Page 1 Recording of this session via any media type is strictly prohibited. Page 1 The GPS Framework: A Comprehensive Approach to Strategic Risk Management Damon Levine, CFA, CRCMP Vice President, Enterprise Risk Management Assurant Inc.
Jan 04, 2016
Page 1
Recording of this session via any media type is strictly prohibited.
Page 1
The GPS Framework: A Comprehensive Approach to
Strategic Risk Management
Damon Levine, CFA, CRCMPVice President, Enterprise Risk ManagementAssurant Inc.
Page 2
Recording of this session via any media type is strictly prohibited.
Legal Disclaimer
The views expressed herein are my own and not necessarily those of my employer, Assurant Inc.
Page 3
Recording of this session via any media type is strictly prohibited.
About the Speaker• Damon Levine is Vice President, Enterprise Risk
Management at Assurant Inc. • 20 years experience in actuarial consulting,
portfolio strategy, and risk management with past presentations to the ERM Symposium, ISO's ERM Forum, and the Actuarial Society of NY.
• Published in The Actuary, SOA's Risk Management journal, and the Society of Actuaries exam syllabus.
Page 4
Recording of this session via any media type is strictly prohibited.
Today’s Key Takeaways• A clear understanding of the approach, tools, methods, and scalability of GPS• The ability to measure all types of threats to strategic execution with the same set of metrics• The capability to manage strategic risks with a powerful and practical method which:
•Creates buy-in and improves both strategic planning and execution•Enables informed management reaction to strategic risks
Page 5
Recording of this session via any media type is strictly prohibited.
AgendaERM, Strategic Risk Management, and GPS
Risk Quantification in GPS
Important GPS Concepts and Tools
An Execution Management Cycle
Quantitative Applications
Page 6
Recording of this session via any media type is strictly prohibited.
A Convenient Fallacy
• Strategy versus strategic objectives
• Risks to strategic objectives
• Does strong Enterprise Risk Management (ERM) imply effective Strategic Risk Management (SRM)?
Page 7
Recording of this session via any media type is strictly prohibited.
GPS: Goals-Progress-Strategy• Goals: Ensure a well defined business goal,
i.e. strategic objective; it identifies key risks to achievement of the objective and its foundational “critical-to-success” goals and lower level checkpoints
• Progress: GPS includes key progress metrics and an Execution Management Cycle (EMC)
• Strategy: the EMC assesses risk evolution and informs risk-intelligent “course correction”
Page 8
Recording of this session via any media type is strictly prohibited.
Risk Quantification in GPS
• Scenario analysis Goals Process: the risk interview Benefits
• Probability• Macro events and conditional probability
Page 9
Recording of this session via any media type is strictly prohibited.
Example: Scenario Approach
Scenario Description Probability Impacts to Business Drivers Year 1 Earnings Impact Company Value Impact
1. Exchange rate stays within 10% of March 1 levels for next 12 months 35%
Assume baseline forecast interval estimate applies
Assume baseline forecast interval estimate applies
Assume baseline forecast interval estimate applies
2. Dollar appreciates vs. Euro by 10-20% 25%$US Sales Down 20% (vs. baseline, post currency translation) -$20M -$54M
3. Dollar appreciates vs. Euro by >20% 20% $US Sales Down 35% -$35M -$95M
4. Dollar depreciates vs. Euro by 10-20% 15% $US Sales Up 15% $15M $41M5. Dollar depreciates vs. Euro >20% 5% $US Sales Down 30% $30M $81M
Statistical Expectation -8M -22M
Downside Conditional Expectation -27M -72M
Challenges [list perceived difficulties in risk prevention or impact reduction]
Mitigations [identify existing risk controls that reduce likelihood and/or expected business effects]
Potential for Action [assess the expected benefit to the company’s risk-reward profile from additional focus or effort on risk mitigation]
Scenario Summary for Foreign Exchange Rate Risk (Dollar vs. Euro)
Page 10
Recording of this session via any media type is strictly prohibited.
Financial Quantification of Risk• Frame each scenario’s impacts in terms of key
performance drivers including sales/other cash flows, expenses, etc.
• Translate scenario impacts into effects on income statement and balance sheets
• Spreadsheet logic and accounting rules make the key metrics simply “fall out” of model (e.g. earnings, ROE, IRR etc.)
Page 11
Recording of this session via any media type is strictly prohibited.
Additional Tools: Old and New
• Well known techniques from ERM play key roles in GPS
Scenario approach to risk ID and quantification Mitigation/control assessment
• Potential for Action (PFA)• Required Recovery Ratio (RRR)• The Logical Framework Approach (LFA)
Page 12
Recording of this session via any media type is strictly prohibited.
Potential for Action (PFA) and Risk Velocity
• PFA is a qualitative assessment of the expected improvement in a company’s risk-reward profile due to increasing or expanding risk controls/mitigation
• Expected “bang for the buck” of additional effort/resources for a specific risk exposure
• Risk Velocity: measure of expected time from risk manifestation to financial impact
Page 13
Recording of this session via any media type is strictly prohibited.
Required Recovery Ratio (RRR)• Assume a strategic objective based on a sales metric and 3-
year forecast horizon: year 1: 100M, year 2: 200M, year 3: 250M (a 3-year total of 550M)
• Assume actual year 1 sales are 50M and define RRR by the “get back on plan” relation:
50 + RRR(200+250) = 550• RRR=111% we must outperform forecast by 11% in years
2-3• Higher RRRs smaller likelihood (estimated at beginning of
year or before project launch!)
Page 14
Recording of this session via any media type is strictly prohibited.
RRR & Probability of Success• RRR is a function of actual performance to date and
plan projections for the remaining time horizon
• Various ranges of RRR are considered before a strategic project launch and corresponding probabilities are estimated
• RRR, in conjunction with other indicators, provides a dynamic estimate of the probability of goal achievement
Page 15
Recording of this session via any media type is strictly prohibited.
The Logical Framework Approach (LFA)• A powerful management tool driving successful
strategic planning and execution • Developed in 1969 for the United States Agency for
International Development (USAID) • Originally used for international development
projects and used in over 35 countries • Very effective in corporate settings to ensure
attainment of strategic objectives• Temporal Logic Model and If-Then Thinking
Page 16
Recording of this session via any media type is strictly prohibited.
Effective Strategic Planning and LFA
• A “time-reversed” causal chain Strategic objective Critical-to-Success Goals Sub-goals
and Tasks Plan left to right and (ideally) execute right to left
• If-Then thinking and progress metrics in an SRM setting
Page 17
Recording of this session via any media type is strictly prohibited.
An Execution Management Cycle• ID Critical-to-Success (CtS) Goals• Goals Progress Strategy
Goal definition and project planning Risk ID and progress assessment with metrics and early
warning indicators (EWI) Adaptive Management: strategic course correction and
tactical actions
• Informed and dynamic execution of an objective from start to finish
Page 18
Recording of this session via any media type is strictly prohibited.
Define SMART Strategic Objective, Critical to Success (CtS) Goals & pre-requisite
tasks
Select metrics/indicators, upfront RRR analysis, and risk ID & quantification through scenario
approach
Calculate and track metric/EWI values, risk velocity & exposures, mitigation effectiveness
(PFA), and RRRs
Report and interpret data and findings, PFAs, and overall risk-reward
outlook/assessment
Apply adaptive management: revise risk mitigation
Business tactics/strategy and inform ‘go/no-go’ calls
Research performance drivers, metrics and EWI. Employ Logical Framework techniques to
analyze CtS goals, sub-goals, and if-then assumptions
Does overall strategy need to be
altered?
No
Yes
Do we needto re-define metrics/EWI
or revise risk
scenarios?
No
Yes
GPS in Action: Managing Strategic Execution
Page 19
Recording of this session via any media type is strictly prohibited.
GPS and Strategic Planning• Strategic objective Critical-to-Success (CtS)Goals
Sub-goals and Tasks• Identify challenges and risks to attaining CtS goals
and the smaller sub-goals or tasks (include LFA’s if-then thinking)
• Quantify potential impacts in light of existing mitigations or controls
• Provide a ranking of key risks and their PFAs to prioritize management action
Page 20
Recording of this session via any media type is strictly prohibited.
Progress Assessment for Strategic Objectives
• At regular time steps GPS will: Identity and quantify risks to key tasks, projects, and CtS
goals and assess corresponding risk controls/mitigations Determine progress metrics, RRR & outlook for success Provide an estimate of the expected benefit of increased
mitigation effort (PFA)
• Crucial performance indicators are tracked and reported for those areas driving success (e.g. training, marketing , sales, profit, etc.)
Page 21
Recording of this session via any media type is strictly prohibited.
Enabling Adaptive Management• The EMC allows for:
Timely assessment of progress toward goals expressed in objective measures and metrics
A dynamic view of risks to achievement of the key foundational tasks and the overarching strategy
A risk-intelligent basis for course correction: high level and tactical revisions to business plans
• Management has a realistic view of the priority for shoring up risk mitigation and increased focus on strategic elements
Page 22
Recording of this session via any media type is strictly prohibited.
Stochastic Approaches
• Scenario impacts need not be point estimates• Uniform, triangle, normal or other
distributions may be assumed for impact estimates of key business drivers or income statement items
• Refining scenario analysis with conditional probabilities
• Macro factors and correlation
Page 23
Recording of this session via any media type is strictly prohibited.
Stochastic Simulation in Excel• Assume a risk source modeled with three scenarios:
S1 = no/mild impact, S2 = moderate impact, and S3 = severe impact & probabilities 60%, 30%, 10% resp.
• Generate random digit in (0,1): r• Use r to determine which scenario has occurred:
0<r<.6 S1, .6<r<.9 S2, and r>.9 S3• Next slide illustrates this concept in Excel; note that
subintervals corresponding to scenarios have widths equal to the scenario probabilities
Page 24
Recording of this session via any media type is strictly prohibited.
Stochastic Simulation in Excel (cont’d)
r in (.6,.9) then S2 Occurs r>.9-->S30 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
if r falls in (0,.6) then S1 Occurs
•Assume P(S1)=.6, P(S2)=.3 and P(S3)=.1•Use “=rand()” in Excel to generate random digit, r, from (0,1)•The rule below associates r with a particular scenario (i.e. the random value of r indicates the scenario which occurs)
Page 25
Recording of this session via any media type is strictly prohibited.
Simulation of a Strategic Objective with Two Risk Sources
•A random digit from (0,1), r1 determines the simulated scenario for risk source 1 •Another independent random digit r2 determines the simulated scenario for risk source 2
r1=.6531
Use r1 and r2 to indicate simulated state of each risk source
risk source 1: scenario 2 is simulated
determine performance for strategic objective X:aggregate impacts from above simulated scenarios
r2=.3215
risk source 2: scenario 1 is simulated
0 0.2 0.7 1
scen 1 scen 2 scen 3prob=.20 prob=.50 prob=.30
0 0.4 0.9 1
scen 3prob=.10prob=.50
scen 1prob=.40
scen 2
Page 26
Recording of this session via any media type is strictly prohibited.
The Macro Factory Overlay • Model macro factors that significantly affect
likelihood of outcomes for risk sources of strategic objective X
• Simulate macro factor states: M1,M2,… (e.g. recession, inflation, pandemic)
• For each macro state i, estimate strategic objective probabilities: pr(scen 1 | Mi), pr(scen 2 | Mi), …
Page 27
Recording of this session via any media type is strictly prohibited.
Simulate performance of Strategic Objective X: Based on the simulated impacts for the scenarios, aggregate the effects on key metrics such as
earnings, IRR, ROE, etc
Simulating Strategic Objective Performance
Generate several random digits from (0,1) to simulate the state of each of the macro factors; i.e. model the “state of the world”
For each risk source (R1, R2,…) affecting the performance of strategic objective X, we note the “activated set” of scenario probabilities
For each of R1, R2, …generate a random digit from (0,1) to simulate which scenario occurs for each of these risk sources
Page 28
Recording of this session via any media type is strictly prohibited.
Macro Factor Based Simulation of a Strategic Objective
Economic scenario 2 is simulated
Regulatory scenario 3 is simulated
r1=.3531
r2=.8108
Simulate each state of each Risk Source X, Y, Z,…thus simulating the strategic objective
0 0.2 1
scen 2prob=.80
scen 1prob=.20
0 0.2 0.7 1
scen 1 scen 2 scen 3prob=.20 prob=.50 prob=.30
Strategic Objective Risk Source XMacro “state of the world” activates conditional probabilities for simulation of risk x:
(15%,60%,25%), (20%,50%,30%), (10%,55%,35%)
The above macro scenarios activate conditional probabilities for each risk source of the objective…
Page 29
Recording of this session via any media type is strictly prohibited.
Simulated Infusions
Risk-adjusted Capital of a Strategic Objective
• Best estimate/baseline forecast for 3 years of objective X: B1, B2, & B3 (e.g. $earnings in years 1-3)
• Denote years 1, 2, and 3 earnings levels in simulation k by E1k, E2k, and E3k
• Notional supplemental flows to meet baseline: B1- E1k , B2- E2k , and B3- E3k for years 1-3 respectively
• Kth infusion = PV(supplement) = (B1- E1k)/(1+i) + (B2- E2k)/(1+i)2 + (B3- E3k)/(1+i)3
Page 30
Recording of this session via any media type is strictly prohibited.
Strategic Objective Risk Capital
Risk-adjusted Capital of a Strategic Objective
• In large number of simulations observe 95th%ile of simulated infusions for strategic objective X
• RACX = risk-adjusted capital = 95th %ile of infusions
• In one run risk capital may be determined for all objectives simultaneously: RAC1, RAC2, …
Page 31
Recording of this session via any media type is strictly prohibited.
Return on Risk-Adjusted Capital
of a Strategic Objective• Calculate in large run: RAC1, RAC2, …
• Observe average “reward” for each objective (e.g. average PV(distributable earnings) or IRR): y1, y2,…
• Define return on RAC for the “strategic objective i” as RORACi = yi/RACi
Page 32
Recording of this session via any media type is strictly prohibited.
The Portfolio View
• Simulation allows for portfolio infusion notion and RACPORT and yPORT, the average reward for the portfolio of strategic objectives
• We may define RORACPORT = yPORT/ RACPORT
• Estimating a strategic objective’s contribution to portfolio RAC and RORAC
• “Risk classes” and exposure to macro factors
Page 33
Recording of this session via any media type is strictly prohibited.
Risk Appetite
• A previously vague or “pie in the sky” risk construct
• Now we may define: Desired risk class allocation of the portfolio of
strategic objectives Support only those objectives increasing RORACPORT Portfolio exposure limits to specific macro factors
Page 34
Recording of this session via any media type is strictly prohibited.
Risk-Intelligent Capital Deployment• Based on upfront risk analysis how does one
decide on the “green light” for an objective or choose among competing objectives?
• Illustrative policy: Priority based on objective or portfolio RORAC Priority as a function of resulting portfolio exposure to macro
factor(s) or maximum risk class allocation Consider performance distribution of the objective
Page 35
Recording of this session via any media type is strictly prohibited.
In the Research Paper• More detail, slower buildup to ideas• A risk-adjusted compensation framework and
100 Day Implementation Plan• Definitions & examples of risk velocity, RRR, &
PFA• Illustrative application to a new product launch• Available online at:
http://www.ermsymposium.org/2013/pdf/erm-2013-paper-levine.pdf
Page 36
Recording of this session via any media type is strictly prohibited.
Closing Thoughts
• These techniques generally apply to ERM and to a large extent, GPS is an application of robust ERM to strategic risks
• GPS will still “work” if only partially implemented and does not commit a company to a single SRM/ERM path
• Contact: [email protected]• Questions?