/ engineering & quantitative dispatches

dubstrata engineering & causal systems blog

architectural releases, live telemetry specifications, and quantitative engineering notes from the builders of Dubstrata Causal-FM and the 3-Tier Verified Causal Knowledge Graph.

wss://api.dubstrata.com/api/v1/ws/accuracy16-channel timesfm-3 + cg-lora
FULL-CASCADE AUDIT & FORWARD OUTLOOKLIVE DEPLOYMENTNovember 2026 • 10 min read

Anatomy of a -23.6% Cascade: Reviewing Every Leg of Bitcoin's November Drawdown & What Our Systems Project Next

Percentage-Normalized Review Across 13 Consecutive Sessions, 10 Intraday Dislocation Legs & Bias-Adjusted Multi-Horizon Projections

Across thirteen consecutive November trading sessions, Bitcoin shed -23.6% from peak to trough across ten distinct intraday dislocations before rebounding +5.4% off the low. Auditing our systems in percentage terms shows why our price-projection layer lagged the initial bull trap, how it locked onto the institutional ETF redemption cascade across subsequent legs (achieving a 100.0% 24h hit rate on the steepest breakdown session), and why our bias-adjusted curve favors -2.6% to +2.0% base-building over unadjusted momentum carryover.

Peak → Trough Move-23.6% (-20.9% Net)
Main Breakdown 24h Hit100.0% Directional
Certified External Share54.4% (80.7% Conf.)
Bias-Adjusted 24h Band-2.6% to +2.0%
Listing Identifier
Neudata Dataset #5101648
Compliance Posture
Neudata DDQ v1.1 • Zero-PII • RFC 3161
Data Representation
16-Channel Tensor X(t) ∈ ℝ^(16×512)
Authoring Desk
Dubstrata Quantitative Research

/ engineering dispatches

all technical articles & release notes

Filter:
full-cascade audit & forward outlookNovember 2026 • 10 min read
Dubstrata Quantitative Research

Anatomy of a -23.6% Cascade: Reviewing Every Leg of Bitcoin's November Drawdown & What Our Systems Project Next

Percentage-Normalized Review Across 13 Consecutive Sessions, 10 Intraday Dislocation Legs & Bias-Adjusted Multi-Horizon Projections

Across thirteen consecutive November trading sessions, Bitcoin shed -23.6% from peak to trough across ten distinct intraday dislocations before rebounding +5.4% off the low. Auditing our systems in percentage terms shows why our price-projection layer lagged the initial bull trap, how it locked onto the institutional ETF redemption cascade across subsequent legs (achieving a 100.0% 24h hit rate on the steepest breakdown session), and why our bias-adjusted curve favors -2.6% to +2.0% base-building over unadjusted momentum carryover.

#Drawdown Audit#Causal Attribution#Spot ETF Redemptions#Cross-Market Positioning#Calibration Analysis#Forward Outlook
/blog/btc-nov4-nov16-crash-forensic-audit-and-forecastread article
executive research note / out-of-sample auditNovember 9, 2026 • 5 min read
Dubstrata Quantitative Research

Why Bitcoin's Nov 5 Breakdown From $80,000 Was Visible Hours Before Spot Cracked

Full-Population Audit: N = 2,727 Sybil-Clean Feature Snapshots & N = 3,263 Multi-Horizon Forecasts (Nov 4–5, 2026)

Between November 4 and November 5, 2026, Bitcoin executed a textbook $80,000 round-trip. To a trader watching spot price or perpetual funding alone, the Nov 4 breakout and the Nov 5 morning consolidation looked structurally identical—under the surface, cross-market conviction in prediction markets (Q2) and knowledge-graph capital flow (Q3) had already inverted completely.

#Executive Research Note#16-Channel Causal Tensor#Spearman Rank IC#Q2 Prediction Markets#Q3 Causal Graph#Out-of-Sample Audit
/blog/btc-nov4-nov5-80k-roundtrip-forensic-tearsheetread article
institutional milestone / alternative dataOctober 21, 2026 • 4 min read
Dubstrata Quantitative Systems & Compliance

Dubstrata Is Now Listed on Neudata: Bringing Sybil-Scrubbed Prediction Market Consensus and Bi-Temporal Causal Graphs to the Institutional Buy-Side

Why most prediction market and event feeds get rejected during hedge fund onboarding—and how we engineered a zero-lookahead, DDQ-verified 16-channel tensor built for systematic desks.

Dubstrata’s Prediction Market Consensus and Causal Event Dataset (#5101648) is officially live on Neudata alongside our completed DDQ v1.1 and Risk Check—packaging four orthogonal layers into a stationary, point-in-time 16-Channel Multivariate Causal Tensor resampled at fixed 5-minute intervals.

#Neudata #5101648#16-Channel Causal Tensor#Neudata DDQ v1.1#Zero-PII#RFC 3161 Provenance#Sybil-Scrubbed CLOB
/blog/dubstrata-live-on-neudata-institutional-alt-dataread article
foundation models & online learningOctober 21, 2026 • 6 min read
Dubstrata Quantitative Systems Desk

Closing the Loop: Convergent Multi-Horizon Accuracy & Causally-Gated LoRA (CG-LoRA)

Why raw price proximity is an incomplete measure of model intelligence—and how we use 6-metric convergent settlement audits to gate 5-minute online LoRA adaptation across our 16-channel TimesFM-3 architecture.

Most quantitative forecasting systems evaluate accuracy in a vacuum. Today, we deployed our Convergent Multi-Horizon Accuracy Emitter (wss://api.dubstrata.com/api/v1/ws/accuracy) and Causally-Gated Low-Rank Adaptation (CG-LoRA) pipeline on top of our 16-channel TimesFM-3 foundation architecture—scoring 7 simultaneous horizons across 6 formal causal and accuracy diagnostics every 5 minutes.

#TimesFM-3#CG-LoRA#HRX Jacobian#Causal Attribution Veracity#Multi-Horizon Telemetry#WebSocket Stream
/blog/convergent-multi-horizon-accuracy-cg-loraread article
causal graph & systemic riskSeptember 21, 2026 • 4 min read
Dubstrata Quantitative Research Desk

Most Risk Models Don’t Fail Because the Math Is Wrong. They Fail Because the World Isn’t a Spreadsheet.

Why rear-view VaR dashboards and flat LLM news summarizers miss systemic supply chain and prediction market cascades—and how Dubstrata traces falling dominos while they're still in mid-air.

When a Tier-1 supplier in East Asia halts production due to an unseasonal reservoir freeze, traditional VaR dashboards show nothing—until 18 days later when regulatory gazettes and Polymarket whale wallets trigger an 8.4% market open shock. Here is why we built a sub-second, point-in-time causal knowledge graph.

#Causal Knowledge Graph#O(1) Graph Traversal#Prediction Markets#Supply Chain Risk#Whale Orderbook Vectors#Systemic Risk
/blog/spotting-falling-dominos-causal-risk-graphread article
/ formal institutional whitepapers

looking for our formal SSRN papers & prediction market post-mortems?

explore the full mathematical specifications for Dubstrata Causal-FM (SSRN: 7539038), the 3-Tier Cryptographically Verified Causal Knowledge Graph (SSRN: 7377639), and our empirical prediction market post-mortems.