the CEO needs to understand that the business plan is one that can move the portfolio in this direction for this may be taking place when the strategic imperative might be to hold or reduce risk.
The CEO and the Board
So much more can be added, but at this point consider the simple question “do we want credit risk in our portfolio to rise in the coming year?” An automatic answer; “of course not” might be very poor strategy for the lag between the default risk cycle and the credit pricing cycle is such that there are certainly times when the price for risk is favorable. Moreover, there may be good reasons why a portfolio might seek a larger number of smaller but somewhat riskier clients because by doing so diversification may be enhanced.
The answer therefore is far from simple yet the question is very powerful. If left to their own devices, by which I mean a revenue target with no risk target and no consideration of the existing portfolio profile, RMs will move to the line of least bonus resistance. Whether they add business that improves the bank’s use of capital is simply a matter of fortune rather than planning. In addition the potential for severe embarrassment once a lot of poor deals are on the books (for term) will be high.
In sum the risk manager should not only be part of the business plan process but indeed should be empowered by the CEO and the Board to set the risk parameters before the revenue plan is determined. The benefits would be very substantial:
- The risk plan can be viewed along with the business plan as the year moves on so that trade‐offs can be determined using both risk and return data.
- The RMs would have a much clearer direction as to the nature of the business that is being sought.
- The critically important bonus plans would reward good deals rather than good (and potentially short term) revenue gains. In addition bonus payments would not be made for simply repricing the maturing part of the current book.
- The business planning process could target business mix changes on the best risk/return businesses potentially reducing the internal bickering over allocations
- The process would allow the risk managers to look at the RMs portfolios and the repricing intentions to better balance pricing and structuring among like clients. The opportunity for consistency may appear in a way that had not been seen before.
- The data gathered should be of great value when using portfolio models and other credit risk management software.
At the end of this however the best result will be that there will be a better understanding of the credit risk profile and a road that will lead to a chosen rather than a random place.
copyright (c) Brian Ranson, March 17, 2010