By Gumersindo Ruiz. Malaga University & Ramon Trias. AIS, Aplicaciones de Inteligencia Artificial
Article originally published in FUNCAS “Cuadernos de Información Económica” (Funcas Economic Information Journal) nº 230 September-October 2012, redrafted in accordance with Act 9/2012.
This article is divided into three sections. The first section explains the need for methodological change in financial regulation and supervision, going beyond simply fulfilling balance sheet ratios and formal operating procedures towards an approach in which the entity’s strategic management plan plays a leading role. The second section proposes a dynamic methodology which aims to project the balance sheet in accordance with possible scenarios, and optimises decision-making on risk and profitability, subject to capital and liquidity restrictions. The third section summarises the process using an example, which leads to an optimum decision on the sale of portfolio.
It must be remembered that this process occurs during the development of Basel III regulations, with a sophisticated methodology for the assessment and weighing of risks overlying a simpler one with stronger capital requirements. As a result, in Spain we are subject to both Basel III regulations and government requirements as established in royal decrees and laws, along with a new orientation towards knowledge of institutional viability, which involves a future projection of the balance sheet, as set out in Act 9/2012, of 14th November, and in Royal Decree 1559/2012, of 15th November, which compiles all regulations in this regard.
We are going to use the new law, Act 9/2012 of 14th of November, which requires financial entities to, in specific circumstances, present a viability plan, in order to justify the methodological change of negotiation. The wording of this Act makes clear the need to develop a strategic plan which can be shown to be viable, to which end we require a decision-making system which envisages capacity to generate resources subject to a series of capital and liquidity restrictions, limiting risks, considering the initial business conditions or microeconomic model and the prevailing macroeconomic environment. We believe that this approach is valid for all entities and not only those in a precarious situation, as set out by law, since the methodology can apply to any entity in order to project its balance sheet in the current environment of volatility and uncertainty.
1. References in Act 9/2012, of 14th November, on the need for a viability plan for the restructuring of financial entities and methodological implications.
Restructuring cases are defined as transient weaknesses of credit entities, which must be identified in advance and require public support. The preamble mentions the recovery of the public funds applied to restructuring through the profits generated by the entity.
It is clearly stated that the fundamental criterion when implementing a restructuring or dissolution (liquidation) process of an entity is its financial viability. Both cases envisage the drafting of a plan to be approved by the Bank of Spain, which will be used to establish the necessary instruments.
Article 5, paragraph 2 refers to determining the economic value of the entity, its assets and liabilities, and is a static view of the balance sheet. However, point 3 makes clear that the assessment will be based on the entity’s economic and financial projections (which will be dealt with in the FROB). This projection is important since it means consideration must be given not only to the situation of the entity at the moment of analysis, but also to its capacity to generate profits.
Article 6 sets out the conditions for restructuring, if there are objective elements or whether it is expected that the entity may not fulfil the requirements of solvency, liquidity, organisation or internal control. This would require an action plan as developed in the following articles. Article 7 states that this plan must set out the actions envisaged to ensure the long term viability of the entity, to be implemented with immediate effect.
One of the ideas we can anticipate is that entities should have updated strategic plans which allow them to act swiftly in the event of alerts issued by the Bank of Spain (objective elements, predicted impairment), as set out in Article 6.
Article 8, 1. a) states that the entity, in addition to analysis of its situation, must present a business plan which includes specific objectives relating to: efficiency, profitability, leverage and liquidity. This is perhaps the area which deals most specifically with business strategy, since the solvency commitments in paragraph b) must be fulfilled by achieving a series of objectives; the organisation and management improvement objectives (section c) are explicitly developed in the methodology proposed below.
The criterion set out in paragraphs 2 and 3 of Article 8 is hugely important, since it envisages the possibility of modifying the plan in the event of a downturn in market conditions or negative evolution of the entity. The Bank of Spain may regulate these procedures, but what is clear is the need for a methodology which can intelligently present the best options for managers by introducing parameterised data.
Both the entity and the supervisor implicitly require a flexible solution which can continually answer all questions arising during the restructuring process. Article 11 therefore refers to monitoring the action plan and FROB information. However, given the current complexity of the information handled and the uncertainty involved in balance sheet projections, it is more than reasonable to propose a methodology which has been specifically designed to permanently provide this information.
Chapter III is devoted entirely to “Restructuring”, whilst Article 16 reinforces the dynamic view of the problem, since it is said that the restructuring plan will include analysis of the entity’s current situation and its future ability to pay back the financial aid received. There is also the possibility of an entity being saved even when the aid cannot be returned, given the impact it would otherwise have on the system. In any case, an optimisation criterion is introduced since it is said that the measures should minimise the use of public resources, along with optimal management of the capital instruments used.
Our interpretation is reinforced in Chapter 5, which deals with “Financial support instruments”. Any decision by the FROB on financial support involving public resources may face opposition from the Ministry of Taxation and Public Administration. The General Secretariat of the Treasury and Finance Policy and the General State Comptroller will issue their own reports in this regard. This makes it necessary to have a methodology which can intelligently answer the questions raised by the different public bodies involved with regards to the restructuring process.
Article 32.2 sets out another circumstance which makes a balance sheet projection tool necessary. The entity in question has to buy or amortise the instruments executed or acquired by the FROB within a maximum of five years. However, this five-year period will be subject to ongoing prospective monitoring, since both the Bank of Spain and the FROB must constantly scrutinise the situation in order to anticipate whether this repurchase or amortisation is in jeopardy.
There are other circumstances which require a dynamic view of the entity, such as Article 34.1 which envisages the possibility of the entity not being able to complete the repurchase or amortisation of the instruments made available by the FROB once the five-year period has passed, whether due to internal or external reasons. In this case, there is the possibility of a two-year extension, although failure to meet the initial target will mean the entity’s performance forecast, strategic decision-making and business plan will require stringent, flexible, ongoing monitoring.
Chapter 7 is dedicated to the management of hybrid capital and subordinated debt instruments. One of the characteristics of the present situation is the complexity of capital and liquidity compared to the more straightforward requirements which have been in place for many years. Article 39 deals with these instruments and the debt of entities subject to restructuring. The aim is to ensure an appropriate distribution of costs among investors, although, as is clear in paragraph 2 of Article 40, any modification of the terms of issue in accordance with the restructuring plan will make it necessary to choose between different alternatives, which will also affect the balance sheet. A system which can offer simulations and provide optimum choice will allow us to answer all these questions with regards to the restructuring process.
Article 44 sets out the criteria for determining which issues or groups of hybrid capital and subordinated debt instruments are subject to this management action. The deferral, suspension, elimination or modification of rights, obligations, terms and conditions of the issues of the entity respond to a range logic of the securities in accordance with their risk. This type of delicate operations, in which the FROB has to establish the repurchase prices for the entity, requires not only a market valuation approach, but also consideration of how the operation will affect the entity’s viability plan, which is the ultimate goal of these actions. The same applies to the powers of suspension of contracts and guarantees set out in Article 70. Moreover the FROB (Article 46) must justify its actions to the Bank of Spain.
The dynamic nature of all that considered is again apparent in Article 48, which establishes that any instability of the entity or of the environment may bring about a change in the management of the hybrid capital and subordinated debt instruments.
The general administrative powers of the FROB, as set out in Article 64, range from determining the economic value of the entity, through to capital increase or reduction operations and the issue and amortisation of securities; this Article makes clear the need to count on a tool which can, at all times, assess the impact of particular actions on the viability of the entity, since the FROB can decide on aspects which determine a banking strategy. This is particularly important in relation to the transfer of assets of the financial entity to the Bad Bank (Additional provision seven to nine).
Finally, transitional provision one of Royal Decree 2/2011, of 18th February, on a fulfilment strategy for capital requirements, which sets out the obligation of the entity to present a fulfilment strategy and schedule, is amended. The deadlines are so short that, in our opinion, the entity needs a strategic plan of the type considered above, allowing it to project its balance sheet and, in consequence, its solvency or requirements over a specific timeframe.
2. Optimal choice as a criterion
This changeover from a general financial equilibrium based on analysis and individual decisions, along with assessments by entity risk models, to a more interventionist, regulated system, has a theoretical basis, an empirical verification in the current crisis, and a review, in which we currently find ourselves.
In the 1950s the economist Kenneth Arrow gave mathematical consistency the idea that the integration of individual interests brings about the best application of resources in an economy, this being the start of the development of general equilibrium theory, laying the groundwork for neoclassical economic theory.
At practically the same time, Markowitz, Tobin and Sharp, as well as other authors of less renown, developed the idea of the operational translation of Arrow’s seminal idea to the efficient allocation of assets by agents on financial markets.
This description of the situation is not so clear since there are, as stated at the beginning, two coexisting currents: the sophistication of risk calculations and their weighting as incumbent upon the entities, and which remain valid in Basel III; and another trend towards simple yet rigorous capital rules. The latter has its clearest representation in Andy Haldane of the Bank of England, and his famous paper at the Jackson Hole meeting in August 2012: “The Dog and the Frisbee”, in which he recalled that dogs do not need any knowledge of physics in order to run after and catch a frisbee, and so, in the same way, it is more effective to establish stronger capital standards than sophisticated risk and weighting calculations. This line has been followed recently in Spain, where indiscriminate capital requirements have been established. In any case, we can say that a liberal view of regulation, in which the decision between risk and profitability is taken by the financial entity, has been replaced in the current context by an interventionist position. The conclusion is that it is necessary to formally incorporate the regulatory restriction in the decision-making process with regards to the entity’s balance sheet.
Asset allocation is the investment strategy which best balances the risk and return of a portfolio, since it provides the combination with greatest profitability for a specific risk. The term was originally applied to equity investments in Markowitz’s portfolio theory in 1952, since when, despite the significant constraints of information, market performance and volatility, it has been a fundamental part of all investment theories.
The problem can be extended to short selling, when the best situation for the investor is for securities to have a negative value, in order to buy at this lower value and thus deliver more securities at a given price. In 1971 David Pyle extended the concept of portfolio optimisation to all of a bank’s assets and liabilities, with liabilities being understood as short selling since their decreased price is the best situation for the financial entity. In addition to the work of Pyle, Pierce and Cohen, other authors developed an economic theory of banking, and the criteria for its practical application. Balternsperger’s article includes an analysis of the different approaches.
This concept, for both assets and liabilities, has been used for the decision-making methodology developed by AIS, incorporating regulatory restrictions (compared to spontaneous decisions on the market) and improving risk assessment through models with complex scenarios.
Specifically, the following is taken into account:
- Some restrictions in the form of fulfilment of capital and liquidity requirements.
- An optimisation function, subject to these restrictions, which should optimise performance for a given risk. The optimisation model responds to continuous information on the consequences of a decision (for example, the sale of an asset at a certain price) on capital and liquidity. Moreover, the effects on capital and liquidity are reintroduced as greater business possibilities.
- A future projection showing the viability of the business plan for a predicted environment, given the entity’s portfolio and a series of specific scenarios.
These three issues can be seen in the current context:
- Liquidity and capital requirements have become complex. Compared to traditional measures, liquidity is conditioned not only by the market and volatility characteristics of assets and liabilities, but also any mismatch between them and the time they become payable. Capital requirements also set out that capital classes, including FROB contributions, must be distinguished, meaning an intelligent methodology which is capable of dealing with this complexity is required.
- The optimisation function, with its restrictions, will show what type of strategy is required and which decisions should be taken with regards to the business and risk. This means an active role for the entity’s directors, who need to move forward with a new business model.
- The business model responds to a plan agreed to with the Bank of Spain and the FROB, and there is a consensus on its viability. In consequence, it is necessary to take into account the initial portfolios and the business conditions of the entity, and see how they are managed in a predicted environment for a series of possible scenarios.
In this context, we have tried to define a risk and performance optimisation function for complex capital and liquidity restrictions, integrating the business strategy into three areas: a macroeconomic environment model, a model which sets out the conditions of the entity, and a model of possible scenarios. The latter resembles the stress test, but, unlike a static balance sheet analysis and how possible future events can affect assets and liabilities, is based on complex modelling which aims to project the future balance sheet. It also allows the integration of geo-referenced business management and measurement of elasticity of demand (for example, for the sale of assets).
Chart 1 summarises the methodology for decision-making.
This methodology is guided by the criteria outlined, and aims to answer the questions below.
i. The goal is to design a platform which integrates the elements which support the strategic decisions of the banking business, identifying the best options in line with existing constraints. It calculates optimal feasible business flows and displays them in business plan format, in the same way as the projected balance sheet, income statement, cash flow statement and charting of the evolution of assets and liabilities.
ii. The current challenge can only be faced properly by integrating decision elements such as:
iii. The person handling the application may set different scenarios and the system will suggest different actions, such as the sale of real estate or mortgage loan portfolio in certain conditions. The scenarios should be relaxed whenever the results are not feasible.
- Simulation and optimisation of investment and financing decisions, including new loans, portfolio sales, FROB contribution (if any) and the maximum current value of the cash flow generated per unit of capital.
- Integrate macroeconomic model forecasts with external scenarios, in a more complex way than the projection of some macroeconomic variables.
- Assess the impact of the situations listed above on assets, liabilities and results, along with consequences on losses in loans, cost, operating and financial income and potential demand for credit.
- Assume the effect of market, regulatory or intervention restrictions such as credit ceilings, regulatory changes in capital, liquidity and leverage, as well as economic capital, and the effect of diversification or concentration of investments.
- Incorporate guidance in the form of restrictions, rules and desiderata, such as maximum exposure in negotiation, loans to SMEs, public administration and the ability to raise capital.
The system will take into account the limitations of the business, its dynamics and the evolution of external variables and their joint variability, and incorporates estimation of external scenario conditions, explores the feasibility of the restrictions and desiderata, identifies investments and funds to maximise the value of the entity, if a feasible solution exists, and defines the actions and their consequences in accountancy format.
This is shown graphically in Chart 2, in the form of operational modules.
3. A practical application, by way of example
Our aim in this example is to illustrate how to achieve optimal allocation of investments whilst respecting the balance between risk and return, adjusting the position of each element in terms of assets, liabilities and capital in accordance with capacity to absorb risk, starting point, objectives and time structure and any regulatory and market constraints.
This example is based on the projection obtained from a macroeconomic model integrated with the definition of external economic scenarios; the projected macroeconomic variables are used to estimate the main flows of the future of the financial entity —default, market— . This methodology integrates randomness, uncertainty and optimisation when projecting viability and business plans, along with strategic and tactical planning. The example shown uses criteria of the RDF method (Risk Dynamics into the Future), the econometric resources of R (“language and environment for statistical computing”), and the Strategic Advisor model.
This combination allows the analyst to try out future macroeconomic scenarios on a financial entity’s data: balance sheet and income statement. The calculation engine will search for the optimum situation within the existing regulatory and market constraints, with the future projection of several periods.
In the current environment of regulatory and macroeconomic uncertainty, it is extremely useful to have a system which can simulate changes in the regulations in different economic scenarios, thus indicating the entity’s needs in times of macroeconomic change and/or transition of regulatory requirements. By incorporating optimisation criteria in addition to statistical and econometric models, any changes in the decisions of the entities, however radical, are taken into account.
Shown below is an example of the methodology based on the data of a fictitious bank, A, in the 4 phases (see chart 2):
The problem can be extended to short selling, when the best situation for the investor is for securities to have a negative value, in order to buy at this lower value and thus deliver more securities at a given price. In 1971 David Pyle extended the concept of portfolio optimisation to all of a bank’s assets and liabilities, with liabilities being understood as short selling since their decreased price is the best situation for the financial entity. In addition to the work of Pyle, Pierce and Cohen, other authors developed an economic theory of banking, and the criteria for its practical application. Balternsperger’s article includes an analysis of the different approaches.
This concept, for both assets and liabilities, has been used for the decision-making methodology developed by AIS, incorporating regulatory restrictions (compared to spontaneous decisions on the market) and improving risk assessment through models with complex scenarios.
A. Preparation
Data is loaded from the last balance sheet: assets, liabilities, income and cash flows (see Figure 1):
It also loads the business parameters which describe the operation of each sub-portfolio of assets and liabilities, envisaging risk premium parameters, interest rates, commissions, operating costs, mean periods, % early payments, provisions, PD and LGD model parameters, % posted on the book value in the sale of portfolios, % irregular credit collected, etc.
Some of these parameters have been defined for this financial year, for example:
- Specific provision of 25% of risk of non-performing loan portfolio.
- Generic provisions: 30% for real estate project financing portfolio, 0.5% for mortgages and 1% for the rest.
- % posted on the book value in the sale of portfolios: 88% in mortgages, 40% in project financing.
- % posted on the book value of non-performing portfolios: 8% in mortgages, 6% in project financing.
- Last real data: fourth quarter of 2011.
- Outlook 6 quarters from now.
B. Simulation
The following macroeconomic scenario has been configured (see Figure 2), with the following characteristics:
Based on the pre-set values in the last quarter of 2012, 2013 and 2014 for four of the seven macroeconomic variables of the model, it has calculated the most probable “route” for the other variables in each period, conditional on complying with the pre-set scenarios, based on the prediction model.
Having defined the macro-scenario and the operating parameters of the entity, Basel III restrictions and the entity’s market restrictions are incorporated. The Basel III restrictions contain the ratio limitations in accordance with the proposed regulations; this example also tests for didactic purposes and, in order for the result to be more visible, brings forward the short-term liquidity limitation schedule, placing it a few quarters ahead of that proposed: second quarter of 2013, instead of the first quarter of 2015.
Accountancy, market and strategic guideline restrictions are added, with different levels of priority, to the macroeconomic situation and the restrictions deriving from the regulations. These restrictions configure the space where the optimisation engine works, the criterion of which is to maximise the value of the entity in line with the limitations. This value is interpreted as the current cash flow value of the periods projected plus net asset value, updated in the last period of the planning schedule. so-ansi-language: EN-GB’>From a basic scenario (macroeconomic model projections), the user sets the value of some macroeconomic variables in some future quarters; these values come from an extra-model scenario obtained from other estimates, expert opinions or an adverse external scenario. The “Conditioned Cases” chart shows the hypotheses chosen by the user.
The system calculates the quarterly projections with the macroeconomic model operating on the platform and conditions them in line with the user’s hypotheses. The “Macroeconomic Scenario” chart of Figure 2 shows the values which result from this calculation.
C. Implementation and D. Solution
For both scenarios the results in Figure 3 are obtained for the conditioned scenario, which seeks to maximise profits in the next 6 quarters. Optimisation takes into account all the models and revenue and cost parameters of each portfolio, along with all the restrictions imposed. It also takes into account the estimated credit losses based on the PD and LGD models.
The optimisation engine takes the decision to sell part of the portfolio, since the profitability that can be achieved with the current structure does not compensate for either the potential default or the capital consumption. Specifically, the system discovers that it is best to do away with part of the mortgage portfolio, allow Mortgage and Project Financing portfolios to mature, keep Consumer Loan at a minimum up to five quarters after the beginning of the financial year, increasing by a minimum in the following quarters, keep new Commercial Credit concessions half-running and maintain an average level in Credit Cards, apart from maximising income from non-financial transactions.
For financing, the algorithm obtains funds from the sale of portfolio and tries for longer term debts than those maintained in the initial balance sheet, since highly liquid assets are required in order to satisfy the short-term liquidity ratio, as demanded by the simulated hypothesis for bringing forward the schedule in this ratio. Figure 4 shows the impacts of these decisions.
One of the advantages of using formal optimisation tools is post-optimum analysis, which can be used to classify the restrictions imposed as “unfulfilled” , “unnecessary” and “active” , showing the amount required in order to be satisfied or active in the first two cases.
In the present example, we can see the effect of the short-term liquidity restriction, which becomes an active restriction, altering investment and debt policy due to its presence. Figure 5 shows the activation of the liquidity constraint imposed in the last period.
The decisions made by the system may seem very sudden. However, relative growth minimum and maximum restrictions would be used in a real environment, meaning that the changes can be relaxed, therefore assuming criteria that do not allow direct modelling.
Conclusions
- Recent legislation, in particular Act 9/2012 of 14th November, provides clear evidence of incorporating strategy and decision-making in order to determine the viability of an entity, rather than a static approach with regulatory compliance.
- We have generalised the new methodology to include entities which are not in the restructuring process, and which carry out balance sheet projection analysis for shareholders, regulators and supervisors.
- The need to use methods, techniques and optimisation algorithms comes from the increased market and regulatory restrictions. When we talk of a “new methodological approach in financial regulation”, we are referring to the change from an individual profitability and risk calculation methodology to a regulated system.
- The financial sector has developed modelling concepts which are heavily biased towards risk and price calculation. Nowadays these are widespread and commonly used for all business aspects: marketing and sales, human resources, strategic planning and financial management. From a methodological perspective, this bias can be seen in projections based on serial analysis, which is less effective when there is a change of cycle or volatility is high, as has been the case over the past four years.
- Decisions are ever more difficult and complex, meaning the integration of existing models, along with their dissemination in the entity, is absolutely fundamental.
- Particularly important is that which refers to integrating internal and external information for commercial purposes and developing a well-designed, effective marketing plan, the cost and results of which can be evaluated.
- We have established a unique orientation for the capital, liquidity and commercial requirements of a financial entity, within a strategy to optimise and project the entity’s balance sheet.
