Skip to content

faq.md

Latest commit

 

History

History
237 lines (145 loc) · 5.39 KB

faq.md

File metadata and controls

237 lines (145 loc) · 5.39 KB

Sentinel Risk OS — FAQ

This FAQ addresses common questions about the Sentinel Risk OS Results repository, its scope, and how the published outputs should be interpreted.

What is Sentinel Risk OS?

Sentinel Risk OS is a private, deterministic engine designed to detect structural risk regimes and stress accumulation in time-series systems.

This repository publishes results only — reports, CSV exports, and derived summaries — without exposing proprietary internals.

Is the engine open source?

No.

The engine is intentionally kept private.

This repository exists to publish observable outputs and allow independent inspection of their structure and consistency, not to share implementation details.

What problem does Sentinel Risk OS solve?

Most risk tools answer:

“How stressed is the system right now?”

Sentinel Risk OS answers:

“What structural phase is the system in, how did stress form, how long did it persist, and how did it transition?”

It operates at the regime level, not at the signal or prediction level.

Is this a trading or prediction system?

No.

Sentinel Risk OS:

  • does not generate trade signals
  • does not forecast prices
  • does not optimize returns
  • does not provide entry/exit recommendations

Its purpose is structural risk interpretation, not trading execution.

How is this different from VIX, OFR Financial Stress Index, or similar indicators?

Indices such as VIX or OFR FSI provide scalar measurements of stress intensity at a given point in time.

Sentinel Risk OS:

  • detects regimes
  • aggregates behavior into coherent stress windows
  • identifies onset, persistence, peak, and transition

The approaches are complementary, but not interchangeable.

Is Sentinel just another index with labels applied?

No.

Regimes and windows in Sentinel Risk OS are constructed structural objects, not thresholds applied to a continuous index.

The system explicitly models temporal persistence and transitions, which scalar indices do not.

Can the same results be obtained by thresholding existing indices?

No.

Thresholding an index:

  • introduces arbitrary cutoffs
  • fragments continuous behavior
  • removes temporal coherence

Sentinel windows are formed through structural aggregation, not post-hoc filtering.

Does Sentinel Risk OS use machine learning?

The published results do not depend on:

  • supervised learning
  • probabilistic modeling
  • neural networks

The internal engine architecture is not disclosed.

Are the results reproducible?

The engine outputs are not reproducible without access to the private system.

However:

  • all derived metrics
  • all figures
  • all run summaries

can be regenerated deterministically from the published CSVs using the provided verification scripts.

Why publish results without the model?

This repository follows a results-first validation model.

The goal is to allow independent inspection of:

  • observable behavior
  • regime structure
  • temporal consistency
  • cross-asset coherence

without requiring access to proprietary internals.

This mirrors how advanced research systems are often evaluated: evidence before explanation.

What do the regime labels mean?

  • Stable — structurally normal conditions
  • Pre-Stress — early structural deviation
  • Core Stress — sustained stress accumulation
  • Extreme — high-intensity structural instability

Regimes describe structural phases, not predictions or outcomes.

Are regime labels asset-specific?

No.

Regime labels are structural abstractions and are applied consistently across assets.

This allows cross-asset comparison without redefining semantics per market.

Why focus on daily resolution?

Daily resolution provides a balance between:

  • noise reduction
  • structural persistence
  • interpretability

The system is designed to capture regime-level behavior, not intraday dynamics.

Does Sentinel Risk OS generalize beyond financial markets?

The engine is designed around structural detection principles, not asset-specific heuristics.

This repository, however, is intentionally scoped to financial time-series results only.

Is Sentinel Risk OS intended to replace existing risk tools?

No.

Sentinel Risk OS can be used:

  • alongside traditional indicators
  • as a higher-level structural layer
  • as a contextual framework for interpretation

It does not aim to replace monitoring or reporting tools.

What should I look for when reviewing the reports?

Focus on:

  • duration of stress windows
  • internal peak timing
  • regime transitions
  • persistence versus transience

The value lies in structure, not isolated values.

Where can I find the latest reports?

Published reports are available via GitHub Pages:

https://johnoliveiradev.github.io/Sentinel-Risk-OS-Results/

What is the long-term roadmap?

The current phase focuses on:

  • single-asset analysis
  • daily resolution
  • public, verifiable outputs

Future phases may include:

  • cross-asset interaction
  • extended validation
  • external audit frameworks

Details will be published only when appropriate.

Final note

This repository exists to publish what the system observes, not how it works.

Interpret the outputs as structural evidence, not forecasts.

Evidence before explanation.