Risk Tiers · Underwriter

Risk tiers & discount bands

A gate plus two paths, not a weighted blend: we reward being healthy, or getting healthier, and you only need one.

Provisional · 2026-Q3, freezes at period close
Tier distribution
96 of 120 monitored members are rated · 24 Unrated
0%2.5%5%7.5%10%No score
Premium discount per band
Score composition
Luna Protect Score = the higher of Attainment and Improvement
Peak Consistency percentile45%
Behaviour — activity, sleep and recovery, ranked within this insurer’s book.
VO₂ Max decile35%
Cardiorespiratory capacity, normed for age and sex.
Resting HR percentile20%
Cardiovascular baseline, normed for age and sex; lower resting HR ranks higher.

Drag to re-weight · normalised to 100% · feeds the tier simulator below

Eligibility gate
80%

Qualifying days in the period. A data-validity check, never scored. Below it a member is

Unrated not in tier 1
Two-path rule

We reward being healthy, or getting healthier, and you only need one.

Device compliance cannot buy a discount: it only decides whether the data is trusted.

About to move

Within 4 points of the next tier up on the provisional score — the only tier signal still actionable before 2026-Q3 closes.

18 members within reach of an upgrade
MemberCurrent tierPoints to next tierQualifying daysPathSuggested action
Samuel Koh
45M · Priority · Singapore
3096%ImprovementRecognise the progress; a challenge keeps the trend going to period close
Terence Ng
47M · Priority Private · India
2096%ImprovementRecognise the progress; a challenge keeps the trend going to period close
M100273
Data sharing not enabled
21100%AttainmentMention what the next tier unlocks at renewal
Lina Pillai
50M · Priority · Hong Kong
2289%AttainmentMention what the next tier unlocks at renewal
Clarice Ng
40F · Priority · Singapore
22100%AttainmentMention what the next tier unlocks at renewal
Elaine Bhatia
45M · Priority Private · Singapore
12100%AttainmentMention what the next tier unlocks at renewal
Aisha Iqbal
39M · Priority Private · India
3286%ImprovementRecognise the progress; a challenge keeps the trend going to period close
Daniel Wong
55F · Priority · Singapore
1293%AttainmentMention what the next tier unlocks at renewal
Zhao Chua
47F · Private Banking · Singapore
42100%AttainmentMention what the next tier unlocks at renewal
M100658
Data sharing not enabled
2296%ImprovementRecognise the progress; a challenge keeps the trend going to period close
Shreya Hassan
50F · Private Banking · Singapore
4296%AttainmentMention what the next tier unlocks at renewal
Yusof Lim
41F · Priority · India
3396%AttainmentMention what the next tier unlocks at renewal
Cheryl Tan
40M · Priority Private · Singapore
2393%AttainmentMention what the next tier unlocks at renewal
M100553
Data sharing not enabled
3386%Run rate 79%ImprovementWear reminder first — the gate is at risk before the upgrade lands
Elaine Seah
27M · Priority · Hong Kong
24100%AttainmentMention what the next tier unlocks at renewal
Gopal Chandran
30M · Priority Private · India
3493%ImprovementRecognise the progress; a challenge keeps the trend going to period close
Priya Pillai
43F · Private Banking · Singapore
2486%AttainmentMention what the next tier unlocks at renewal
M100693
Data sharing not enabled
4493%AttainmentMention what the next tier unlocks at renewal

Plausibility checks

Loading daily data…

This is detection, not a determination. These are rule-based checks computed from daily aggregates, surfacing implausible patterns for human review — never a fraud finding on their own. A flag never voids or changes a tier.
Steps without sleep record
…
members flagged

Steps recorded but no sleep record for 7+ consecutive days

Action · Flag for review
Implausible step count
…
members flagged

Daily steps above 60,000

Action · Day excluded from qualifying days
Steps / Activity score mismatch
…
members flagged

15,000+ steps with an Activity score under 40, or under 1,000 steps with a score over 80

Action · Flag for review
Baseline discontinuity
…
members flagged

Median resting HR shifts more than 8 bpm between adjacent 7-day windows, with no transition

Action · Flag for review

Path split monitor

Share of upgrades earned on the improvement path. A low share means the scheme is rewarding only the already healthy.

Upgrades via improvement
65%
11 of 17 upgrades vs 2026-Q2
Low-water mark
40%
Tenant-configurable alert threshold
Status
Within range

Members are qualifying by getting healthier, not only by already being healthy.

Tier simulator

Underwriter only · non-destructive. Nothing here writes to live tiers.

Settings
Eligibility gate80%
Tier count5
Maximum discount10%
T1 · 0%T2 · 2.5%T3 · 5%T4 · 7.5%T5 · 10%
Input weights · normalised to 100%
Liability change
Modelled
−$3,310
/ yr
$35k → $32k
Discount up
0
members
Discount down
18
members
Improvement path
48%
of rated · live 48%
Redistribution
At live settings the simulated book matches today
5 tiers → 5 tiers
LiveSimulated
Attainment is recomputed from its inputs — Peak Consistency percentile within the whole monitored book, VO₂ decile and resting HR percentile — and reproduces 102 of 120 stored tiers at live settings. On the Improvement path only the Peak Consistency 90-day change is in the sample; the VO₂-decile and RHR-percentile changes are held at the 50th percentile, so reweighting moves their share of the score but not their rank.
Modelled Liability uses modelled premiums per band.Lower discount spend than live.