Asset and liability management (ALM) is the practice of managing the risks arising from mismatches between assets and liabilities on a financial institution’s balance sheet, while earning a return commensurate with the risks.
Asset and liability management (ALM) is the practice of managing the risks arising from mismatches between assets and liabilities on a financial institution’s balance sheet, while earning a return commensurate with the risks. ALM is also administering policies that address the financial risks posed by changing interest rates. Financial institutions that optimize the risk/return relationship are deeply invested in the ALM process because it involves understanding and modeling every financial asset and liability of the organization. An effective ALM modeling process clearly identifies a financial institution’s opportunities and risk. Armed with a deep understanding of its ALM position, management can chart and execute a successful future path for the organization. Financial institutions that want to develop a successful business model over the long term have ALM processes that include:
- Selecting an ALM model that is consistent with the organization’s complexity and risks;
- Developing robust input assumptions
- Back testing the assumptions and updating them as necessary
- Engaging talented professional(s) to run the model; and
- Controlling and overseeing the ALM process in order to optimize the risk/return relationship for future business activities.
Wilary Winn believes that many organizations do not have adequate ALM processes and procedures in place to measure and mitigate ALM risk. Signs that a financial institution has an inadequate ALM solution include:
- It cannot efficiently and effectively run requested scenarios or stress tests using its internal ALM software
- It must run global credit losses because the ALM model being used does not support the input of default and loss severity assumptions at the loan or cohort level
- It uses loan and deposit input assumptions that are overly simplistic
- The analyst in charge of running the ALM model continually provides reports with mistakes or results contrary to what common sense would indicate
- The external ALM provider uses the same “cookie cutter” approach for all of its ALM clients
- The external ALM provider’s compensation for the other activities it provides the institution overshadows the compensation it receives for the ALM advice resulting in a potential conflict of interest. Examples include commissions received for buying and selling securities, correspondent banking fees, etc.
- Its asset and liability management committee (ALCO) is focused on addressing regulatory concerns at the minimum cost versus using the ALM modeling results to effectively manage the organization
If any of the above items describes your organization’s current ALM, we strongly urge you to consider alternatives as your organization has settled for an inadequate ALM solution.
Model Selection
The ALM model itself is a critical component of an effective ALM process. Wilary Winn licenses the ZM Desk model. We selected it because using it we can:
- Model credit losses at the loan cohort or individual loan level
- Perform both static and dynamic analyses
- Model on both a stochastic and deterministic basis
Wilary Winn believes a financial institution’s ALM risk measurement and management processes can be significantly improved by incorporating credit risk on a bottom up approach versus applying a global ALLL assumption or importing credit losses from another model. Incorporating losses obviously improves the measurement of interest rate risk because only interest income that is expected to be received is included in the forecasts. More importantly, incorporating estimated credit losses into the forecasts results in better measurements of overall net income and capital. While nearly all ALM models in the marketplace today allow for prepayment rates as an input, very few support the use of conditional default rate and loss severity assumptions to model credit losses at the loan or cohort level. We further believe while the ability to model credit losses is important now, it will be critically important going forward. The banking regulators are focusing on concentration risk and capital stress testing in addition to interest rate risk. The largest institutions are now being required to perform rigorous stress tests on capital adequacy and we believe these requirements will find their way into exams of smaller institutions. In addition, FASB’s issuance of the Current Expected Credit Loss (CECL) model is imminent. Institutions than can model credit risk can plan ahead and be fully compliant when required.
ALM model selection is therefore critical. We advise financial institutions not to sign a long term contract with an ALM provider or for ALM software that lacks the capability to perform detailed loan loss calculations. If a financial institution’s ALM model cannot calculate lifetime losses on individual loans or loan cohorts, the cases to be made for capital adequacy and the CECL calculations are weaker than they could otherwise be.
Model Inputs
Developing robust ALM inputs is the next step in a successful ALM process. We have spent years developing and back testing our ALM inputs. Our model inputs are informed by the other services we offer. For example, we perform mortgage servicing rights valuations for 200 clients located across the country providing us with a broad and comprehensive perspective regarding prepayments. As another example, we have developed credit loss assumptions on billions of dollars of residential real estate and consumer loans through our fair value business line. We have formed static pools using the information our clients provided and we continually back test and adjust our input assumptions based on our clients’ actual experience and changes in existing and forecasted economic conditions. While some of our competitors also license ZM Desk – we believe we are years ahead in the development of input assumptions. Of further note, while we believe it is possible for a financial institution to develop input assumptions using only its data, we caution that the sample sizes may not be large enough to be statistically valid.
Staffing
Even if a financial institution licenses a sophisticated ALM model such as ZM Desk and works to develop robust input assumptions, it still needs both a talented professional(s) to apply and adjust the assumptions and run the model, as well as an ALCO that has both control and oversight over the ALM process. Think of auto racing as an example. Even if a team has the fastest car, it still needs both a talented, well-trained driver and pit crew to win the race. With ALM, the best model is essentially useless if assumptions are overly simplistic, incorrectly applied or the ALM process lacks oversight.
Wilary Winn analysts are intelligent, highly trained financial professionals. They have years of experience in developing and applying assumptions (repayment, default, loss severity, decay, beta and effective maturity) and analyzing ALM results. This includes projecting lifetime losses based upon loan attributes. Our experienced financial analysts use numbers to tell your institution’s story. The end of the story is the presentation of options and opportunities related to the ALM position of the organization. Every financial institution is different with respect to its business model and the way in which it views risk. Understanding the uniqueness of the financial institution is a key element to providing customized ALM recommendations and optimizing the risk/return relationship.
ALCO
Wilary Winn believes that an empowered ALCO is a critical component of an effective ALM process. Ideally, ALCO determines the what-if and stress scenarios in a proactive way as opposed to reacting to requests from its regulators. ALCO’s role also is to provide control and oversight to the ALM process. Are the ALM assumptions reasonable? What are the results of assumption back-testing? What are the impacts of new business strategies on the ALM position? Should derivatives be used to bring the ALM position back into compliance? Furthermore, an empowered ALCO ensures ALM policies are detailed and reflect the operating strategies of running the financial institution. Ideally, the policies cover management of interest rate risk, liquidity, investments, concentration risk and contingency funding in sufficient detail as to provide a road map should something unforeseen occur.
There is no shortage of outsourced ALM solutions in the marketplace. The quality of the work varies considerably in comparing external ALM vendors. At the one extreme, there are very simplistic and inexpensive cookie cutter ALM solutions which are run based on call report data. This simplistic approach can work if a financial institution has a very simple balance sheet with low ALM risk. At the other extreme, financial institutions with more complex balance sheets and elevated risks can benefit greatly from vendors that are able to provide extremely detailed ALM calculations and customized advice. Given the importance of ALM to financial institutions, we believe that financial institutions that outsource strongly consider the motivation and business philosophy of its external provider when determining whether or not it has an adequate ALM process in place. For example, if an ALM provider’s primary interest is to sell securities and it is paid substantially more to do so, then how unbiased and credible is it advice? If the advice is self-serving – sell these securities and buy these securities – then we have found that the outsourced ALM is often loss leader for the provider. The goal here is to spend the least amount of time on ALM and maximize the time spent on sale of securities. As another example, how effective is the ALM advice if a vendor applies the same non-maturity input assumptions to every client regardless of each financial institution’s business practices. If your ALM provider is using a cookie cutter approach to ALM reporting, they are likely more interested in cutting their labor costs than providing good ALM service.
Wilary Winn provides independent, fee based ALM advice. We do not sell securities or earn commissions. That said, for those interested, we can provide the names of securities dealers that are experts in finding the “sweet spot” on the yield curve and constructing cash flow waterfall strategies that can be helpful in both earning yield and managing liquidity. Smart, long term investing is a key element in successful financial institutions.
Wilary Winn provides customized reporting solutions and recommendations for our ALM clients. We also provide clients with risk-based pricing analytics, concentration risk assessments and detailed capital stress testing.
Wilary Winn follows these best practices with respect to its ALM modeling:
- Explicit calculation of credit losses from the bottom up using robust and fully back tested input assumptions
- Use of sophisticated input assumptions for prepayment
- Use of accurate structured models for CMOs (Agency and private label)
- Use of level yield calculations for premium expense amortization
- Use of detailed and supportable assumptions for non-maturity deposit behavior
- Use of forward curves for discounting and reinvestment
- Use of higher discount rates for credit impaired or delinquent loans
- Inclusion of stress scenarios and what-if analyses
- Evolvement to stochastic analyses on mortgage loans and investments for those with complex balance sheets and elevated ALM risk
Wilary Winn follows these best practices with respect to managing ALM positions:
- Recommending the construction and management of cash flow ladders
- Helping our client understanding the risk/return tradeoffs
- Base case vs. rate change
- Short term vs. long term
- Calculating risk based pricing including requiring a return for estimated credit losses
- Monitoring the duration variance between assets and liabilities
- Improving policies to provide early warnings of undue risk
- Understanding current and proposed regulations on our clients’ business models
- Evaluating derivative and non-derivative strategies for managing interest rate risk
We urge financial institutions not to settle for an inadequate ALM solution. If your organization is interested in developing a deeper understanding of its ALM position and optimizing its risk/return relationship, then please call our office at 651-224-1200 to discuss our ALM solutions in more detail.