This is a weekly series for The Regulatory Reporting Data Model Working Group. The RRDMWG is a collaborative group of insurers, regulators and other insurance industry innovators dedicated to the development of data models that will support regulatory reporting through an openIDL node. The data models to be developed will reflect a greater synchronization of data for insurer statistical and financial data and a consistent methodology that insurers and regulators can leverage to modernize the data reporting environment. The models developed will be reported to the Regulatory Reporting Steering Committee for approval for publication as an open-source data model.
openIDL Community is inviting you to a scheduled Zoom meeting.
One tap mobile +16699006833,,98908804279# US (San Jose) +12532158782,,98908804279# US (Tacoma) Dial by your location +1 669 900 6833 US (San Jose) +1 253 215 8782 US (Tacoma) +1 346 248 7799 US (Houston) +1 929 205 6099 US (New York) +1 301 715 8592 US (Washington DC) +1 312 626 6799 US (Chicago) 888 788 0099 US Toll-free 877 853 5247 US Toll-free Meeting ID: 989 0880 4279 Find your local number:https://zoom.us/u/aAqJFpt9B
Two weeks ago we went through content of document provided by Mr. Harris (Mr. Antley pulled this up on the screen and pointed to .xls file in wiki
Mr. Harris great analysis of what data call is doing - aggregation/review of question that are being commonly asked
Next level of this looking at report we're working toward producing. Defined Base 1.0 /Derived 1.0 and calculations applicable to each.
Two additional tabs added to worksheet - premium record, loss record - run through by Mr. Antley
Gap: does accounting from mid-month term impact this? Answer: 80% same
Move into auto coverage report
Calculations- are various ones derived or base? Column C: up for discussion - will start here
First column - EP - broken down w/aggregations on the left, filtering criteria on right. Report broken out with aggregation on the top, coverage on side (y axis) - picture of auto transacted over course of year
Car years - derived field, bc it comes from premium record, and calculated by exposure (transforming into car years) - we came up with derived for this. Queried possible objections from group
Mr. Harris: isn't months covered going to give us car years. Mr. Antley: he's more on tech side. Pulled open stat plan. Way stat plan calculates based on exposure. How would we from a biz perspective calculate this. Mr. Harris: # of vehicles on policy, months covered - in premium record - we have VIN and months covered and coverage and deductible. If months cov'd = 12, that is one car year. For each VIN.
Mr. Antley queried Mr. Madison. At what level do we get to derived inst. of base.? Mr. Braswell: months + VIN/12 = total car years. Mr. Harris: sorting/filt. criteria = coverage, state, deductible (based on covg.), limit, amount bc we have to determine basic vs. excess. We would calculate VIN months covered by coverage state ded. and limit amount which only applies to BI for liability portion. Summing months covered / 12. Mr. Naik: depends on duration for which we're doing aggregation. Mr. Reale: if we round and it's less than a year, we will always get a 0 or a 1 for a specific vehicle. Mr. Harris: are we doing any statistical reports on a quarterly basis today? Mr. Antley: fast track work happens on a higher frequency but never more frequently than quarterly.
Mr. Sayers: a combination or carrier-by-carrier basis? Mr. Harris: on everything. Mr. Antley: we run extraction on multiple carriers and aggregate it at the end and then round.
Mr. Harris: we need to go back and look at how we do it today re: rounding. Mr. Reale: in 8 months no one will remember this discussion so this needs to be checked and documented. Mr. Harris: we're giving you car months so we're rounding anyway. 2 1/2 wks we will round to one month. Mr. Braswell: as we get into larger and larger calculations rounding becomes less significant.
Mr. Antley: raw data will not leave HDS, so we're making report on each carrier and submitting to HDS and then having analytics node combine all those reports. Mr. Reale: i.e., if we get 5000 lines/carrier for two carriers we submit all lines and then round. Mr. Antley: weighted issue also potentially - in some cases how do we combine multiple reports without skewing results. Mr. Reale proposed two different strategies for how to approach this.
Mr. Braswell: there are no ratios being computed here. Mr. Harris: given the information here we can only calculate a loss ratio. Mr. Sayers: before we start aggregating we have every company's data and if we reverse this order do we start breaking things?
Mr. Reale: customization is all we have to worry about now. (Mr. Sayers agreed - we may eventually have to split functionality btwn carrier and analytics nodes by necessity). Mr. Reale: for now we can't send out data pre-aggregated. Mr. Sayers: we have to calculate ratios after we have the raw #s. Mr. Reale: maybe in the future if there is a data call that needs something different, we can send an aggregation with more sig. digits or metadata (e.g. weighted average). Data payload would be different.
Mr. Antley: Logic for combining info on a particular report needs to be completely configurable (the others agreed).Mr. Reale: each data call is bespoke to the request - configurability is a must. (Others agreed).
Mr. Antley: will talk to stat reporting team to to see if there is anything we can highlight from other lines. Mr. Harris: even if so, this wouldn't break it. We are still bringing raw #s together. Mr. Braswell: we aggregate and then compute don't try to do anything with ratios themselves, that won't work
Provided in the loss file. Mr. Harris: sum of incurred losses. Loss amount and transaction type.
Mr. Antley: sum the loss amt grouped by transaction type. Mr. Harris: a specific transaction type that is giving you incurred loss. Mr. Harris: sum of paid losses and outstanding losses, which is the loss amount based on transaction type.
Mr. Antley: Paid losses: combination of loss amount and transaction type
Mr. Antley: Incurred count = sum of claim count from loss records
Mr. Harris: Basic limit loss = sum of paid losses and outstanding losses for a limit = state minimum. How do we do this today? Mr. Antley: no comment today, will explore/comment next week
Mr. Harris: Excess limit loss = diff. btwn incurred loss and basic limit loss.
Mr. Antley: basic limit, pip excess limit. Will reach out to chief actuary @ AAIS on how this is figured.
Mr Harris: column i we need to add transaction type on Base 1.0. Suggested that we add a column btwn C&D. EP based on premium amount. Car years based on months covered. Loss amount. Claim count. Loss amount (basic limit loss), loss amount (excess limit loss), loss amount (basic limit), loss amount. (pip excess limit)
Mr. Harris: if we go back to his exhibit, how do we calculate WP? This points back to middle section which is essentially business rule (no different depending on what sorting and filtering criteria are). Will only change based on how info is delivered and how specific we want information to be. All written premium, but more finely tuned, but this can be day 2. This may be where Mr. Hamilton is going re: creating biz rules ard the aggregations.
Mr. Antley: what comes next? (Referenced implementation). Mr. Harris: we can change one of variables but it will be an iterative loop. When you get to bodily injury, deductible must be factored in. Comp and then collision. Mr. Antley: once we get into comprehensive how do #s changed? Mr. Harris: deductible = full. Mr. Antley: a secondary sort on top of physical damage. Same thing for collision? (Mr. Harris affirmed, yes this is correct).
Mr. Antley: Homeowners report and stat plan. Mr. Antley: what do we want to do next?
Mr. Sayers: get what we have working and then call this group to order based on issues that arise out of that work?
Mr. Antley: next week the RRDMWG should be more of a focused working group based on implementation. Not a mandatory meeting - not everyone needs to be there
Mr. Antley: nevertheless a touch point will be helpful next week. Mr. Harris affirmed that he cannot be here next week.
Mr. Antley: will he exclusively write the code on implementation or will others be involved as well? Happy to make an attempt and have an update next Friday. Mr. Sayers: available to help as well
Consensus is that the discussion will resume on this point next week
C. Adjournment - Meeting adjourned at 1:45 EST with plans to resume discussion on 8/6/22 at 1pm EST.