Intelligence Assessment ACH + Horizon Scan Unclassified

Current Strategic Forecast:

Analysis of competing hypotheses across 20 pieces of evidence, combined with an 11-signal horizon scan of emerging indicators.

Date: 01 APR 2026 Method: ACH + Horizon Scan Evidence items: 20 Horizon signals: 11
\Analysis of Competing Hypotheses

Hypothesis scores

Five competing hypotheses were tested against 20 items of evidence (OSINT and GEOINT). Each piece of evidence was scored for consistency (C/CC) or inconsistency (I/II) with each hypothesis, weighted by credibility and relevance. Higher scores indicate more evidence support.

#5 — Least supported
H5: Gray-zone escalation / accidental conflict
−14.2
#4
H1: No significant action in 2026
9.4
#3
H4: Blockade or quarantine
28.5
#2
H2: Coercion without invasion
33.7
MOST SUPPORTED
#1 — Best supported
H3: Continued military build-up / preparation
64.2

Key finding: The evidence overwhelmingly supports H3 as the leasing hypothesis, indicating that China may engage in an accelerating military build-up and preparation phase but is not on a path to invasion in 2026. H2 (coercive pressure without kinetic action) is the second most-supported hypothesis, consistent with Beijing's preference for political and economic coercion.

The strongly negative score for H5 (gray-zone escalation / accidental conflict) is notable: while the risk of miscalculation has risen sharply following the CMC purges, the structured evidence does not support an uncontrolled escalation trajectory as the most probable outcome. The near-zero score for H1 suggests that pure status quo is equally implausible. The strategic environment is shifting materially even if invasion is not imminent.

ACH Evidence Matrix

Evidence consistency ratings

Each cell records whether a piece of evidence is consistent (C), strongly consistent (CC), inconsistent (I), strongly inconsistent (II), or neutral (–) with each hypothesis. Evidence weighted by credibility × relevance.

CC
Strongly consistent
C
Consistent
II
Strongly inconsistent
I
Inconsistent
Neutral / not applicable
Cat. Evidence summary H1
No action
H2
Coercion
H3
Build-up
H4
Blockade
H5
Accidental
Horizon Scan

Signal register: 11 key indicators

Signals ranked by composite score (Impact × Likelihood × Novelty). Priority classification reflects urgency for decision-maker attention within the 9-month horizon (April–December 2026).

Part II — Signal Priority Ranking

Composite scores by signal

Analyst Assessment

Overall judgment: invasion in 2026 is unlikely

The combined weight of the ACH matrix and horizon scan supports a high-confidence assessment that a deliberate, coordinated PLA invasion of Taiwan in 2026 is unlikely. The most probable near-term trajectory is continued coercive pressure, including intensified exercises, gray-zone maritime operations, and political interference, with the PLA continuing to build and rehearse amphibious strike capabilities for a post-2027 window.

Three structural factors dominate the assessment: (1) the CMC purges have created genuine command paralysis, reducing the near-term viability of a coordinated joint campaign; (2) Beijing's 15th Five-Year Plan reveals a leadership preference for long-term economic and technological coercion over costly kinetic action in 2026; and (3) the Trump-Xi diplomatic context, while introducing deterrence risks, simultaneously reduces Beijing's incentive to act before summit-derived political gains can be consolidated.

The primary risk within the 9-month horizon is not invasion but compounding deterrence erosion: Taiwan's defense budget deadlock, US arms sale delays, and Operation Epic Fury munitions depletion are creating a structural vulnerability window that Beijing will exploit through political means.

Methodology note: ACH scores computed using a credibility-weighted consistency matrix. Credibility: High = 2.0, Medium = √2 ≈ 1.414, Low = 1.0. Relevance multiplier: High = 2.0, Medium = 1.0. Rating values: CC = +4, C = +2, N = 0, I = −2, II = −4. Horizon scan composite = Impact × Likelihood × Novelty (each scored 1–5). All evidence sourced from open-source intelligence as of 1 April 2026. This assessment is unclassified and produced for analytical purposes only.