Learnings from project implementations - (surprises, landmines) Joanna Starczewska RQS, Advisory Business Solutions Manager, Insurance
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Learnings from project implementations
(surprises, landmines)
Joanna Starczewska
RQS, Advisory Business Solutions Manager, Insurance
Company Conf ide ntial – For Inte rnal U se Only
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.
Project complexity drivers
There are various types of factors which may strongly impact the amount of work within the project.
Architecture, volumes Methodology
• Scope of implementation • Different approaches used (BBA, PAA, VFA)
• Centralized vs decentralized • Level of detail of calculations (lob, cohort, contract, unit of exposure)
• Number and complexity of entities • Complexity of grouping algorithm
• Number and complexity of source systems • Complexity of approach to calculation of RA, CSM, LC
• Number of data packages to be handled • Complexity of approach to reinsurance held
• Number of UoA and their subgroups Project • Number and complexity of posting rules
• Number and complexity of reconciliation rules
complexity • Complexity of approach for transition period
Data
• Quality of input data
Process
• Number and complexity of validation rules • Number and complexity of processes to be handled
• Number of different data sets required by actuarial • Running the proces on the subset of the whole portfolio
➢
tools and complexity of their content • Performance requirements
• Number of data sets generated by actuarial tools, • Level of traceability and auditability required
validation rules to be run on them, amount of Reporting
further processing • Number and complexity of reports changed within IFRS 17 (solo for all entities and for group )
• Number and complexity of data to be generated for • List of external reports (if any) defined as required (solo for all entities and for group )
accounting system.
• Number and complexity of reports within „internal reporting”
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Scope and approach to implementation
IFRS17 specific calculations,
Actuarial, postings and reporting
Accounting
modeling tools
RA_init Calculations IFRS 17 Calculation IFRS17 Subledger
1. Gather detailed data 1. GL Closing
Calculations of IFRS17 Generate posting entries
2. Generate expected cashflows specific measures related to reserves. 2. Consolidation
on the level of contracts IFRS 17 Grouping Generation of IFRS17 Run trial balance and other
Assignment of contracts accounting events, validation rules. 3. Strategic Planning
3. Prepare actual cashflows on into groups for which
Incl RA Calculations
the level of Unit of Account CSM/LC is calculated
DATA MANAGEMENT + WORKFLOW + REPORTING
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.What is the Unit of Account (UoA)?
NonLife
business
In most of the cases
Portfolio
actuarial models are
working on portfolio
level and actuaries
need to allocate down.
Unit of account
?
Life
business
Actuarial models work
on policies or model
points and at initial
Contracts
recognition policies exposures
(contracts) need to be
assigned/aggregated
to UoA
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Obtaining the UoA
How to assign RA at How to define Should several options How to derive the
z
initial recognition to the „significantzpossibility”? of groupingzbe tested? z UoA based
attributes of
contract? on contracts’ ones?
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Obtaining the UoA
Derivation of attributes
Attributes of UoA that steer the calculations
- Measurement approach
Could be set depending on the line of - Method of RA calculations (and parameters to be used)
business or product - Approach to finance income and expenses
- Significant finance component (separately for LRC and LIC)
- Approach to amortization of acquisition expenses
- Approach to Analysis of Change
- IR to be used for discounting
Min initial recog, min Begin, max end - Initial recognition date, Begin, end of coverage date
Based on initial recognition - UoA subgroups
Data and parameters
Sum - Expected cashflows,
- Actual cashflows
Based on parameters on contracts, but: several questions - Amortization paremeters (for CSM, premium, acq expenses)
arise:
• It should rather be weighted, but with what?
• Should it be recalculated every reporting period (to
make allowance for the derecognition for instance)
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Availability of data
Experience Adjustment, Actual cashflows
Expected CFs
(for PV and RA,
by different legs)
UoA Id
(info about ins period,
acc approach, options)
Actual CFs
(by accounts)
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Loss Component
In case of negative impact of change in expectations
unwinding Exp_adj (premium,
Initial acq exp, inv comp)
recognition Chng in Expectations
(in PV, RA)
CSM Release CSM_EoP
=0 =0
CSM_BoP
LC
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Loss Component
Loss Component should be Consistent presentation of
identified and detailed disclosure impact on SCI is required
of its release is required
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Availability of data
Change in expectations for reinsurance held (par 66c)
2015Q4 2016Q1 time
Re_Goc_2015
One needs to identify which
cashflows are coming from
[underlying] UoAs which are And some other coming
profitable, and which ones from other Gocs
from [underlying] UoAs which New reinsurance programme (with higher proportion)
are onerous at inception or Re_Goc_2016
become onerous. And some other coming
from other Gocs
Reinsurance held
Gross_Goc1_2015
Gross_Goc2_2015 Becomes
onerous
New contracts from 2016 will
be reinsured by Re_Goc_2016
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.
Assumed businessAvailability of data
Disocunting done in actuarial models
Detail of portfolio
Insurance contract Unit of Account
PV or Input option 2: Input option 4:
Movements PV and/or PV and/or
movements for movements for
contracts groups of
IFRS 17 Grouping contracts
Assignment of contracts IFRS 17 Calculation
into groups for which Calculations of IFRS17
CSM/LC is calculated specific measures
Detail of data
Generation of IFRS17
Discounting Discounting accounting events
Input option 1: Input option 3:
Undiscounted Undiscounted CFs PV and/or
CFs for contracts movements for
IFRS 17 Grouping groups of
Assignment of contracts contracts
into groups for which
CSM/LC is calculated
Default
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d. oneAvailability of data
Analysis of change
t-1 t t+1 t+2 t+3 t+4
t-1
Expected CFs
obtained last period as
Closing ones
Δ→
t derecognition
Δ → „Chng in non-
econ assumptions”
Δ → „Chng in econ
assumptions”
Recognition
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Challenges of PAA
Assuming that we have a group of contracts that
coverage period
- starts in 2017.04.01 and ends in 2018.03.31
- Single premium paid upfront (1100)
- Acquisition expenses paid upfront (100) 2017 2018 2019 2020 2021
- Premium release pattern (80%,20%)
Development of claims incurred in 2017
- Claims ratio 78%
Cl2017,2017 Cl2017,2018 Cl2017,2019
- Claims paid following pattern (70%, 20%, 10%)
Development of claims incurred in 2018
Cl2018,2018 Cl2018,2019 Cl2018,2020
1. Is group of contracts onerous? 2017.12 2018.12 2019.12 2020.12
2. How to calculate Onerous Contract
Liability (OCL)? LRC • Expected cashflows: • Coverage period No LRC No LRC
3. What should be the pattern of premium, Acq costs • Amortization parameter
amortization of premium (and acq • Actual cashflows: premium,
expenses)? Acq costs
• LRC at the end should be 0
4. Should discounting be applied? • Coverage period
• Amortization parameter
If OCL calculated then the
If OCL calculated then the
same data as for BBA are
same data as for BBA are
required
required
• Expected cashflows (EXI) • Expected cashflows (EXI) • Expected cashflows (EXI) • No expected CFs
1. How to obtain the expected cashflows
LIC
for LIC just for this group of contracts
(issued in 2017, by proftability group)?
• Actual cashflows • Actual cashflows • Actual cashflows • Actual cashflows
2. Are all claims settlement expenses
available by group of contracts? • Risk adjustment data
Locked-in curve for claims Locked in curve for claims
3. Are adequate data to show the Locked in curve for claims incurred In 2017=2.3, for Incurred in 2018 =2.6 LIC at the end should be =0
reestimation of reserves Incurred in 2017 =2.3 those incurred in 2018 =2.6
avg IR=2.3 avg IR=2.6
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.
avg IR=2.5 avg IR=2.4Challenges of PAA
Onerous Contract Liability
Onerous Contract Liability is But this would mean that this approach To avoid the need of providing such detailed data
calculated as difference between : isn’t that simplified after all, at least not and running such verification all the time, one
• The carrying amount of the liability as it comes to the data required. should be able to set the flag to trigger the
for remaining coverage onerosity test.
• The fulfilment cash flows measured
as under the general model
1. How to calculate the release of LC
in subsequent periods? As delta in
LC differences between carrying
amount of LRC and FCFs?
LRC FCFs LRC FCFs LRC FCFs
Carrying Carrying Carrying
amt amt amt
2015 2016 2017 2018 2019 2020
changes in
expected fulfilment
of cash flows
(claims that will be
incurred in 2018
and 2019 will be
400 each instead of
150 as assumed at
the moment of
recognition)
onerous
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Flexibility of posting logic
Accounting events – details of results of calculations
Unique technical Code of accounting Id of GoC Value of
key of accounting event predefined in transaction
event in given run dimension, derived for each acc
based on event of each
preconfigured logic GoC
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Flexibility of posting logic
Posting logic as configurable element
Accounting Events
Configuration
ACCOUNTING_EVENT_TYPE_DIM
ACCOUNTING_EVENT_TYPE_CD ACCOUNTING_EVENT_TYPE_DESC
LRCCSMRL LRC - Contractual Service Margin - Revenue - CSM release
RULE_SL_ACCOUNT_EVENT_TYPE
SL_ID ENTITY ACCOUNTING_EVENT_TYPE_CD POSTING_GROUP_ID
SL_IFRS17 WW_INSURANCE LRCCSMRL LRCCSMRL
Derivation of
ACCOUNT_POSTING_GROUP
SL_ID POSTING_GROUP_ID GL_ACCOUNT_ID DC_CD MAPPED_COLUMN_NM journal entries
SL_IFRS17 LRCCSMRL IFRS17_COA_LE_Ins_InsCov_CSM D POSTING_AMT
SL_IFRS17 LRCCSMRL IFRS17_COA_PL_IR_CSM_CSMRel C POSTING_AMT
COA_GL_DIM
GL_ACCOUNT_ID GL_ACCOUNT_DESC GL_ACCOUNT_TYPE
IFRS17_COA_PL_IR_CSM_CSMRel Insurance revenue - Release service margin INCOME
IFRS17_COA_LE_Ins_InsCov_CSM Insurance contract related liabilities - RC Excluding loss component-Service margin LIABILITY
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Traceability
Illustration of traceability of calculations
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Down-stream integration
Add details of UoA Detailed data
Adding attributes defined for Detailed information on
UoA that will enable detailed UoA level
Reporting tools
Postings analysis by all required
dimensions
reconciliation,
Postings for UoA register Feedback info
With info on posting rule,
account_id, credit/debit Add details of UoA
Map SoA, Aggregate
• Adding attributes defined for
UoA that are required for
dimensions used in GL;
UoA attributes • aggregate by some of those Aggregated postings
dimensions; map internal SoA By dimensions required by GL, GL system
into Master SoA following Master SoA, on
• Add info about version of data more aggregated level than
UoA
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Reconciliations Between IFRS17 subledger and General Ledger Ensure consistency between results kept General Ldger and and those in IFRS17 subledger. It may be obtained by storing the feedback information about the effects of postings in GL, including the document id in GL. This will enable detailed drill through to the data in the IFRS17 subledger as well. Between IFRS17 and SII Comparison of some BS accounts is planned to be performed. In most of the cases on more aggregated level (UoA are not defined in context of SII, neither SII lobs in context of IFRS17). Between IFRS17 and local accounting standards Comparison of the BS accounts is planned to be performer on more aggregated level. Between IFRS17 disclosures and internal reporting In some cases, for internal purposes, more detailed analysis of results is done by dimensions which are not obligatory from perspective of IFRS17 disclosures but interesting form profitability management point of view. For this purpose, the IFRS17 measures are sometimes allocated down to more detailed level and presented in several internal reports. It is important that the information presented in such reports is consisten with results in disclosure reports Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.
Reconciliations
Between actuarial calculations and accounting postings
Ensure consistency between results of actuarial calculations and balance of account resulting from posting entries
All these steps need to be performed by
„actuarial” module/part of the solution to be able
to calculate the release of CSM and determine
whether UoA is profitable of not.
Results of calculations are provided to the
subledger as acocunting events and based on that
the posting entries are generated. Based on
them, the final balance of each account is
determined.
These balance values should be consistent with
the EoP values calculated by the „actuarial”
module.
Between actuarial calculations in subsequent periods
Very rarely but still [especially for LIC], the predefined movements do not explain the difference between EoP
values of subsequent periods. In such situations, additional „other” element is added.
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Project complexity drivers
There are various types of factors which may strongly impact the amount of work within the project.
Architecture, volumes Methodology
• Scope of implementation • different approaches used (BBA, PAA, VFA)
• Centralized vs decentralized • Level of detail of calculations (lob, cohort, contract, unit of exposure)
• Number and complexity of entities • Complexity of grouping algorithm
• Number and complexity of source systems • Complexity of approach to calculation of RA, CSM, LC
• Number of data packages to be handled • Complexity of approach to reinsurance held
• Number of UoA and their subgroups Project • Number and complexity of posting rules
• Number and complexity of reconciliation rules
complexity • Complexity of approach for transition period
Data
• Quality of input data
Process
• Number and complexity of validation rules • Number and complexity of processes to be handled
• Number of different data sets required by actuarial • Running the proces on the subset of the whole portfolio
➢
tools and complexity of their content • Performance requirements
• Number of data sets generated by actuarial tools, • Level of traceability and auditability required
validation rules to be run on them, amount of Reporting
further processing • Number and complexity of reports changed within IFRS 17 (solo for all entities and for group )
• Number and complexity of data to be generated for • List of external reports (if any) defined as required (solo for all entities and for group )
accounting system.
• Number and complexity of reports within „internal reporting”
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Implementation approach
ASSESSMENT CUSTOMIZATION APPLICATION TRANSITION
• Gap Analysis First model based on Applying the model Adjustments
• Full business scope (iterative or
• Initial roadmap defined use-cases parallel if possible)
Generation of disclosure
reports for comparative
• All to-be-applied IFRS17 • Based on full data studies
approaches • Integration with existing IT
• Based on priorities and availability infrastructure
(people, data, models)
• Covering representative use cases
• no integration yet
Steps may be repeated in iterations
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Possible variations of use cases
• Having or not having investment component
• Applying different amortization parameters
• Having the date of recognition equal to date of inception or not equal (day before etc) and having different values of IR quotes for them
• With or without TVOG
• Based on annual or quarterly reporting
• For quarterly one – simulate situation of NB during the year.
• single premium paid at the beginning or at the end of coverage
• claims are paid when incurred and claims paid following development pattern
• With experience adjustment - difference between expected and paid premiums, acquisition expenses, investment component
• change in expected cashflows.
• IR is changed (with situation when it is changed during the coverage and should still impact or when it happens in last period)
• OCI possible options
• different currencies in cashflows or has different currency than entity
• Derecognition
• profitable UoA becomes onerous and opposite.
• Various approaches of risk adjustment
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.Thank you
sas.com
Copy rig ht © SA S Institute Inc. A ll rig hts re se rve d.You can also read
