Paolo Laureti: ERM platforms are enterprise-wide frameworks to identify financial risks and manage them in accordance with a set of business objectives. The main goals are to protect the firm from adverse events and create value for internal or external stakeholders – owners, customers, auditors, rating agencies, and regulators.
The ERM framework should be complete and interactive, which means that business users are able to access an end-to-end workflow, with the ability to generate stochastic projections of the entire balance sheet and compute analytics at any aggregation level. The outcome is a set of reports, which typically include market risk, credit risk, and potentially other risk types; they should be accessible interactively, as users need to run ‘what if’ analyses and stress tests independently of IT.
Whether it is accessed as a service on cloud or on-premises, as long as the platform is user-friendly, robust, and trustworthy, ERM can be used for compliance and beyond, to support risk-aware business decision making.
Paolo: The scope of ERM can be very wide, and its evolution depends on the initial need of the firm. Broadly speaking, the drivers for an ERM platform could be either risk and compliance, or ALM and investments: most of the time it’s a mixture of both.
"After the pandemic, we are effectively witnessing a new wave of applications
for Internal Models, which typically require robust ERM foundations."
ERM platforms allow firms to build a risk culture inside the firm, improve its management and as a result, potentially reduce capital charges. Therefore, risk and compliance are often the entry point, with the goal to compute Economic Capital. The push comes from rating agencies that are interested in risk tolerance at the enterprise level, and from regulators who advocate a principle-based approach to risk management.
The breakthrough in this respect is Solvency II, in effect in the EU since 2016, and equivalent regimes such as those in Switzerland and Bermuda. After the pandemic, we are effectively witnessing a new wave of applications for Internal Models, which typically require robust ERM foundations.
Insurance Capital Standards are increasingly being applied at a global level, e.g., following the initiative of the International Association of Insurance Supervisors. In some Asian countries, for example, risk-based capital regulations are coming into effect at the same time as IFRS 17.
Paolo: The other common reason to build an ERM platform is to obtain a business advantage by improving the Asset Liability Management (ALM) and liquidity risk processes, with the goal to optimise the Strategic Asset Allocation (SAA) and align it with the firm’s risk appetite framework.
This is particularly important in the current volatile market conditions, both in terms of risk (hedging, capital management) and in terms of opportunities (increase returns, generate alpha).
The advantage of an ERM approach to complement actuarial projection systems is a complete view of market-consistent balance sheet projections under large Monte Carlo scenario sets. Analysts can effectively calculate quantities such as cashflow mismatches and key rates duration, while monitoring market risk and credit risk at the same time in a stochastic framework. Investment portfolios can be optimised leveraging the same infrastructure.
Paolo: Asset modelling is the starting point of an effective ERM approach.
After more than a decade of low-interest rates, insurers have introduced complex assets into their portfolios in the form of derivatives, real estate investments, and alternatives, including private equity and debt. To capture the resulting multi-factor exposures, often hidden into funds and funds of funds, precise asset modelling is crucial. Regulators in different geographies have started to emphasise this explicitly.
"We need not only to run large simulations but give
users the ability to run a stress test in real time."
For example, with instruments such as a convertible bond, the solution must be able to indicate the likely exercise date under normal conditions, stress conditions, and value at risk (VAR) scenarios to capture cashflow generation correctly. Another example is that of complex exposures within funds which, if ignored, can lead to a severe underestimation of concentration risk. An advanced platform should support the full look-through approach into the funds, as well as other simplified models applicable with the principle of materiality.
Native Asset Modelling becomes essential in periods of high volatility when the behaviour of risk factors is somewhat erratic and can potentially diverge from normality by a great amount. In today’s high inflation and energy crisis environment, we need not only to run large simulations but also give users the ability to run a stress test and a ‘what if’ analysis in real time, to react quickly to market changes.
Paolo: To generate reliable projections of the market consistent balance sheet, companies need to model the liabilities correctly and consistently alongside the assets, and under the same risk factor universe.
The most common approach is to calibrate proxy models to represent liability portfolios, so as to keep the simulation time within hours. Examples of proxy methodologies are Curve Fitting, Least Squares Monte Carlo, Portfolio Replication, and distribution functions. The choice of the proxy method depends on the nature of the business (life or P&C) and of the policies, the accuracy of the sample data and the scope of the solution, e.g., do they need cashflow-generating functions. Some machine learning techniques can be very helpful in improving the robustness of those proxies.
Once assets and liabilities have been modelled in a consistent framework and projected under real world shocks, it is easy to aggregate them into portfolios and generate analytics for risk and investment management purposes if we deploy an advanced risk dashboard.
Paolo: Accurate capital management is one of the main objectives of ERM, especially in highly regulated markets. Leveraging the infrastructure just described, scenario-based aggregation allows users to account for full diversification effects, perform risk attribution and calculate the risk contributions of each entity and risk type.
"The reporting dashboard must allow users to slice and dice
tail risks under changing condition."
Typical questions that need answers are: what is the VAR at the group level? What is the contribution to the group VAR of interest rate, credit risk, spread, or FX? How much of it comes from a specific legal entity? What are the shocks that drive these capital requirements and why? And how can they be acted upon to mitigate those risks?
The reporting dashboard must allow users to slice and dice tail risks under changing condition, to attribute the source of extreme losses to sub-modules and business lines, down to single instruments and individual risk factor levels. The stability of risk measures, as well as their decomposition into risk factors and comparison across different time periods, have been some of the use cases the companies have been focusing on in recent years.
Paolo: One of the advantages of leveraging an ERM platform for strategic asset allocation purposes is that it allows to define and enforce a consistent risk tolerance and appetite framework at all levels of the reporting hierarchy – from individual portfolios to legal entity and group level.
"With LDI, insurers seek to optimise return on their investment portfolios
while ensuring that they match the outflows from their liabilities."
The portfolio optimisation tool must be seamlessly integrated with the infrastructure used to project assets and liabilities, aggregate risk and calculate analytics. The foundation of an efficient workflow to devise Liability-Driven Investment (LDI) strategies must ensure consistency of data and models, coupled with the management of users’ roles and permissions,
With LDI, insurers seek to optimise return on their investment portfolios while ensuring that they match the outflows from their liabilities. At the same time, they must keep the capital requirements under control. To support this type of design, users need to define objectives and apply constraints on at least three dimensions: return, risk, and capital. Each of these dimensions can be an absolute or active measure, i.e., relative to a benchmark, which can be the liability portfolio itself.
The outcome of this process is a solid framework that allows to improve not only risk and capital management, but also asset-liability and investment management, thus providing a competitive advantage.
