Credit scores run from 0 to 100. Each score maps to a rating band using fixed thresholds derived from the model design. The mapping is transparent and consistent — no committee discretion, no black box.
Investment Grade
Speculative Grade
Score Explorer
Drag the slider to see how a score maps to a rating in real time.
Score
65
Rating
Position
Grade
Investment
Sub-Parameter Scoring Anchors
All sub-parameters are scored on a 1–5 integer scale. Anchors are defined to reduce ambiguity and improve consistency across participants.
1
Severe Weakness
Material default risk, near-insolvent metrics, or total institutional breakdown. Reserved for the most distressed sovereigns.
Immediate creditworthiness concern. High risk of structural failure or imminent default.
2
Structural Weakness
Elevated but manageable risk. Identifiable structural problems without immediate default threat. Requires monitoring.
Chronic vulnerability. Persistent fragility that, without correction, could escalate toward distress.
3
Mixed / Transitional
Ambiguous profile. Some positive indicators offset by meaningful concerns. Trajectory matters more than current position.
Coexisting strengths and weaknesses. Direction is unclear — could improve or deteriorate.
4
Structurally Sound
Good underlying position with no systemic threat. May have isolated vulnerabilities but overall trajectory is stable or improving.
Solid fundamentals. Identifiable risks are manageable and do not threaten creditworthiness in the near term.
5
Very Strong / Resilient
Exceptional performance. Very low risk, strong institutional buffers, and high capacity to absorb shocks. Rare — reserved for top-tier sovereigns.
Best-in-class. Minimal vulnerability and highly resilient to external shocks or policy stress.
Outlook Model
Participants vote a directional expectation for each category: improving (+1), unchanged (0), or deteriorating (−1). The weighted average of votes determines the overall outlook.
↑
Positive Outlook
Weighted average > +0.20
The aggregate view expects conditions to improve over the medium term.
→
Stable Outlook
Between −0.20 and +0.20
No clear directional signal. Conditions expected to remain broadly unchanged.
↓
Negative Outlook
Weighted average < −0.20
The aggregate view expects conditions to deteriorate over the medium term.
Confidence Index
Each submission includes a self-reported confidence level indicating how well-evidenced and stable the analyst's view is. This feeds into the rating's overall confidence index alongside participation depth and score stability across revisions.
Low
Limited Evidence Base
The analyst has a directional view but limited access to primary data, or the situation is rapidly evolving. Score should be treated as indicative and weighted accordingly in aggregation.
Moderate
Reasonable Evidence Base
The view is grounded in credible sources with some gaps. The analyst is broadly comfortable with the assessment but acknowledges meaningful uncertainty. This is the default for most well-informed submissions.
High
Strong Evidence Base
The assessment is backed by primary sources, direct country knowledge, or strong quantitative evidence. The analyst is highly confident in both the score and its stability over the near term.
Confidence levels also feed into the rating's displayed confidence band — when the majority of submissions carry Low confidence, or disagreement is high, the rating displays a wider uncertainty range.
How a rating is produced
Each rating is produced by aggregating independent analyst assessments through a transparent weighted formula. Analysts score up to 36 sub-parameters across seven categories. Each category carries a fixed weight; sub-parameter scores are averaged within each category, then combined into a final score from 0–100 mapping to one of 17 rating bands from AAA to CCC.
Expert submissions carry full weight; quick submissions carry half weight. Four automatic guardrails protect against edge cases where mechanical aggregation would produce a misleading result. All thresholds and rationale are documented below.
Technical Specification
1. Category Weights
Seven categories cover the full sovereign credit profile. Weights reflect the relative contribution of each dimension to default probability, informed by academic literature and agency practice.
Category weights
A
Fiscal Strength
20%
D
Institutional & Political
20%
B
External Liquidity
15%
C
Growth & Resilience
15%
E
Financial Sector
10%
F
Default Willingness
10%
G
Event Risk
10%
2. Score Calculation
Sub-parameter scores within each category are averaged to a category score. Category scores are combined using a renormalised weighted sum, then converted to the 0–100 scale.
Calculation steps
1
Average sub-parameters
Mean score of all rated params within a category → Ck on a 1–5 scale
2
Weighted combination
Renormalised weighted sum of Ck across scored categories → rawScore (1–5)
3
Normalise to 0–100
Linear map: 1→0, 3→50, 5→100. Anchors the scale symmetrically around the midpoint.
3. Submission Weighting
Expert submissions carry full weight; quick submissions carry half weight, reflecting shallower parameter coverage. This is not a judgement on analytical quality.
Weights by submission mode
Expert
w = 1.0
Full parameter coverage
Quick
w = 0.5
Shallower coverage
Half-weight is not a quality judgement — it reflects that quick submissions cover fewer sub-parameters and carry proportionally less information.
4. Outlook Signal
Each category carries an outlook vote (+1 Positive, 0 Stable, −1 Negative). The signal is category-weight-aware, so higher-weighted categories (A, D) count more. Thresholded to produce a three-state output. Quick-mode submissions provide a single overall vote rather than per-category votes; this is treated as a unit-weight signal by design.
Three-state output
O > +0.20
−0.20 to +0.20
O < −0.20
Positive
Improving trajectory expected
Stable
No directional consensus
Negative
Deteriorating trajectory expected
Votes: improving = +1 · unchanged = 0 · deteriorating = −1
5. Confidence Index
Derived from self-reported level, a depth bonus for scoring all categories, and a revision bonus for deliberation. Confidence reflects observable behaviour, not only self-assessment.
Base values
Low
Expert
0.40
Quick
0.35
Moderate
Expert
0.65
Quick
0.55
High
Expert
0.90
Quick
0.80
Behavioural bonuses
All 7 categories scored (Expert mode)
+0.05
Score revised during deliberation
+0.03
6. Disagreement Index
Measures the spread of analyst opinions using submission-weighted variance. Expert disagreement carries more weight than quick-submission disagreement.
Disagreement bands
σ > 12
6 < σ ≤ 12
σ ≤ 6
High
Structurally uncertain rating
Moderate
Some divergence of opinion
Low
Strong analyst consensus
High disagreement does not invalidate a rating — it signals material divergence of expert opinion, which is analytically meaningful.
7. Guardrails
Four rules fire automatically when threshold conditions are met. They protect against edge cases where mechanical aggregation would produce a misleading result.
G1
Rating Cap
catScore[F] < 2.0 or catScore[G] < 2.0 → cap at BB+
A sovereign with strong fiscal metrics but demonstrated default willingness cannot hold investment grade. The 2.0 threshold marks where structural concern becomes disqualifying. Applied at aggregate level for published ratings and per-submission for individual scores.
▲ Rating Capped
G2
High Uncertainty
disagreement = High AND outlook = Negative → “Uncertain”
Either condition alone may reflect analytical complexity. The combination signals a declining trajectory assessed with material expert divergence.
⚠ High Uncertainty
G3
Indicative
submissions < 15 → labelled “Indicative”
Below 15 submissions the aggregated view is not statistically meaningful as a crowd-sourced signal.
● Indicative
G4
Minority View
expertShare < 0.55 → minority view note recorded
Flags when quick submissions drive a disproportionate share of the aggregate, reducing the effective weight of expert assessment. Requires a minimum of 3 submissions — countries with fewer are not flagged to avoid noise from near-empty samples.
⚖ Minority View
8. Known Limitations
Ratings are produced by a transparent model but inputs are human judgements. The disagreement index and confidence score are designed to surface — rather than mask — the cases where these effects are most likely to be material.
Structural
Participation Bias
Self-selection by country prominence — smaller sovereigns attract fewer submissions and higher individual-analyst variance, making their ratings less stable. Mitigated by the G3 Indicative guardrail.
Behavioural
Agency Anchoring
Analysts may unconsciously anchor to Moody’s / S&P / Fitch ratings despite the parameter-first guidelines, compressing the spread of independent views.
Temporal
Recency Bias
Over-weighting of recent events relative to structural fundamentals. Partially mitigated by the deliberation window, which allows scores to be revised in light of peer submissions.
Social
Herding / Deliberation Influence
During deliberation, analysts can see peer scores. This may suppress independent views even while appearing to reduce disagreement — revisions reflecting conformity rather than genuine updating of beliefs.
Design
Category Weight Subjectivity
The category weights (A=20%, D=20%, B=C=15%, E=F=G=10%) are design choices that embed a prior about what drives default risk. Different weight schemes would produce different ratings — treat these as a modelling assumption, not an empirical finding.
“These limitations are inherent to any crowd-sourced analytical system. They are disclosed here so users can apply appropriate judgement when interpreting ratings.”
Not Investment Advice. Ratings are crowd-sourced assessments produced by the TITAN division at the Atlas Institute. They do not constitute investment advice, are not produced by a regulated credit rating agency, and should not be used as the sole basis for investment decisions.
What you're doing
Contributing to a crowd-sourced sovereign rating
You are contributing to a crowd-sourced sovereign credit rating — scoring a country's likelihood of meeting its debt obligations across seven weighted categories. Each submission is combined with others to produce an aggregate that is independent of the three major agencies.
7
Weighted categories
Fiscal, governance, monetary, external and more
1–5
Scoring scale
Current fundamentals only — trajectory captured via outlook
AAA–CCC
Output rating
Aggregate converted to a standard notch scale
Scoring reference
The 1–5 scale
Score current fundamentals — where the country stands today, not where it is headed. Trajectory is captured separately via the outlook vote.
1
Severe weakness
Material default risk or near-total institutional breakdown
2
Structural weakness
Elevated risk without immediate default threat. Requires monitoring.
3
Mixed / transitional
Coexisting strengths and weaknesses. Direction unclear.
4
Structurally sound
Good position, isolated vulnerabilities. Stable or improving.
5
Very strong
Exceptional fundamentals. Rare — reserved for top-tier sovereigns.
When in doubt between two scores, choose the lower one. A 3 that captures genuine ambiguity is more valuable than a forced 4.
Submission type
Which mode should I use?
Expert
Full weight · w = 1.0
For analysts, researchers, and practitioners with direct country knowledge. Scores all 36 sub-parameters individually across seven categories, with an outlook vote per category. Includes a structured deliberation phase.
Quick
Half weight · w = 0.5
For informed observers who want to register a view without scoring every parameter. Provides a single overall score per category. Takes under two minutes. Carries half weight because it covers shallower depth — this is not a quality judgement.
Expert
Step by step
1
Select the sovereign you are assessing from the map or country list.
2
Score each of the seven categories (A–G) and their individual sub-parameters on the 1–5 scale. Add an outlook vote (+1 / 0 / −1) per category.
3
Set your confidence level (Low / Moderate / High) reflecting how well-evidenced your view is.
4
During the deliberation window you can see the aggregate distribution and revise any score once. This is optional — you are not required to move toward consensus.
5
Submit. Your assessment enters the aggregate at full weight (w = 1.0).
Good practice: Score current fundamentals, not trajectory — outlook is captured separately. Avoid anchoring to published agency ratings.
Quick
Step by step
1
Select the sovereign from the map or country list.
2
Give a single overall score (1–5) for each of the seven categories. You are rating your overall view of that category, not individual sub-parameters.
3
Set your overall confidence level (Low / Moderate / High) and indicate the overall outlook direction.
4
Submit. Your view enters the aggregate at half weight (w = 0.5) and feeds into the disagreement and confidence indices.
Good practice: Only submit on sovereigns where you have genuine familiarity. A quick submission from someone with real regional knowledge is more valuable than a speculative one.
After you submit
What happens to your score
Immediately
Your score is incorporated into the aggregate on the Ratings page. A confirmation screen shows how your rating compares to the current consensus. Ratings update in real time.
Updating a submission
Requires a signed-in account. You can re-submit for the same country at any time — your most recent submission replaces the previous one in the aggregate. Older versions are stored for transparency but not counted.
Ready to contribute?
Submissions are open to all — no account required. Create an account if you want to track your history, view your past assessments, and update your scores over time.
Submission standards
All submissions are governed by the Ratings Integrity framework. Participants should not submit on sovereigns where they hold a direct financial position, act in an advisory capacity to the government, or have another conflict of interest that could compromise objectivity. The Atlas Institute reserves the right to exclude submissions that show evidence of coordination or deliberate anchoring to published agency ratings.
Common questions about how TITAN Ratings works, who can contribute, and how to interpret what you see.
About the rating
What is TITAN Ratings? ▾
TITAN Ratings is a sovereign credit rating application built by the Atlas Institute. It produces crowd-sourced sovereign credit ratings by aggregating assessments from analysts and informed observers worldwide. Unlike traditional agency ratings, every component score is explicit, weighted, and independently verifiable.
How is this different from Moody's, S&P, or Fitch? ▾
Traditional agencies produce ratings through internal committees whose deliberations are confidential. TITAN Ratings makes the analytical framework fully transparent — every parameter, its weight, and the aggregate score behind any rating is visible. Ratings are also produced by a distributed community of analysts rather than a single institution, which reduces concentration risk and ideological uniformity.
How often are ratings updated? ▾
Ratings update in real time as new submissions are incorporated into the aggregate. There is no fixed review calendar — the rating reflects the aggregate of all current assessments at any given moment. If a significant event occurs, participants can revise their scores immediately, and the aggregate will update accordingly.
What does the Outlook mean? ▾
The outlook — Positive, Stable, or Negative — reflects the direction participants expect conditions to move over the medium term. It is separate from the current score and is derived from weighted outlook votes submitted alongside each assessment. A Negative outlook does not mean the rating will necessarily change, but signals the aggregate view expects deterioration.
What is the Confidence Index? ▾
The confidence index reflects how well-evidenced the aggregate rating is. It combines three signals: the number of participants who have submitted, the average self-reported confidence level of those submissions, and the stability of scores across revisions. A low confidence index means the rating is based on fewer or less certain views and should be treated with more caution.
The methodology
How does a score become a rating? ▾
Each participant scores 7 categories on a 1–5 scale. Category scores are multiplied by their weights and aggregated into a composite score between 0 and 100. That score is then mapped to a rating band — for example, a score of 70–74 corresponds to an A rating. The full mapping is shown in the Rating Scale tab.
Why are some categories weighted more than others? ▾
The weights reflect the empirical importance of each dimension to sovereign creditworthiness, calibrated against historical default and distress events. Fiscal Strength and Institutional & Political Strength each carry 20% because they are the most consistent predictors of sovereign risk over the medium and long term. Weights are fixed across all sovereigns — they do not vary by country or region. Full category definitions are in the Parameters tab.
Why does Ratings include "Default Willingness" as a category? ▾
Traditional agencies treat willingness to pay as a qualitative overlay rather than a scored variable. Ratings makes it explicit because the historical record shows that some sovereigns have chosen not to honour obligations even when fiscally capable of doing so. Separating ability to pay from willingness to pay produces a more complete picture of credit risk.
What is the difference between Expert and Quick mode? ▾
Expert Mode scores all 36 sub-parameters individually and carries full weight (w = 1.0) in the aggregate. Quick Mode provides a single score per category and carries half weight (w = 0.5). The weight difference reflects depth of analysis, not the quality of the participant. A Quick submission from someone with deep regional knowledge is still valuable — it contributes a directional signal and feeds the confidence index. See the Scoring tab for anchor definitions.
Contributing
Who can submit a rating? ▾
Anyone can submit — no credentials or institutional affiliation are required. The system is designed so that analytical quality is reflected through confidence levels and score consistency over time, not through gatekeeping. That said, participants are expected to submit only on sovereigns they have genuine familiarity with.
Do I need an account? ▾
No account is required to submit a rating. However, creating an account lets you track your submission history, update scores over time, and see how your views compare to the aggregate across revisions. Without an account, your submission is counted but cannot be revised.
Can I change my score after submitting? ▾
Yes, but you need a signed-in account to do so. You can re-submit for the same sovereign at any time — your most recent submission replaces the previous one in the aggregate. All previous versions are stored for transparency and are visible in your submission history, but only the most recent counts.
What if I have a conflict of interest? ▾
You should not submit on sovereigns where you hold a direct financial position, act in an advisory capacity to the government, or have any other relationship that could compromise objectivity. TITAN reserves the right to exclude submissions that show evidence of bias or coordination. If you are uncertain whether a relationship constitutes a conflict, err on the side of not submitting.
What is the deliberation phase? ▾
Expert mode submissions include an optional deliberation window during which participants can see the aggregate score distribution and revise any single score once. Full steps are in the Analyst Guide tab. The purpose is to allow genuine reconsideration in light of peer views — not to encourage convergence. You are not required to move toward consensus, and there is no penalty for diverging from the aggregate.
Interpreting ratings
Are TITAN Ratings comparable to agency ratings? ▾
The rating scale is structurally equivalent — AAA through CCC uses the same notation as Moody's, S&P, and Fitch. See the Rating Scale tab for the full band mapping. However, TITAN Ratings reflect the aggregate view of the participant community and may diverge from agency ratings, particularly where the community has access to information or perspectives not fully reflected in published ratings. Divergence is itself informative.
What does a wide disagreement range mean? ▾
Some rating pages display a disagreement range showing the spread of participant scores. A wide range indicates that participants hold meaningfully different views on the sovereign's creditworthiness — which may reflect genuine uncertainty, rapidly changing conditions, or differing analytical frameworks. It is a signal to look more closely, not a flaw in the system.
Can I use TITAN Ratings for investment decisions? ▾
Ratings are provided for informational and analytical purposes only. They do not constitute investment advice, and ATLAS makes no representation that they are suitable for use as the basis for investment decisions. Users should conduct their own due diligence and consult qualified advisors before acting on any information presented in the application.
How do I delete my account? ▾
To request account deletion, email info@atlasinstitute.org from the address associated with your account. We will delete your profile and remove your personal information. Anonymous submission data may be retained in aggregate for research integrity purposes.
Need more information? For any questions not covered above, contact us at info@atlasinstitute.org.