6.10 Sector cycles: capex, credit and PSU banks

6.10.1Valuing cyclicals#

Secular versus cyclical Secular compounders (consumer brands, private banks with cheap deposits, IT services historically, parts of specialty chemicals) are judged on runway × ROIC × reinvestment. Cyclicals (metals and mining, cement, autos, capital goods, PSU banks, real estate, sugar, shipping) earn returns that average near or below WACC over a cycle with large swings. For cyclicals:

• Value on normalised (mid-cycle) earnings, not current earnings.
• P/E is misleading: high at the trough (depressed E), low at the peak. Lynch's rule is to buy at high P/E or peak losses and sell at low P/E or peak earnings.
• Use P/B and replacement cost, and track the spread that drives profit: steel price minus iron-ore and coking-coal cost; cement realisation per tonne; net interest margin and credit cost for banks.

6.10.2The Indian credit cycle, 2003–2026#

India's capex and credit cycle, as a teaching narrative The dates below are well established; individual figures should be checked against RBI and Economic Survey sources before you rely on them.

• 2003–08: infrastructure and capex boom; heavy corporate borrowing; public-sector (PSU) banks lend aggressively to power, steel, roads.
• 2011–18: projects stall; over-levered corporates and stressed banks produce the "twin balance sheet" problem (Economic Survey 2016-17, ch. 4). The RBI's Asset Quality Review (2015) forces banks to recognise bad loans; the Insolvency and Bankruptcy Code is enacted in 2016. PSU-bank gross NPAs peak around March 2018.
• 2018–19: NBFC shock. IL&FS, a AAA-rated infrastructure lender, defaults from September 2018 across group entities; group debt is reported at about ₹91,000 cr. Its failure was an asset-liability mismatch: long-gestation infrastructure assets funded with short-term commercial paper and inter-corporate deposits. When short-term funding froze, DHFL and others followed.
• 2020: Covid; RBI loan moratorium and restructuring.
• 2021–24: deleveraged corporates, cleaner bank balance sheets, government-led capex; PSU banks and PSU companies re-rate.
• March 2026: secondary summaries of the RBI Financial Stability Report (June 2026) report scheduled commercial banks' GNPA ratio at about 1.8%, a multi-decade low.

6.10.3Bank multiples through the cycle#

Mechanism in this module's terms A credit cycle runs: easy credit → over-investment at RONIC < WACC → NPAs → recognition → clean-up → re-rating. Through (6.18), a bank's P/B is $(\text{ROE} - g)/(k - g)$. At the bottom, expected ROE is below $k$ (credit costs eat profits) and P/B is well below 1. As credit costs fall, expected ROE rises above $k$ and P/B re-rates above 1. The return to an investor who bought at the bottom comes from both earnings recovery and multiple expansion. The risk is that "cheap on book" is a value trap if the book is overstated (Yes Bank, 6.12.3): P/B is only meaningful if B is real.