KOPS Core

KOPS Core is the primary index feed product. It publishes continuous, constructed indices on-chain at a configurable cadence. Each index is synthesized from multiple independent signal layers — historical performance, market consensus, and live event data — combined via a weighted composition function.

Core feeds are suitable for any downstream application requiring a persistent price signal for a non-financialized domain: derivative protocols, structured products, risk management systems, analytics platforms.

KOPS Live

KOPS Live handles real-time event processing. When an event begins — a data release, an active event, a policy announcement — the Live layer activates automatically, injecting sub-second signals from live market sources into the index.

At event boundaries, state transitions are deterministic: the Live layer activates at event start and deactivates at settlement, with no manual intervention or ambiguous handoffs.

KOPS Settle

KOPS Settle is the verification and settlement layer. At each event conclusion, Settle executes the deterministic settlement function, applies the causal monotonicity clamp, and publishes the full settlement record to a public log.

Any observer can independently verify that:

• The settlement output matches the deterministic formula given the inputs
• No positive event produced a negative index change
• No negative event produced a positive index change

Design Principles

A KOPS index must satisfy constraints that traditional price oracles never face. There is no external "true price" to converge to — the index is the price. This imposes three requirements:

1. Constructive Integrity — Every index movement must trace to a verifiable real-world event.
2. Causal Monotonicity — Positive outcomes produce non-negative movement. Negative outcomes produce non-positive movement. Enforced by construction.
3. Continuous Arbitrageability — The index must remain anchored to external liquid markets at all times.

Multi-Layer Architecture

KOPS indices are constructed from multiple independent signal layers, each capturing a different information regime:

Multi-layer composition with monotonic price transformation

Historical Performance Layer

Tracks cumulative entity performance through an adaptive rating system. Ratings are solved simultaneously across the entire network, eliminating error propagation inherent in naive sequential update schemes.

Simultaneous equilibrium solve — unique fixed point by construction

Market Consensus Layer

Incorporates forward-looking information from distributed information aggregation mechanisms. Market-implied assessments are transformed into a representation compatible with the rating space via context-specific transformations.

Live Event Layer

During active events, real-time market signals provide continuous signal injection into the index. This layer activates deterministically at event boundaries and deactivates at settlement.

Settlement Mechanics

At event completion, the index undergoes a deterministic settlement update. The settlement function updates the posterior distribution over entity ratings conditional on observed outcomes and prior market-implied expectations.

The update rule decomposes into two components: an innovation term that captures the informational surprise of the outcome relative to the pre-event consensus, and a coherence term that enforces long-run consistency between independently maintained signal layers. The relative weighting of these components is calibrated per-domain based on the statistical properties of the underlying event distribution.

Deterministic settlement — parameterized per domain, constrained by causal monotonicity

Importantly, the settlement function satisfies the semimartingale property under baseline conditions — the expected index change is zero absent genuine informational surprise. This ensures that the index does not systematically drift in either direction, a necessary condition for fair downstream derivative markets.

Causal Monotonicity Enforcement

The cause-effect clamp is the signature property of KOPS indices:

if outcome = POSITIVE:
    ΔR ← max(ΔR, 0)

if outcome = NEGATIVE:
    ΔR ← min(ΔR, 0)

This guarantees that positive events can never decrease an entity's index value, and negative events can never increase it — regardless of market expectations.

Network Equilibrium Solve

Unlike sequential rating systems that update entities one at a time, KOPS solves for the entire network simultaneously. The solve finds the optimal rating configuration under the observed interaction graph, operating over a rolling window calibrated to each domain's characteristic timescale.

Price Transformation

Raw index values exist in an internal rating space whose units are not directly meaningful to market participants. The price transformation maps this latent space to a consumer-facing price space via a smooth, monotonic function chosen to satisfy several desirable properties.

The transformation belongs to the class of log-linear mappings — functions that are linear in log-price space, ensuring that equal movements in the latent rating produce equal percentage movements in price regardless of an entity's absolute level. This property is critical for uniform market microstructure across indices of varying magnitude.

Smooth monotonic transformation from rating space to price space

The transformation is parameterized by a domain-specific constant θ that controls the sensitivity of price to rating changes — effectively setting the volatility regime of the index. Key guarantees:

• Strict Positivity — The image of 𝓣 is contained in ℝ₊, ensuring prices remain positive for all rating values
• Homogeneous Volatility — Equal rating differentials produce equal percentage price differentials across the entire entity universe
• Semimartingale Preservation — If the latent process is a semimartingale, so is the transformed price process, ensuring compatibility with no-arbitrage pricing theory

Trust Model

Traditional oracles derive trust from redundancy — multiple nodes fetch the same price and an aggregation function produces a canonical value. KOPS indices have no external reference price. Trust is derived from three complementary mechanisms:

1. Constructive Verification

Every index update is deterministically traceable to its inputs. Given the public formula, public event outcomes, and public market data, any observer can independently reconstruct the index value.

2. Market Anchoring

The index is continuously anchored to the deepest information markets in the world. This is structurally analogous to how VIX is anchored to S&P 500 options — the index is constructed from market data, not discovered through trading.

3. Cause-Effect Guarantee

The causal monotonicity clamp is a hard constraint enforced on every settlement:

• An entity with a positive outcome will see its index increase or remain flat. Never decrease.
• An entity with a negative outcome will see its index decrease or remain flat. Never increase.

Verifiable by inspecting the settlement log — no statistical modeling, no probability, no trust in the oracle operator.

Manipulation Resistance

Mechanism	Description
Source Diversity	Multiple independent signal sources. No single source can unilaterally influence the index.
Bounded Sensitivity	Each layer has bounded influence. Maximum single-update movement is clamped.
Settlement Determinism	Post-event settlement is fully deterministic. No oracle discretion.

Settlement Log

All index data is published to a public settlement log:

Field	Description
entity_id	Unique identifier for the indexed entity
event_id	Identifier for the triggering event
pre_event_index	Index value before settlement
post_event_index	Index value after settlement
outcome	Verified event outcome
baseline_state	Pre-event index baseline
clamp_applied	Whether the causal monotonicity clamp was active
timestamp	Settlement timestamp

Feed Format

Each index update contains:

{
  "index_id":    uint16,   // Index identifier
  "price":       int128,   // Index value in price space
  "confidence":  uint64,   // Confidence interval
  "timestamp":   int64     // Unix timestamp
}

The confidence interval widens during high signal uncertainty and narrows during stable periods.

Update Cadence

Regime	Example	Cadence
Stable	Between economic releases, inter-event periods	30-60s
Approaching event	Pre-release positioning	30s
Live event	During active event, FOMC press conference	1-2s
Settlement	Post-event outcome incorporation	Once

Use Cases

• Derivative protocols — Oracle feed for futures, options, or structured products
• Structured vaults — Yield strategies referencing macro expectations or sector risk
• Risk management — Sovereign risk indices as inputs for on-chain insurance or hedging
• Data products — Analytics and research on continuous constructed indices
• Prediction aggregation — Composite signal from multiple prediction market sources

Index Catalog

Domain	Coverage	Status
Macro Expectations (Inflation)	CPI contracts, Fed funds futures	Research
Macro Expectations (Employment)	NFP contracts, claims data	Research
Sovereign Risk	Election markets, policy contracts	Research
Competitive Networks	Distributed market data, verified outcomes	Live

Macroeconomic Expectations

Sovereign & Political Risk

Competitive Networks

Term	Definition
Causal Monotonicity Clamp	Deterministic constraint ensuring positive events produce non-negative index changes and vice versa.
Composition Weights	Relative contribution of each signal layer to the final index. Fixed per domain.
Constructed Index	An index synthesized from multiple signal sources, vs. one that relays an existing price.
DIAM	Distributed Information Aggregation Mechanism. Any market aggregating beliefs through economic incentives.
Live Event Layer	Signal layer activating during real-time events with sub-second live market pricing.
Market Consensus Layer	Signal layer incorporating forward-looking probabilistic assessments.
Network Equilibrium Solve	Simultaneous optimization finding the unique set of entity ratings consistent with all observed outcomes.
Oracle Cycle	Complete processing loop from signal ingestion to index publication. Default: 30s.
Performance Layer	Signal layer tracking cumulative entity performance via adaptive rating updates.
Price Transformation	Monotonic function mapping rating space to price space. Ensures positivity and equal-percentage sensitivity.
Settlement	Deterministic incorporation of event outcomes into the index, including the cause-effect clamp.
Signal Layer	Independent information channel contributing to the constructed index.
Signal Source	External data provider contributing raw information to the oracle pipeline.