Skip to main content

India CPI-IW Inflation Data

Purpose​

INDIA_CPI_IW_YOY is a macroeconomic time series representing the point-to-point year-on-year inflation rate based on India's Consumer Price Index for Industrial Workers (CPI-IW).

It is intended for:

  • Asset-allocation research
  • Inflation-regime analysis
  • Macro regime filters
  • Real-return calculations
  • Portfolio drawdown analysis
  • Historical backtesting

Important: CPI-IW is not the same as India's modern headline CPI Combined. CPI-IW covers the Industrial Worker population and should therefore be stored and identified separately.


Official Source​

The primary source is the Labour Bureau, Government of India.

The Labour Bureau publishes the point-to-point rate of inflation in CPI-IW on a monthly basis.


What Does CPI-IW Inflation Mean?​

The published inflation rate is a year-on-year comparison:

CPI-IW Inflation (YoY)
=
(CPI-IW current month / CPI-IW same month previous year - 1)
× 100

Example:

CPI-IW January 2025 = 150.0
CPI-IW January 2024 = 145.0

Inflation
= (150 / 145 - 1) × 100
= 3.45%

For the historical Labour Bureau dataset, use the published inflation rate rather than recalculating it unless there is a specific data-validation requirement.


Frequency​

The series is:

Frequency: Monthly
Observation: Year-on-year percentage

There are normally 12 inflation observations per calendar year.

Recommended database date:

YYYY-MM-last-day-of-month

Examples:

1996-01-31
1996-02-29
1996-03-31
...
1996-12-31

TradeEntry Asset Definition​

Store CPI-IW as a macro asset rather than treating it as a tradable security.

Recommended te_asset (corrected 2026-10-02 to match the real schema - asset_class/value_kind are CHECK-constrained in backend/AA/models.py to equity|precious_metal|fixed_income|cash|economic_data and level|rate_pct respectively; MACRO/RATE below were never valid values and this table previously stated them in error - see backend/AA/seed_cpi_iw_inflation.py, the actual implementation, for the values really used):

FieldValue
codeINDIA_CPI_IW_YOY
nameIndia CPI-IW Inflation YoY
asset_classeconomic_data
sub_typecpi_iw
value_kindrate_pct
unit%
source_publisherLabour Bureau, Government of India

The corresponding observation is stored in te_price.close.

Example:

te_price
------------------------------------------------
asset_id | price_date | close
------------------------------------------------
10001 | 1996-01-31 | 9.00
10001 | 1996-02-29 | 8.60
10001 | 1996-03-31 | 8.90

For inflation, open, high, and low should remain NULL. The close field represents the published inflation rate.


Historical CPI-IW Base Years​

CPI-IW has undergone base-year revisions.

For historical data, the relevant base series are:

PeriodCPI-IW Base
Before January 20061982 = 100
January 2006 – August 20202001 = 100
September 2020 onward2016 = 100

The base-year change does not mean the published YoY inflation series should be treated as a new economic variable. The Labour Bureau provides the official inflation series and linking methodology.

For auditability, retain the base_year associated with each observation.

Example:

date inflation_rate base_year
1996-01-31 9.00 1982=100
2005-12-31 ... 1982=100
2006-01-31 ... 2001=100
2020-08-31 ... 2001=100
2020-09-30 ... 2016=100
2026-07-31 ... 2016=100

te_asset​

code = 'INDIA_CPI_IW_YOY'
name = 'India CPI-IW Inflation YoY'
asset_class = 'economic_data'
sub_type = 'cpi_iw'
value_kind = 'rate_pct'
unit = '%'

te_price​

asset_id
price_date
close

Example:

SELECT
price_date,
close AS inflation_rate
FROM te_price
WHERE asset_id = :cpi_iw_asset_id
ORDER BY price_date;

CSV Import Format​

The normalized CSV should use:

date
inflation_rate
frequency
series
source
base_year
is_official

Example:

date,inflation_rate,frequency,series,source,base_year,is_official
1996-01-31,9.00,monthly,INDIA_CPI_IW_YOY,Labour Bureau,1982=100,true
1996-02-29,8.60,monthly,INDIA_CPI_IW_YOY,Labour Bureau,1982=100,true
1996-03-31,8.90,monthly,INDIA_CPI_IW_YOY,Labour Bureau,1982=100,true

CPI-IW vs CPI Combined​

India has multiple CPI series.

CPI-IW​

Population: Industrial Workers
Publisher: Labour Bureau
Frequency: Monthly
Historical coverage: Long
Primary use here: Long-term macro regime analysis

CPI Combined​

Population: Rural + Urban
Publisher: MoSPI / NSO
Frequency: Monthly
Modern headline inflation measure
Primary use: Current headline inflation analysis

They should be stored as different assets:

INDIA_CPI_IW_YOY
INDIA_CPI_COMBINED_YOY

Do not silently splice them into a single series.


For the asset-allocation project, maintain separate macro assets:

MACRO
│
├── INDIA_CPI_IW_YOY
├── INDIA_CPI_COMBINED_YOY
├── INDIA_WPI_YOY
├── INDIA_10Y_GSEC_YIELD
├── INDIA_VIX
├── RBI_POLICY_RATE
└── USDINR

Then combine these with market assets:

ASSETS
│
├── NIFTY 50
├── GOLD
├── SILVER
├── T-BILL
├── G-SEC
├── CORPORATE BOND
└── CASH

This creates a clean foundation for a macro-regime asset-allocation model.


Data Integrity Rules​

  1. Never overwrite the original downloaded source.
  2. Store the official source URL.
  3. Store the download/retrieval date.
  4. Store the CPI base year.
  5. Keep the published inflation rate rather than reconstructing it unnecessarily.
  6. Do not mix CPI-IW and CPI Combined.
  7. Do not fill missing months by interpolation without explicitly marking them.
  8. Validate that there is normally one observation per month.
  9. Enforce a unique constraint on (asset_id, price_date).
  10. Preserve revisions when the official source revises historical observations.

Recommended PostgreSQL constraint:

CREATE UNIQUE INDEX IF NOT EXISTS
ux_te_price_asset_date
ON te_price(asset_id, price_date);

Important Limitation​

CPI-IW is a population-specific CPI, not India's current headline CPI Combined.

Therefore:

1996 → present
CPI-IW

is appropriate for a long historical macro regime variable, but for current headline-inflation analysis also maintain:

2012 → present
CPI Combined

from MoSPI.


Official References​