Reading economic data: time series, growth rates and high-frequency indicators

Detail

1. What a time series is

• A time series is data arranged in time order. Examples: India's population in each census year, or exports in each year.
• A cross-section is different. It shows many units (states, households) at one point in time.
• An economist splits a time series into four parts to read it correctly:

Part	Time span	What it is	Indian example
Trend	Many years	Long-run direction	Rising exports and imports, 1993-2014
Seasonality	Within one year	Pattern that repeats every year	Kharif/rabi crop arrivals, festive demand, monsoon vegetable prices
Cyclicity	More than one year	Business cycle: boom → slowdown → recession → recovery	Investment cycles (phases → growth-theories-business-cycles)
Irregular	One-off	Sudden shock that does not repeat	COVID lockdown, 2020-21

2. Trend

• The trend is the long-run movement once short-term ups and downs are ignored.
• It is easiest to see on a line graph (time on the x-axis, value on the y-axis).
• NCERT example: both exports and imports rose from 1993 to 2014. Imports stayed above exports, so the gap between the two lines shows the trade deficit.

3. Seasonality and why data is quoted year-on-year

• Seasonality is a regular pattern inside a year. It comes from weather, crop cycles and festivals.
• Vegetable prices (tomato, onion) usually rise in the monsoon months because rain damages supply.
• Consumer demand rises around Diwali (October-November).

• Mandi arrivals jump after the kharif and rabi harvests.

• Why this matters:

• Compare November with October and you will see a big jump in sales. That jump is mostly the festival, not real growth.

• So CPI (Consumer Price Index) and IIP (Index of Industrial Production) are quoted year-on-year, i.e. this month against the same month last year. Both months have the same season, so the seasonal effect cancels out.

• Seasonally adjusted (SA) data:

• Statisticians estimate the usual seasonal pattern and remove it. What is left is the underlying movement.
• Only then can two back-to-back months (m-o-m) be compared fairly.
• Example: RBI's OBICUS reports seasonally adjusted capacity utilisation. It was 75.3% in Q3:2024-25, above its long-term average [4].

4. Cyclicity and irregular movements

• Cyclicity means waves longer than one year. These are business cycles: expansion, peak, contraction, trough.
• Seasonality has a fixed length (one year). Cycles do not. A cycle can last 3 years or 8 years.
• Irregular movements are random, one-time shocks, such as COVID in 2020-21, demonetisation or a war-driven oil price spike. They cannot be forecast from past data.

5. Growth arithmetic

(a) y-o-y vs m-o-m

• Year-on-year (y-o-y) growth = (Value this month − Value same month last year) ÷ Value same month last year × 100.
• Month-on-month (m-o-m) growth = (Value this month − Value last month) ÷ Value last month × 100.
• m-o-m is noisy and seasonal. y-o-y compares like with like.
• The trade-off: y-o-y is slow to show a turning point, because it carries 12 months of history. SA m-o-m shows a turn sooner.

(b) Base effect

• Base effect means the growth rate depends on how high or low last year's figure (the base) was.
• Low base → high y-o-y growth, even when there is little real improvement.

• High base → low y-o-y growth, even when activity is healthy.

• Worked example (index values):

• April 2019 = 100. April 2020 (lockdown) = 60. April 2021 = 90.
• y-o-y growth in April 2021 = (90 − 60)/60 = +50%.
• Yet output is still 10% below April 2019.

• This is why growth after the 2020-21 slump looked high. A large part of it was the base effect.

• Inflation works the same way. If prices jumped last year, this year's CPI inflation can look low even while prices keep rising (a "favourable base").

(c) Percentage vs percentage points

• A percentage point (pp) is the simple difference between two percentages.
• A per cent change is that difference divided by the starting value.
• Example: the unemployment rate falls from 6.1% to 3.2%:
• Fall = 6.1 − 3.2 = 2.9 percentage points.

• Relative fall = 2.9 ÷ 6.1 = ~48%.

• Policy rates are discussed the same way. A repo rate cut from 6.50% to 6.25% is a cut of 0.25 pp (25 basis points), not "0.25%". (1 basis point = 0.01 pp.)

(d) Compound annual growth rate (CAGR)

• CAGR is the one steady yearly rate that would take a value from its start to its end over n years.
• Formula: CAGR = (End value / Start value)^(1/n) − 1
• CAGR is a geometric mean of the yearly growth factors, not an arithmetic (simple) average.
• Why not a simple average? Growth compounds.
• If something grows +50% one year and then falls −50% the next, the simple average is 0%.

• But 100 → 150 → 75 is a real loss. Its CAGR = (75/100)^(1/2) − 1 ≈ −13.4%.

• NCERT census example:

• Decadal population growth in 2001-11 = 17.7%.
• Annual rate = 1.177^(1/10) − 1 ≈ 1.64% a year, which matches NCERT.
• Dividing 17.7 by 10 = 1.77% overstates it, because it ignores compounding.

(e) Nominal vs real

• Nominal figures are at current prices. Real figures remove inflation (they are at constant prices of a base year).
• Rough rule: real growth ≈ nominal growth − inflation. For example, nominal GDP growth 10% with inflation 4% gives real growth of about 6%.
• Details → national-income-accounting / inflation-price-indices.

6. Purchasing Managers' Index (PMI)

• PMI is a monthly, survey-based business indicator. Purchasing managers are asked whether key business variables went up, down or stayed the same compared with the previous month.
• In India: the HSBC India Manufacturing, Services and Composite PMIs. They are compiled by S&P Global from panels of about 400 firms each.

• Composite PMI = manufacturing and services combined.

• Diffusion index: it measures how widely a change is spread across firms, not how big the change is.

• Each component = (% of firms reporting "higher") + 0.5 × (% reporting "same").

• Example: 40% say higher, 45% same, 15% lower → 40 + 22.5 = 62.5 → expansion.

• Reading it: >50 = expansion, <50 = contraction, 50 = no change.

• Trap: PMI shows the direction of change, not the level of activity. A fall from 58 to 54 still means growth, only at a slower pace.
• Manufacturing PMI weights:

Component	Weight
New orders	30%
Output	25%
Employment	20%
Suppliers' delivery times (inverted)	15%
Stocks of purchases	10%

• Why suppliers' delivery times are inverted: slower deliveries usually mean suppliers are busy because demand is strong. So a longer delivery time pushes the PMI up.
• Timing: a flash (early) estimate comes out mid/late month. The final print comes early the next month. This makes PMI one of the earliest data points on any month, weeks before official data such as IIP.
• Limitation: it is a private survey of a small panel, mostly formal firms. It is based on opinion, not measured output, so it can differ from IIP.

7. Indicators by timing

Leading indicators (turn before the economy)

• Used to spot turning points early.
• PMI new orders: orders today become production tomorrow.
• Stock prices: investors price in expected future profits.
• Slope of the yield curve: the gap between long-term and short-term government bond yields.
• Normal (upward) slope → markets expect growth.

• Flat or inverted (short rates above long rates) → markets expect a slowdown.

• Consumer confidence: households that feel confident plan to spend more.

Coincident indicators (move with the economy)

• IIP (Index of Industrial Production):
• MoSPI released a new IIP series with base year 2022-23 on 1 June 2026 [8].
• Weights now come from National Accounts (base 2022-23) and the Annual Survey of Industries [8].
• It reports four sectors: mining & quarrying; manufacturing; electricity & gas; and water supply, sewerage & waste management [8].

• IIP grew 5.1% in May 2026 [9].

• Index of Core Industries (ICI):

• The new series with base year 2022-23 has nine core industries, not eight (NCERT: "eight core industries"). Iron ore has been added [7].
• Steel is now measured by gross production, to match IIP [7].
• Coal now covers only raw coal, to avoid double counting [7].
• ICI weights are taken from the matching items in IIP (2022-23) [7].

• The provisional ICI is released on the 20th of the following month [7].

• GST collections and e-way bills: e-way bills track the movement of goods.

• Power demand: electricity use rises and falls with activity.

Lagging indicators (turn after the economy)

• Unemployment: firms hire only after they are sure demand has recovered.
• CPI inflation: prices react with a delay to demand pressure.
• Bank NPAs (non-performing assets, i.e. loans not being repaid): bad loans show up months or years after a downturn.
• Use: they confirm a turn that has already happened. They do not predict it.

8. RBI's forward-looking surveys (read by the MPC)

• The Monetary Policy Committee (MPC) sets the repo rate (the rate at which RBI lends to banks for a short time). It targets future inflation, so it needs to know what firms and households expect.
• RBI runs five main enterprise and household surveys: IESH, CCS, IOS, SIOS and OBICUS [2]. It also runs the SPF.

Survey	Who is asked	What it tells RBI	Key facts
OBICUS (Order Books, Inventories and Capacity Utilisation Survey)	Manufacturing firms	New and pending orders, inventories, capacity utilisation (how much of the available machinery is actually used)	Quarterly since 2008 [3]; CU 75.3% (SA), Q3:2024-25 [4]
Industrial Outlook Survey (IOS)	Manufacturing firms	Their view of production, orders, prices and employment, now and in the next quarter	Business expectations index
Services & Infrastructure Outlook Survey (SIOS)	Services and infrastructure firms	The same kind of outlook for the non-manufacturing sector	One of RBI's regular surveys [2]
Consumer Confidence Survey (CCS)	Urban households	Views on the economy, jobs, prices, income and spending	Bi-monthly, 19 major cities [5]; Current Situation Index (CSI) and Future Expectations Index (FEI); 100 = neutral; CSI 94.0 (Nov 2024), 6,078 respondents [5]
Inflation Expectations Survey of Households (IESH)	Urban households	Expected inflation 3 months and 1 year ahead	Started Sept 2005 (quarterly); now bi-monthly, 18 cities, ~6,000 households [2]
Survey of Professional Forecasters (SPF)	Professional economists and analysts	Forecasts of GDP growth, inflation and other variables	Bi-monthly [6]

• Why expectations matter:
• Households expect high inflation → they demand higher wages and buy goods early → actual inflation rises.

• So the MPC watches IESH closely, even though households usually expect higher inflation than the official CPI shows.

• Why capacity utilisation matters:

• High CU (above about 75%) → firms are running out of spare capacity → they are likely to invest in new plants.
• High CU also warns that demand may start pushing prices up.

Prelims Hooks

• CAGR = (End/Start)^(1/n) − 1. It is a geometric mean. Decadal growth of 17.7% (2001-11) ≈ 1.64% a year.
• Unemployment rate from 6.1% to 3.2% = a fall of 2.9 percentage points, or ~48%. Watch for "percentage" vs "percentage point" traps.
• PMI > 50 = expansion; < 50 = contraction. It is a diffusion index of m-o-m direction, not level. Compiled by S&P Global (HSBC India PMI) from panels of ~400 firms.
• Largest weight in manufacturing PMI: New orders (30%). Suppliers' delivery times (15%) are inverted.
• Leading: PMI new orders, stock prices, yield-curve slope, consumer confidence. Coincident: IIP, core industries, GST/e-way bills, power demand. Lagging: unemployment, CPI inflation, NPAs.
• New Index of Core Industries (base 2022-23) has nine industries. Iron ore was added [7].
• New IIP series has base year 2022-23, released 1 June 2026 by MoSPI [8].
• OBICUS (quarterly since 2008) measures capacity utilisation in manufacturing [3]. IESH covers 18 cities; CCS covers 19 cities [2][5].
• Base effect: a low base inflates y-o-y growth, and a high base deflates it.
• CPI and IIP are quoted y-o-y to remove seasonality.

Mains Points

• Reading growth data correctly:
• Headline y-o-y numbers after 2020-21 were inflated by base effects.
• A better test compares with the pre-COVID level (2019-20) or uses a multi-year CAGR.

• This matters for judging whether "V-shaped recovery" claims are true (GS-III: growth and development).

• Old base years mislead:

• Base revisions (IIP and ICI to 2022-23) bring in new sectors (iron ore; water and waste management) and updated weights [7][8].
• Without them, indices over-weight old industries and miss new ones.

• The cost of revision: comparisons across series need linking factors, and people may question the credibility of the numbers.

• High-frequency and survey data help policy work in real time:

• GDP comes with a lag of about two months. So the MPC relies on PMI, GST/e-way bills, power demand and RBI surveys (IESH, OBICUS, CCS) to judge demand and inflation expectations early.

• The limits: these sources are mostly urban and formal-sector, so they under-represent the rural and informal economy.

• Anchoring expectations:

• IESH shows households expect higher inflation than official CPI. This shows why the credibility of inflation targeting matters.
• If expectations stay unanchored, a price shock turns into a wage-price spiral.