SciRisk 100 · Semiconductors · Rank 13 of 100

TSMC

TSM · Taiwan · Semiconductors SC4 High

Economic Supply Chain Risk Capital

$197.0m

95% one-year Value-at-Risk · interval $193.5m to $199.5m (3.0% wide)

Risk Capital Intensity

0.219%

of $90.1bn revenue · rank #11 of 100

Global rank

#13

of 100 by ESCRC

Sector rank

#4

of 10 in Semiconductors

Expected Shortfall

$261.3m

Mean loss beyond ESCRC

Expected annual loss

$47.5m

Ordinary year

Revenue / COGS

90.1 / 39.5

USD bn · gross margin 56%

Suppliers modelled

11

3 correlated geographies

What does this ESCRC number mean?

An ESCRC of $197.0m is the estimated economic supply-chain risk capital requirement generated by the SciRisk ESCRC model for TSMC under the stated assumptions: a 365-day horizon, 100,000 simulated years, and the 95% tail of the resulting loss distribution. Beyond that threshold the average loss — the Expected Shortfall — is $261.3m; in an ordinary year the expected loss is $47.5m.

Value-at-Risk: the portfolio loss not exceeded with the chosen confidence level. Read as the ESCRC risk capital — the loss the firm should be prepared to absorb in a bad year. Model-implied and intended for screening and ranking, not a calibrated capital figure.

Supply Chain Risk Capital Profile

Where the risk capital comes from. Every column is a production-engine output for this network — nothing is inferred.

By tier

Tier 1

97%

Tier 2

2%

Tier 3

1%

Share of summed standalone VaR. Deeper tiers are attenuated by the engine's echelon buffer factors and capped by flow conservation.

By supplier geography

Japan

47%

USA

33%

Netherlands

17%

Germany

2%

Major-event risk is correlated within a geography group (rho = 0.35), so concentration here compounds.

By supplier industry

semiconductor

99%

chemicals materials

1%

equipment precision

0%

Buckets are the calibrated-priors industry classes that set each supplier's minor-event rate.

Largest supplier share

23%

Tokyo Electron

Top-3 concentration

60%

of summed standalone VaR

Tier-1 share of risk

97%

Direct suppliers

Diversification benefit

65.5%

vs. sum of standalone VaR

Supplier	Tier	Country	ExposureUSD m	Standalone VaRUSD m	Expected lossUSD m	Tail intensityrevenue-days	RoESCRCx
Tokyo ElectronCoaters/developers	T1	Japan	5000	132.5	12.9	4.2	48.2
Shin-Etsu ChemicalSilicon wafers 300mm	T1	Japan	4000	106.0	10.2	4.2	48.2
Applied MaterialsCVD/PVD equipment	T1	USA	8000	105.3	7.4	2.1	97.2
ASMLEUV/DUV scanners	T1	Netherlands	7200	99.2	6.4	2.2	92.8
Lam ResearchEtch systems	T1	USA	6000	79.0	5.6	2.1	97.2
JSR CorpPhotoresist	T1	Japan	1200	31.8	3.1	4.2	48.2
Merck KGaAEUV resist/chemicals	T2	Germany	800	6.6	0.4	2.2	92.8
EntegrisChemical filtration	T2	USA	1000	5.0	0.4	1.3	154.7
ZeissEUV optics	T3	Germany	1080	4.5	0.3	2.2	92.8
Wacker ChemiePolysilicon	T3	Germany	720	1.8	0.1	1.3	154.7
KLA CorporationProcess control/inspection metrology instruments	T1	USA	1500	0.0	0.7	—	—

The calculator ships a researched network for TSMC

This index entry uses calibrated bucket defaults for every supplier's disruption rate, recovery time and revenue-at-risk fraction — the uniform rule applied to all 100 companies — and prices severity pro-rata, by each supplier's share of cost. The ESCRC Calculator ships a researched network for this company ("TSMC"): supplier-specific parameters, priced instead by each supplier's declared criticality — the share of output that halts when it stops. Same engine, same seed, same settings, but two layers differ, the parameters and the severity formula:

SciRisk 100 entry (bucket defaults, pro-rata): $197.0m · 0.219% of revenue
Calculator preset (researched, criticality): $3.72bn · 4.129% of revenue
Ratio: 18.89x

This difference is not a measure of what supplier-level research is worth: it compounds two separate changes, and the formula change is generally the larger of the two. Pro-rata pricing bounds a supplier's loss by its share of cost, so it understates a concentrated, low-spend bottleneck by construction; criticality pricing does not, and rests on declared criticality and unrecoverability values that are operator judgment rather than model outputs. The two figures answer different questions, and neither is used to adjust the other. It is why index figures are labelled preliminary.

Open the ESCRC Calculator

Scenario Analysis — coming in Phase 2

Recovery-time, supplier-diversification, inventory and geographic-diversification scenarios for TSMC are not part of this release. The ESCRC Calculator already supports what-if analysis on a network you build yourself.

Sector peers

Nearest-ranked companies in Semiconductors. TSMC is highlighted.

Open in Compare

Rank	Company	Ticker	Country	Sector	RevenueUSD bn	ESCRCUSD m	ESCRC / Revenue%	Risk Grade	Change
3	NVIDIA	NVDA	USA	Semiconductors	130.0	577.3	0.444	SC5 Very High	0.0%
9	Samsung Electronics	005930.KS	South Korea	Semiconductors	300.0	269.5	0.090	SC3 Moderate	0.0%
10	Broadcom	AVGO	USA	Semiconductors	51.6	242.0	0.469	SC5 Very High	0.0%
13	TSMC	TSM	Taiwan	Semiconductors	90.1	197.0	0.219	SC4 High	0.0%
16	Qualcomm	QCOM	USA	Semiconductors	39.0	158.6	0.407	SC5 Very High	0.0%
47	Intel	INTC	USA	Semiconductors	54.0	57.6	0.107	SC4 High	0.0%

See this for your own supply chain

You are looking at TSMC's supply-chain risk capital: $197.0m against $90.1bn of revenue, driven mostly by Tokyo Electron and Japan. The calculator runs the same engine on a network you build yourself.

Calculate your ESCRC Load TSMC's network Compare with peers

ESCRC calculated using SciRisk ESCRC Model v5.4.2 · calibrated priors 2026.09-5
Company data as of 2026-06-30 · ESCRC calculated 2026-09-21 · dataset 2026.09-23 Beta

Simulation: 100,000 Monte Carlo paths · 95% confidence · 365-day horizon
Seed: 20260821 · regional correlation ρ = 0.35
Input basis: Curated network, calibrated-prior parameters
Input fingerprint: 88b0165959e0

Input provenance for TSMC

Financials: Production calculator preset focal-firm figures.
Network: Production calculator preset network (TSMC) — supplier names, tiers, geographies, products and purchase volumes as shipped.
Parameters: lambda1/lambda2 = calibrated ESCRC priors 2026.09-5 bucket defaults; r1/r2 = median of the shipped production preset networks within the same industry x geography cell when that cell has enough independent data, else the same industry-only bucket median, or (for industries with no preset supplier coverage) a directly-cited external recovery-event duration where one was found; ic = the same preset-network median (no external benchmark exists for this parameter for any industry). All three clamped into externally-cited recovery-time/BI-loss bounds per field.

SciRisk ESCRC is an analytical estimate generated using the SciRisk Economic Supply Chain Risk Capital methodology. It is not a credit rating, investment recommendation or statement of financial condition. Figures are model-implied and intended for screening and comparison, not as calibrated capital requirements.

Beta / preliminary dataset. Company inputs are analyst-curated from public annual reports and public supply-chain disclosure. ESCRC itself is computed by the production ESCRC engine — only the INPUTS on this page are preliminary.