Full Signal Analysis — Singapore Macro Intelligence

Executive Summary | Pipeline Architecture | GitHub

Three signals, one decision framework:

Signal 1 — CPI Inflation — how fast your cost floor is rising. Below 2%: negotiate and delay. Above 4%: lock prices immediately.

Signal 2 — USD/SGD and CNY/SGD — what your SGD buys in each market this month. A 10% SGD weakening against USD equals a 10% price increase on all US-sourced goods before your supplier quotes you anything.

Signal 3 — Import Volumes — what the entire market is doing. Rising volumes mean competitors are building stock, lead times will extend. Falling volumes mean suppliers need your order — maximum leverage.

Signal 1 — Inflation: Your Cost Floor

2008 peak +7.2% — Global commodity supercycle. Fixed-price contracts signed pre-2007 saved 12–18% vs spot buyers.

2015–2016 at -1.1% — Oil crash deflation window. Optimal moment to renegotiate supplier contracts. Most teams were too cautious to act.

2022 peak +6.1% — Post-COVID supply shock plus Ukraine commodity surge. Dual-source supply chains maintained continuity. Single-source buyers paid 20–40% spot premiums with 6–12 month lead time extensions.

2026 at +1.8% — Cooling within MAS target. Supplier capacity available. This is the negotiation window.

Counter-intuitive finding: Singapore's inflation correlates more with global commodity prices than with the SGD exchange rate. MAS's active SGD management absorbs much of the FX-inflation transmission. Watch Brent crude and DRAM spot prices — not just MAS policy.

Signal 2 — Currency: The Silent Price Multiplier

The USD/SGD line has held between 1.25 and 1.45 for 25 years — MAS policy, not coincidence. The CNY/SGD line shows the RMB's dramatic strengthening from 2005–2014. Chinese goods have become structurally more expensive in SGD terms over 20 years — though they remain far cheaper in absolute cost per unit than US equivalents.

The USD/CNY Relative Sourcing Power Ratio:

When ratio is above 7.5 — China is relatively cheaper, favour Chinese suppliers. When falling — Chinese cost advantage is compressing, reassess US and ASEAN alternatives. Current ratio approximately 7.0 in 2024, down from 8.0 in 2005. China's cost advantage has compressed 12% in SGD terms over 20 years.

The Medical Equipment Case Study

GE Healthcare (USA) vs Mindray (China) — SGD 50M capital equipment budget. Only variable: FX timing.

At USD/SGD 1.45 (2022 peak): SGD 50M buys USD 34.5M of GE equipment. At USD/SGD 1.25 (2014 low): same budget buys USD 40.0M — USD 5.5M more purchasing power on identical spend. For a hospital group on a 3-year capital cycle: the difference between 6 MRI units upgraded or 7.

Signal 3 — Import Volumes: What the Market Is Doing

2009 GFC — Imports fell 30%. Companies placing 18-month orders in Q2 2009 locked pricing before the 2010–2011 recovery surge.

2016 oil crash — Imports fell 18%. Forward contracts signed in 2016 entered 2017's stable period with protected input costs.

2020 COVID — Imports fell 22%. The counter-cyclical window lasted 4 months. Companies that acted avoided 2021–2022 shortage premiums entirely.

Log-scale view — reveals growth rate, not just level:

Singapore's import growth rate has been slowing since 2012. The era of automatic volume growth is over. Margin management now matters more than volume capture.

Correlation Analysis — Does Weak SGD Cause Inflation?

High inflation periods are scattered across the full range of USD/SGD values — not concentrated in weak-SGD periods. Singapore's inflation is commodity-driven, not currency-driven. Track Brent crude futures, Baltic Dry Index, and USD/SGD simultaneously — when all three move adversely together, procurement costs spike hardest.

Annual Scorecard — 12 Years of Prescribed Actions

Companies following the 2020 DEFLATION signal and locking contracts saved 15–25% on input costs entering 2022. Those ignoring the 2021 ELEVATED signal and continuing spot-buying absorbed the full 2022 HIGH PRESSURE spike.

Pipeline Architecture

This dashboard is a data product — automated, tested, versioned, trusted.

15 automated dbt tests validate every source record before it reaches this page. Idempotent ingestion scripts prevent duplicate data. Every raw record carries an ingestion timestamp — full audit trail from government API to dashboard. Schema changes in source APIs fail loudly rather than silently serving wrong numbers.

Extract: SingStat API (CPI) — Python + exponential backoff — Bronze layer

Extract: MAS via data.gov.sg (FX) — Python + pagination — Bronze layer

Extract: Enterprise SG (Trade) — Python + idempotent load — Bronze layer

Transform: Bronze to Silver to Gold — dbt Core, 5 models, 15 tests

Serve: Gold layer — DuckDB — Evidence.dev dashboard

The same patterns are used by data teams at DBS Bank, GovTech Singapore, and commodity trading firms running production analytical systems.

Data sources: SingStat CPI (monthly since 1961) · MAS via data.gov.sg (15 currencies since 1988) · Enterprise Singapore via data.gov.sg (merchandise trade since 1976)

Pipeline last executed: April 2026 · Tests: 15 of 15 passing · Models: 5

Data: SingStat (CPI 1961–2026) · MAS via data.gov.sg (15 currencies 1988–2024) · Enterprise Singapore (trade 1976–2026)
Pipeline: Python 3.13 · DuckDB 1.2 · dbt Core 1.11 (5 models, 15 tests passing) · Evidence.dev · Netlify · GitHub Actions
AI development: Claude Sonnet 4.5 (Anthropic) · Claude.ai Projects · Chart decisions, SQL window functions, correlation framing
Last executed: April 2026 · View source on GitHub · Star schema