Synpute Loss Index
Public data on compute credit.
Default, recovery and time-to-re-lease data across registered clusters, published with its methodology. The loss history lenders, insurers and rating agencies need before compute risk can be priced like any other.
Loss IndexB200 · all regions
- Default rate
- —
- Recovery
- —
- Re-lease
- —
% / yr
% of exposure
days, median
Methodology first
No data yet. Figures publish once registered clusters have enough history.
Cohorts: hardware generation · site · tenant type
Why it matters
Lenders are asked to price a risk nobody has measured.
Older credit markets rest on decades of default and recovery studies. Compute has no public loss data, yet loans already outlive the contracts behind them.
5 years
Maturity of CoreWeave's DDTL 5.5
Source: Motley Fool, 2026
~3 years
Average term of the contracts behind it
Source: Motley Fool, 2026
Methodology
From registered clusters to a published index.
Built from the registry, not from surveys. Every event in the index happened to a verified cluster under a recorded contract.
01Registry
Register
Every cluster in the index is registered on a Synpute ID, with verified parties and contracts.02Synpute Monitor
Record events
Monitor records payments and performance; defaults, Transfers and re-leases are logged as events on the same record.03Cohorts
Group
Events are grouped by hardware generation, site and tenant type, so like is compared with like.04Anonymized
Aggregate
Figures are aggregated and anonymized; no operator, buyer or contract can be identified.05Versioned
Publish
Each release states its methodology version. The methodology is reviewed by an advisory board of lenders, operators and buyers.
Sample metrics
What the index measures.
The definitions the index will publish against. Values are placeholders: there is no Loss Index data yet.
01% of contracts / yr
Default rate
02% of exposure
Recovery rate
03days, median
Time to re-lease
04% of prior rate
Replacement rate
Placeholder values. No Loss Index figures have been published; cohort definitions may change during methodology review.
Subscription tiers
Access, by how you use the data.
Indicative pricing, to be confirmed at launch. The Loss Index is a Phase 2 product, built from the data Phase 1 financings put on the registry.
Anyone
Published summary
Free
- Headline default, recovery and time-to-re-lease figures
- Methodology notes and version history
Lenders, insurers, investors
Data subscription
$75K–$120K
per institution a year
- Loss Index by cohort: hardware, site and tenant type
- Synpute Price Reference marks
- Registry access for comparable data
Rating agencies and analysts
Data license
On request
- Loss Index data feeds
- Synpute Standard data feeds to build criteria on
Indicative prices, not an offer. Final tiers and pricing are set at launch.
Join the waitlist
Be first to see the data.
Waitlist members receive the methodology notes first and get first access when the index publishes.
What happens next
- 1We confirm your place within one business day.
- 2You receive methodology notes as they're published.
- 3First access to the index and subscription terms.
FAQ
Questions we hear
Straight answers to what comes up in the first call. If yours isn't here, ask us directly.
05 questions · reply within one business day
Compute credit needs a loss history.
Every financing on the registry adds to the history. The sooner it starts, the sooner compute credit has the data other asset classes take for granted.