One of the unresolved issues of the 2008–09 global financial crisis and of the Great Recession was what viable substitute there could or would be for the credit-ratings agencies. As a former investment practitioner, I have been looking for firms trying to provide an answer to my question: What new ways are there for performing better credit quality assessments?
As the former head of Global Mergers and Acquisitions for Citibank and founder of two successful investment banking and financial firms, Kamal Mustafa, along with his partner of many years Leonard DeRoma, has a new business called Invictus Consulting Group that has a unique solution to credit quality assessment.
CFA Institute: Could you please describe what Invictus Consulting Group does?
Kamal Mustafa: We are a specialty financial adviser, as well as risk manager, focused on the banking industry. Our customers are of several varieties: banks, banking regulators, investors, and D&O [directors and officers] insurance underwriters, among others.
Our primary product is the Invictus Capital Assessment Model, which we call ICAM. To our knowledge this is the only forward-looking stress-testing methodology able to make predictions about a bank’s post-stress Tier I capital position, as well as other important performance metrics.
When you say “forward-looking,” over what time frame are you talking about?
The intermediate term.
If I understand you correctly, you use sophisticated computer models fed with tremendous amounts of data to execute scenario planning for financial institutions. Do I have that right?
How is this different from how credit risk is currently assessed?
We take all of the possible spooky actions into account to create a sliding scale of capital quality. This, by the way, is in harmony with how banks actually operate. So the ICAM looks at loans and breaks them down into classification tables in terms of asset quality and then looks at how those loans might perform under stress.
Currently the way investors, ratings agencies, and hedge funds look at these things is to compare banks side to side that have similar capital ratios and earnings, and then they analyze the financials based on these “single moment in time” figures. Under normal, steady state assumptions that would work well in most cases. But now — post the Great Recession — it is completely wrong.
When banks make loans they are like a stack of pancakes, with the amount of the loan of a particular type being the thickness of the pancake and the length of the loan as the breadth of the pancake. When you look at banks using current methods, it is very similar to cutting a vertical slice down through the stack of pancakes. But unfortunately, when you do this, you miss any variation in the thickness of an individual pancake over its breadth. In other words, the current methods just provide a “single moment in time” snapshot of the portfolio. Over time each set of loans has its own unique set of characteristics including origination date, term, amortization, interest rates, collateralization, among others.
Another big difference in how we look at credit risk is that we explicitly factor in maturity structure. Loan maturity structure has a real impact on the volatility of earnings of banks.
What was the germination for Invictus? That is, what made you want to invest an important portion of your life into it?
It was a series of accidents and timing. Six months before the recession hit, I was approached to establish a bank consultancy. So I looked at the technologies available as background before considering doing such a thing. What I discovered was that much of the product innovation at banks in the 1990s was in liabilities management, rather than deposit or asset growth. So banks did not have in place any asset management technologies. Consequently both Basel I and Basel III say that the amount of capital a bank must hold is based on a steady state assumption. By the way, the mathematical assumptions of these standards favor the big banks.
Our approach, the concept of stress testing, recognizes that the Basel system already put in place is useless. So there seemed to be a large gap in bank risk management, which is why we [Invictus] felt a system was needed to focus on assets and, consequently, how those assets would perform under various stress scenarios, which we began creating.
My experience as a portfolio manager is that often times companies do not like to be told that their laundry is dirty. What has the response been to your new, seemingly more transparent and predictive approach to credit-risk assessment?
Well you are right, and it took us some time to figure that organizations don’t like their dirty laundry revealed. So the going was tough at first.
But then we got our first break when the FDIC [Federal Deposit Insurance Corporation] heard about us and our approach. That was quickly followed by interest from the CSBS [Conference of State Bank Supervisors — the 50 state bank regulators organization]. Both regulators approached us to give them several educational seminars.
Invictus demonstrated to them how we approach the problem. As a matter of demonstration, we took the banks they regulate and ran them through our system using generic trends. Eventually we ran all 7,500 U.S. banks, stress tested them, and put out detailed reports with a ranking of each of the banks. Then we did an additional round of seminars with the regulators. They subsequently encouraged us to approach the banks with our product.
Banks were very interested in stress tests. However, when the tests were run on the banks, their executive teams and counsel wanted Invictus to burn the reports! In other words, there wasn’t exactly an avalanche of orders from the banks. But that gave us an opportunity to improve.
So we spent a lot of time with regulators learning how to tweak our system. Then in 2010 the regulators came to us and started talking about the fact that stress tests would be a part of future regulatory structure. Eventually bank boards of directors began approaching us to use our model to help with strategic planning.
Now we are finding banks being receptive because we showed them how to communicate more effectively with bank regulators in that capital quality needed to be viewed in three ways with the overall theme that capital quality could not be looked at monolithically and in stasis. First, we showed them how to help regulators recognize that there were loans made before the recession; and second, that there were loans made during the recession whose yield spreads and other characteristics did not have the same pre-recession mistakes built into them. Last, we showed them that a stress test with a two-year time horizon could be used to show regulators how the banks planned to change their loan portfolio to deal with the stress test results. This has helped our customers to have a better dialogue with regulators.
Who, or what, do you consider to be Invictus’ biggest competition?
There are competitors, such as Trepp, that have come out with stress tests that are not quantified and that use guesses to identify weaknesses in banks. In other words, their approach is not as consistent as is Invictus’ approach.
Many potential competitors have approached us. Two of the major ratings agencies, for example, have spoken with us about acquiring us. But we are not interested in being acquired at this time and feel that we have a very competitive product.
We have a massive database in place of the banks and their loan portfolios that we spent the recession creating that would be very difficult to replicate. Furthermore, our ICAM model has undergone many rounds of verification and validation that would also be very difficult to replicate.
Speaking of which, how do you measure the strength of your assumptions, and ultimately the quality of your model?
In running our reports for regulators each quarter we compare our predictions relative to actual outcomes. This allows us to fine tune our results and has made our technology even better.
How would you go about improving banking regulation?
After the recession ensued, all of the financial protections collapsed, but without any direct effect on bank loan portfolios. Then the actual impact of the recession hit banks’ balance sheets. Finally, the third phase occurred when bad loans made during the recession started affecting capital ratios. Depending on when you looked at the banks they would have gone from barely affected to near collapse.
Basel III is a joke because it has refused to acknowledge that its mathematical system to determine capital adequacy on the credit side of things relies on data that has disappeared. To deal with this, they are changing the definition of what capital is rather than looking at the root cause of the problem: bank customers are in trouble so the bank assets are in trouble.
But, because of the arbitrary rules of Basel III, you are taking financial institutions and forcing them to have higher capital ratios than are necessary — which means that banks stop making loans. The ultimate problem is that it is focusing on a short term Band-Aid approach rather than the key issues — i.e., bank customers are getting creamed.
Instead, a longer-term, more flexible, scenario approach is needed in order to provide both proper assurance of appropriate risk management, but also adequate return for banks and financial system liquidity.
Thank you very much for your time today Kamal, and good luck to you and Invictus.