AI Supply Chain Weather Map
Weather Map
supply-chain
momentum
geographic
Bottom Line
Loading weather assessment...
Real-time pressure, momentum, and turbulence across the AI supply chain,
derived from relative strength indicators and news sentiment.
▲
Turbulence
▼
Low Pressure · High Turbulence
Avoid
Dislocated and unstable. No edge, no trend. Wait for a catalyst or momentum inflection before committing capital.
High Pressure · High Turbulence
Reduce & Hedge
Strong but volatile—storm conditions. Take partial profits, tighten stops, and expect sector rotation within 2–4 weeks.
Low Pressure · Low Turbulence
Monitor
Quiet underperformance. Early-stage recovery may be forming. Build a watchlist and size in only on confirmed momentum turn.
High Pressure · Low Turbulence
Conviction Position
Strongest regime. Sustained relative strength with low noise. Increase allocation with the trend—this is where most alpha accrues.
◀
Pressure
▶
-- articles
Overview
The AI Supply Chain Weather Map treats the semiconductor-to-software
ecosystem as a weather system. Sustained relative strength creates
high pressure zones. Rapid momentum shifts generate
wind vectors. Extreme turbulence produces
storm cells. Supply chain linkages create
contagion paths through which pressure propagates.
Supply Chain Layers
- Compute/Semi — GPU designers, foundries, memory, semiconductor equipment, EDA tools
- Networking/DC — Switches, optical interconnect, data center REITs, cooling infrastructure
- Cloud/Platform — Hyperscale cloud providers, enterprise cloud, AI infrastructure
- AI Apps/Software — Enterprise AI, observability, automation, data platforms
- Power/Bottleneck — Utilities, power management, grid infrastructure, nuclear/renewables
Strengths
- Maps abstract financial dynamics to intuitive spatial metaphors
- Reveals supply chain contagion and rotation patterns
- Identifies pressure buildups before breakouts
- Daily-updating with 60-day animation history
Limitations
- News sentiment coverage varies by company
- Supply chain weights are heuristic, not estimated
- Geographic layout is approximate (HQ location != economic exposure)
- Storm detection threshold requires calibration
How to Read the Weather Map
- Company dots are placed at real HQ coordinates. Size reflects conviction (distance from neutral). Color indicates supply chain layer.
- Pressure heatmap shows accumulated relative strength. Red = bullish pressure (outperforming). Blue = bearish pressure (underperforming). Intensity = magnitude.
- Storm cells (pulsing rings) appear where turbulence exceeds the threshold. Larger rings = more intense storms. Red = bearish, green = bullish.
- Supply chain arcs connect upstream to downstream layers (e.g., Compute to Cloud). These are the pathways through which pressure propagates.
- Animation plays the last 60 trading days. Watch pressure build, storms form, and momentum rotate through the supply chain.
Feature Engineering
Relative Strength
$$RS_{i,t} = \frac{P_{i,t}}{P_{\text{bench},t}}$$
Pressure
$$\text{Pressure}_{i,t} = 0.4 \cdot (RS\text{-Ratio}_{i,t} - 100) + 0.3 \cdot z(\sigma_{i,t}) + 0.3 \cdot z(r_{i,t}^{21d})$$
Turbulence
$$T_{i,t} = z\!\left(\text{std}_{5d}(\Delta \text{Pressure}_{i})\right)$$
Storm Score
$$\text{Storm}_{i,t} = \max\!\left(0,\; |P_{i,t}| + T_{i,t} - \theta\right)$$
Future: State-Space Model
Latent State Vector
$$\mathbf{x}_{i,t} = \begin{bmatrix} P_{i,t} \\ M_{i,t} \\ N_{i,t} \\ T_{i,t} \\ C_{i,t} \end{bmatrix}$$
Transition with Network
$$\mathbf{x}_{i,t+1} = \mathbf{A}_i \mathbf{x}_{i,t} + \mathbf{B}_i \mathbf{u}_{i,t} + \mathbf{G}_i \sum_j W_{ij} \mathbf{x}_{j,t} + \mathbf{w}_{i,t}$$