India Macro · Quarterly Nowcast

India GDP NowCast.

Target  FY27 Q1 · Apr–Jun 2026
Official print  ~29 Aug 2026
Model  Combination of bridge equations

REAL GDP · year-on-year

6.75%

Holding near trend; services-led

prev quarter 7.83%

R² 0.17 · dir 88%

actual nowcast (median) central band (±MAE, ~57%)

Central band5.2 – 8.3%

Drop-a-group range5.6 – 8.3%

Hard data in hand83%

Central band = ±MAE (~57% range). Drop-a-group = how far the nowcast moves if any one indicator is removed. Hard data = share of the quarter’s three months already published (vs extrapolated).

RBI projection · real GDP · RBI · Jun 2026 MPC1Q27 6.62Q27 6.33Q27 6.54Q27 6.8FY27 6.6%

NOMINAL GDP · year-on-year

8.45%

Credit- and payments-led; GST base corrected

GST corrected for a stale-base error that showed a false drag; now +5.3%, in line with credit & UPI.

prev quarter 9.12%

R² 0.39 · dir 92%

actual nowcast (median) central band (±MAE, ~57%)

Central band6.3 – 10.6%

Drop-a-group range7.8 – 9.7%

Hard data in hand78%

What the nowcast says

Real GDP nowcasts at 6.8%, down from 7.83% last quarter, sitting 0.15pp above the RBI MPC's 6.6% projection. Services PMI (58.63) props growth up — dropping it pulls the nowcast to just 5.62% — while dropping ICI (0.86) lifts it to 8.26%, showing industry is the drag. With R2 of just 0.17 and industrial gauges structurally blind to services, the 5.18-8.31% band, not the point estimate, is the honest read.

Nominal GDP nowcasts at 8.5%, down from 9.12% prior, with the value indicators broadly aligned: dropping GST (5.26%, now rebuilt on the correctly revised base) lifts the nowcast to 9.69%, while dropping Bank credit (15.83%) drags it to 7.79% — the tension is credit-led strength versus a GST print still carrying a modest, fading drag from September's rate cut. UPI value (21.4%) is closest to consensus, dropping it barely moves the nowcast to 8.4%.

The dominant caveat is band width, not data quality: real GDP's 5.18-8.31% central band spans over 3pp and the model explains only 17% of quarterly variance (R2 0.17), so directional accuracy (88%) is far more trustworthy than the level. Nominal's band (6.27-10.63%) is similarly wide despite stronger fit (R2 0.39), and its GST input should now be read as clean, not as a residual distortion.

How much to trust it · out-of-sample backtest

Metric	Real GDP	Nominal GDP
R² — variance explained vs naive	0.17	0.39
Correlation, nowcast vs actual	0.76	0.57
Directional hit-rate	88%	92%
Typical miss (RMSE)	±2.1 pp	±3.3 pp

Walk-forward backtest, expanding window, FY19 Q1 onward, COVID quarters excluded (n = 24). Real GDP tracks direction well but adds little on level — it is inherently smooth; nominal carries more genuine skill.

FY27 Q1 indicators

Indicator	Latest	Unit	Hard months
ICI — Eight Core Industries	0.86	% y/y	2 / 3
Services PMI	58.63	index	3 / 3
Manufacturing PMI	54.73	index	3 / 3
Air passengers (domestic)	5.07	% y/y	2 / 3
GST collections	5.26	% y/y	2 / 3
Bank credit	15.83	% y/y	2 / 3
UPI payment value	21.4	% y/y	3 / 3
PV retail (4W)monitored	23.1	% y/y	3 / 3
2W retailmonitored	15.08	% y/y	3 / 3

Values feeding the current-quarter estimate. “Hard months” = actual data in hand of the three in the quarter; the rest is extrapolated. Rows tagged monitored are tracked for context and do not feed the model — on this sample they carry no predictive signal for real GDP.

Model-free z-scores · FY26 Q1 → FY27 Q1

Indicator	FY26 Q1	FY26 Q2	FY26 Q3	FY26 Q4	FY27 Q1
ICI — Eight Core Industries	-0.44	+0.10	-0.30	-0.19	-0.55
Services PMI	+0.98	+1.32	+0.92	+0.78	+0.87
Manufacturing PMI	+1.32	+1.50	+0.99	+0.55	+0.36
Air passengers (domestic)	-0.27	-0.36	-0.30	-0.32	-0.26
GST collections	-0.29	-0.25	-0.64	-0.55	-0.39
Bank credit	-0.38	-0.33	+0.16	+0.57	+0.88
UPI payment value	-0.60	-0.62	-0.62	-0.50	-0.61
PV retail (4W)monitored	-0.30	-0.38	-0.11	-0.11	-0.02
2W retailmonitored	-0.22	-0.33	-0.05	+0.04	-0.11

above long-run average below long-run average Standard deviations from each indicator’s own history — independent of the model. monitored = tracked for context, not a model input.

Terms

Drop-a-group range: The span of nowcasts obtained by removing each indicator in turn. A tight range means no single series is driving the result; a wide one means the forecast leans on specific indicators.
Hard data in hand: Share of the current quarter’s monthly indicator readings that are already published, versus model-extrapolated. Higher means less of the nowcast rests on extrapolation.
Bridge equation: A simple regression linking quarterly GDP to monthly indicators aggregated up to the quarter, plus a lag of GDP itself. This model estimates every one- and two-indicator bridge and averages their forecasts — robust when the sample is short and relationships shift.
MAE — Mean Absolute Error: The average size of past nowcast misses, ignoring direction. The ±MAE band shown on each chart is the ~57% confidence range.
R² (out-of-sample): Share of the variation in actual GDP the nowcast explains in backtest, measured against a naive running-average forecast. Higher means more genuine skill.
RMSE — Root Mean Square Error: Typical miss, penalising large errors more heavily than small ones.
Directional hit-rate: How often the nowcast gets the direction right — whether GDP rose or fell versus the prior quarter.
YoY — Year-on-year: Change versus the same quarter or month one year earlier.
PMI — Purchasing Managers’ Index: Monthly business survey; 50 = no change, above 50 = expansion, below = contraction.
ICI — Index of Eight Core Industries: Combined output of coal, crude oil, natural gas, refinery products, fertilisers, steel, cement and electricity — about 40% of industrial production.
GST — Goods & Services Tax: Monthly indirect-tax collections; a nominal, services-inclusive proxy for activity.
UPI — Unified Payments Interface: Value of real-time digital payments each month; a fast proxy for consumption.