Using an Efficient Frontier Analysis Approach to Improve

Using an Efficient Frontier Analysis Approach to Improve ERM Implementation Prepared by: Ward Ching, Vice President, Risk Management Operations, Safeway, Inc. Loren Nickel, FCAS, CFA, MAAA, Regional Director and Actuary, Aon Global Risk Consulting RIMS Session ERM 009 Recording of this session via any media type is strictly prohibited. Page 1

Before we get started… • This is a technical discussion of the use of EFA techniques as applied to hazard risks within an hypothetical organization. • Yes, there will be some math. The obligatory equation and assumption set is gratuitously included for your reading and napping enjoyment. • Yes, there will be a number of graphs that have reasonably • • squiggly lines in them. If we feel like it, there may be a quiz at the end of the session, just to see if you were awake. Cheating may be necessary. Start taking notes now, just in case. If your head hurts, even slightly, we’ve done our jobs. Recording of this session via any media type is strictly prohibited. Page 2

Anyone who wants to leave, now is the time to do it… You are now well and truly warned. . . Recording of this session via any media type is strictly prohibited. Page 3

We won’t point fingers at you (well maybe) or snicker too loudly… Recording of this session via any media type is strictly prohibited. Page 4

Your dignity is our main concern…well, sort of. Recording of this session via any media type is strictly prohibited. Page 5

Over the past 25 years, advanced quantitative financial and behavioral analysis has received much attention • • Need to better understand predict performance impact of hazard risk as a financial portfolio Limitations of single discipline modeling often lead to misreading or under valuation of complex business risks across operational categories. Need to answer broader risk related questions (insurance, finance, audit and operations) to seek out a more integrated and data robust approach. Questions regarding the dynamic impact of enterprise-wide risk portfolios has given rise to robust Enterprise or Strategic Risk Management techniques. Recording of this session via any media type is strictly prohibited. Page 6

The RIMS Strategic Risk Management Framework Recording of this session via any media type is strictly prohibited. Page 7

Framework highlights… • The focus of the Framework is on the impact of risk on strategies • • How do differing risk profiles create economic or operational tradeoffs? How do the tradeoffs impact risk transfer or risk assumption decisions? Are all risk assumption decisions the same? How do risk tolerances change as an entity’s strategy changes? • Do the “analytically informed” risk adjusted strategies yield competitive or exploitable opportunities? Recording of this session via any media type is strictly prohibited. Page 8

Modern Portfolio Theory is the analytical foundation that underpins Efficient Frontier Analysis • MPT is a mathematical method developed in the mid-1950 s. • Theory of Finance focused on the maximization of portfolio return while minimizing risk subject to an expected predetermined financial return. • MPT simulates the impact of various “risk adjusted” strategies against fixed targeted returns. • The objective of MPT is to understand the sources of volatility across a defined portfolio or mix of assets. • MPT models generally assume normal distribution functions and views risk as the standard deviation of return. Recording of this session via any media type is strictly prohibited. Page 9

Types of questions asked • What is the economic value of an organization’s material risk profile when characterized as a financial portfolio? • How can the economic or operational volatility of an organization’s risk profile be examined dynamically and intertemporally? • Are an organization’s risk mitigation strategies and methods efficiently matching its risk profile? • If an organization changes its operations in a material way, what impact can be visualized across the organization's risk portfolio? • If risk retention and risk transfer are considered two independent variables in an organization's’ risk profile distribution, how can the two be maximized through insurance purchase? Other Techniques Used: Dynamic Financial Analysis, Capital Asset Pricing Modeling, Behavioral Economics. Recording of this session via any media type is strictly prohibited. Page 10

What is Efficient Frontier Analysis? • EFA is a quantitative technique that allows for the graphical depiction of an organization's’ portfolio of risk. • In many cases EFA is performed in conjunction with higher order simulation methods such as the Monte Carlo application in insurance. • EFA focuses on understanding the trade offs between risk assumption and risk transfer given an optimal risk portfolio. Recording of this session via any media type is strictly prohibited. Page 11

EFA Questions • What is the shape, size and volatility of an organization’s risk frontier? • What drives changes in the shape of an organization’s risk frontier? • Do the risk mitigation methods being used by the organization to manage/assume/transfer risk correspond to the risk frontier? • Are there instances where the risk mitigation methods are inefficient relative to the risk frontier (i. e. , under defend volatile or difficult risks and over defend trivial risks? ) • Can the use of EFA influence insurance-related retention efficiency or layering decisions? • Can the use of EFA suggest different or integrated approaches to financially engineering risk across hazard, market, operational, or reputational risk spaces? Recording of this session via any media type is strictly prohibited. Page 12

Benefits to Risk Management The four benefits to risk management as defined by James Lam are: • Managing risk is management’s job • Managing risk can reduce earnings volatility • Managing risk can maximize shareholder value • Risk management promotes job and financial security Recording of this session via any media type is strictly prohibited. Page 13

Financial Efficient Frontier Formulas Transformed into Insurance Framework Standard Expected Loss Formula: E(rc) = rf + y[E(rp) – rf] We will replace E(rc) with E(rsp) = expected risk spend; which is expected retained loss + premium paid to transfer risk Standard Risk Formula: σ2 p=w. A 2σA 2+ w. B 2σB 2+2 w. Aw. B σA σB ρAB We will replace σ2 p with TVAR 95= Tail Value at Risk and is the average of the 5% of the worst outcomes in a simulation for the portfolio of risk Recording of this session via any media type is strictly prohibited. Page 14

Efficient Frontier Framework Portfolio Preference Recording of this session via any media type is strictly prohibited. Page 15

Sample Case Study – Current Program We have a sample insurance program which includes three lines of coverage: Earthquake, Workers’ Compensation and General Liability. The following table shows the retention and limits purchased, along with the current expected losses retained by line and for the portfolio. Recording of this session via any media type is strictly prohibited. Page 16

Sample Case Study – Mathematical Options Using the mathematical method, we select options that are mathematically close to the current program and then extend outward. Under the mathematical method, there is no recognition of marketplace dynamics; it is purely a linear selection of options. The availability approach (not shown here) selects options that are known to be available in the marketplace, and may not be linear. Recording of this session via any media type is strictly prohibited. Page 17

Sample Case Study – Simulated Losses Recording of this session via any media type is strictly prohibited. Page 18

Sample Case Study – Efficient Frontier Recording of this session via any media type is strictly prohibited. Page 19

Sample Case Study - Findings The portfolios no longer follow the efficient frontier, as some of the options lie above the efficient frontier line. The slope of the efficient frontier is steep, and follows the risks that contribute to the portfolio (earthquake in this instance is driving the steep curve) If the organization is using a risk appetite for the entire portfolio, then they would look at the efficient frontier below the $20 million dollar Tail Value at Risk level. (Only option #4 qualifies) Recording of this session via any media type is strictly prohibited. Page 20

Several Modern Portfolio Concerns Contained in the Framework Asset returns are (jointly) normally distributed random variables • Addressed by using TVAR and non-normal assumptions Correlations between assets are fixed and constant forever • Addressed by using copulas and/or correlation matrix All investors have access to the same information at the same time • Insurance markets clearly have this concern, where anyone with more information can be at an advantage All securities can be divided into parcels of any size • Still a concern, as insurance products are not fractional. However, more information may lead to a product that is better than current offerings. Risk/volatility of an asset is known in advance/is constant • Addressed by using simulations allowing for parameter risk in addition to process risk Recording of this session via any media type is strictly prohibited. Page 21

Behavioral Concerns of Efficient Frontier The different types of informational errors are: Forecasting errors • Over-relying on recent experience Overconfidence • Are you ‘lucky’ or ‘good’? Conservatism • Slow to react to new information Sample size neglect and representativeness • Using too small a sample size forecasting Recording of this session via any media type is strictly prohibited. Page 22

Takeaways Businesses will begin to focus on insurance risk in a portfolio Our new framework is robust, and appropriate for today’s computing power and ‘big data’ storage Experience is an important criteria for any framework, i. e. knowing when it is appropriate and when it is not Recording of this session via any media type is strictly prohibited. Page 23

Ok…you can leave now… Recording of this session via any media type is strictly prohibited. Page 24