12.2 · Market cycles & sentiment¶
Why this matters: the same company, with the same cash flows, has traded at 15x and 45x within five years because the market's mood changed. A fundamental investor cannot time the mood, but can measure it — index valuation bands, flows, IPO froth, small-cap premia — and adjust behaviour: more cash and tighter margins of safety when the pendulum is at one extreme, more aggression at the other. India's small-cap cycles (2008, 2018, 2024–25) are the reference, and September 2026 — Nifty below its median P/E, record domestic flows, FPIs at a 15-year-low ownership — is the live reading.
Learning objectives — after this lesson you can:
- Read index valuation measures (P/E and P/B bands, earnings yield vs bond yield, market cap/GDP) and state their limits.
- Track flows (FPI vs DII, SIPs) and primary-market activity (IPOs, QIPs) as sentiment gauges.
- Describe India's small-cap cycles and the behaviour that survived them.
- Apply Howard Marks's pendulum to behaviour: what to do at extremes, and what not to do in the middle.
Prerequisites: 12.1 Macro for equity investors, 06.5 Multiples · Time: ~60 min
1. Measuring the market's mood¶
| Measure | How | Reading (Sep-2026) | Limits |
|---|---|---|---|
| Index trailing P/E vs its history | Nifty 50 P/E vs 10–20-year bands (median, ±1σ) | ~19.5x vs a 5-year median of ~22x; historical band 2 of 5, from which 5-year forward returns have historically been positive with a median in the low teens (IndexPE; PrimeInvestor) | Earnings at a cycle peak make P/E look low (FY22, FY25); index composition changes (consolidated EPS from 2021 lowered the P/E series — compare like with like) |
| Index P/B | Nifty P/B vs history (typically 2.5–4.5x) | Mid-range | ROE changes shift the fair band |
| Earnings yield vs 10-year G-sec | 1/PE vs yield | ~5.1% vs 7.07%: a negative gap — equities offer less current yield than bonds | India has almost always had a negative gap because of growth; use the change in the gap, not its sign |
| Market cap / GDP ("Buffett indicator") | Total mcap ÷ nominal GDP | Elevated vs history (India has listed more of its economy; new listings inflate it) | Structural drift; compare across countries with care |
| Small-cap vs large-cap premium | Nifty Smallcap 250 P/E ÷ Nifty 50 P/E | Above 1 for most of 2023–25; compressed in the 2025 correction; check current | Small-cap indices include loss-makers; use median P/E |
| Breadth | % of stocks above 200-day averages; advance/decline | — | Technical; useful as a sentiment thermometer, not a valuation |
| Implied vol (India VIX) | Level vs history | Low VIX at high valuations = complacency | Short-dated; regime indicator only |
None of these predicts the next year. Their use is to know where in the range the market sits, and hence how much of any stock's multiple is the market's mood rather than its merits.
2. Flows¶
| Flow | Source | 2025–26 reading | Why it matters |
|---|---|---|---|
| FPI net equity flows | NSDL/SEBI, daily | Net outflows of ~₹1.8 lakh Cr in FY26; selling persisted into 2026 (Multibagg); FPI ownership 15.8% of NSE-listed, a 15-year low | Historically the marginal buyer; now the marginal seller |
| DII net flows (MFs, insurers, pension) | SEBI/AMFI, monthly | Record ₹8.5 lakh Cr in FY26, of which mutual funds ₹6.4 lakh Cr; ownership 17% — a record (SEBI annual report via OmmCom); H1 2026 DII inflows ₹4.3 lakh Cr (Business Standard) | The regime: domestic savings, not foreign capital, set prices |
| SIP flows | AMFI, monthly | ₹25,000+ Cr/month (verify current) | Sticky, but not permanent; SIP stoppage ratios rise in drawdowns |
| Retail direct participation | Demat accounts; NSE active clients; F&O volumes | Demat accounts > 20 crore (verify); F&O activity cut by SEBI's 2024–25 measures | Retail is now a structural holder and a sentiment gauge |
| Primary market | SEBI: IPOs, QIPs, OFS by month | 2024–25 record IPO volumes; NSE's ₹22,562 Cr IPO in Sep-2026 (5.7x subscribed) | Supply absorbs flows; a heavy IPO calendar at high valuations is a late-cycle marker |
| Promoter selling | SAST/bulk deals | Elevated in 2024–25 | Informed sellers meeting uninformed buyers |
The flow regime cuts two ways. Domestic flows have made Indian valuations resilient to FPI selling — the 2025–26 experience — and they have made the market dependent on a single behaviour (households' monthly SIP) that has never been tested by a prolonged bear market. The base-rate question is not "will SIPs stop?" but "what would the market look like if SIP inflows halved for a year?" — and the honest answer is that nobody knows, because it has not happened.
3. IPO cycles and primary-market froth¶
Primary markets follow a recognisable sequence: quality issuers at reasonable prices → strong listings → a rush of lower-quality issuers at higher prices → retail oversubscription and grey-market premia → listing-day pops attracting more supply → a few high-profile failures → the window closes. Markers of the late stage: subscription multiples in the hundreds; SME IPOs with tiny floats oversubscribed 500x (2023–24); IPOs priced on "basis of offer price" arguments that reference no earnings; promoters selling most of the issue (OFS-heavy deals); and post-listing underperformance across the cohort (2021's new-age tech — case I8). The rule from 03.5: read the DRHP, note the OFS share and the pre-IPO placement prices, and treat listing-day trading as noise.
4. India's small-cap cycles¶
| Cycle | Run-up | Drawdown | What was said at the top | What survived |
|---|---|---|---|---|
| 2003–08 | Small caps up several-fold on the capex/credit boom | ~−70–75% peak to trough (2008–09) | "India decoupled"; infrastructure super-cycle | Companies with cash flow and no debt; investors with cash |
| 2013–18 | Small-cap index roughly tripled to Jan-2018 | ~−40–50% (2018–20), with many individual names −80%+ | "Quality at any price"; "formalisation" | Genuine compounders bought before 2016; the "quality" cohort de-rated but survived; the fraud cohort (Manpasand, Vakrangee, etc.) did not |
| 2020–24 | Post-COVID rally; small caps up ~4–5x from the March-2020 low; retail participation and SIPs | 2024–25 correction of ~−25–30% in the small-cap index with wider dispersion (verify magnitudes) | "Amrit kaal"; defence/rail/PSU narratives; SME mania | Names with earnings and governance; the narrative cohort de-rated |
Patterns: each cycle has a narrative that justifies the multiple, a flow that funds it, a supply response (IPOs, QIPs, promoter selling) that meets the flow, and a governance tail that is exposed on the way down. The survivors were bought with a margin of safety before the narrative took hold and were held through the drawdown because the thesis, not the price, was the reason for holding.
5. The pendulum — and what to do about it¶
Howard Marks's image: markets swing between greed and fear, euphoria and depression, credulousness and scepticism — and spend little time at the "fair" midpoint. The investor's job is to notice where the pendulum is (the measures above), not to predict when it turns. Behaviour by position:
| Pendulum position | Signs | Behaviour |
|---|---|---|
| Euphoria | Valuations at historical highs; IPO froth; "this time is different"; retail leverage; low VIX; every idea passes the kill test at a price you would not pay | Raise cash by refusing to buy without margin; trim positions above the bull case; tighten thesis-breakers; do not short (timing); keep researching — the watchlist is being built for later |
| Optimism | Above-median valuations; flows strong; earnings good | Normal process; smaller sizes; more attention to what the price implies |
| Midpoint | Median valuations; mixed flows | Bottom-up work dominates; the market is not the story |
| Pessimism | Below-median valuations; FPI selling; earnings downgrades; "value traps everywhere" | Deploy watchlist triggers; sizes toward the upper bound of the rules; expect to be early |
| Depression | Historical lows; forced selling; liquidity gone; "equities are dead" | Buy the best businesses at any size the rules allow; accept that the bottom cannot be identified except in hindsight |
Where is September 2026? Below-median index valuations on a slowing earnings cycle, record domestic flows, FPI pessimism, a completed small-cap correction and a still-heavy primary calendar: optimism in flows, pessimism in earnings, midpoint in valuation — a market where the bottom-up work decides, and where the flow regime is the risk to name in every portfolio pre-mortem.
Trader's lens
Sentiment measures are the equity market's vol surface: level (valuation bands), skew (small-cap premium, IPO froth), and term structure (flow persistence). You already know that the surface tells you the price of risk, not its direction, and that the trade is regime-relative positioning — smaller when risk is cheap and everyone is long, larger when it is dear and nobody is. The same applies here; the only difference is the timescale, which is years, and the discipline, which has to survive being early.
Common mistakes
- Timing the market on valuation bands (they say "expensive", not "when").
- Confusing a low index P/E at peak earnings with cheapness.
- Treating domestic flows as permanent — or as about to reverse.
- Selling everything in euphoria (missing the last, largest leg) or buying nothing in depression (missing the turn) — the rules are about sizes and margins, not all-or-nothing.
- Reading sentiment from the same media that manufactures it.
Key terms¶
| Term | Meaning |
|---|---|
| Valuation band | Historical range of an index multiple with median and dispersion |
| Earnings yield gap | Earnings yield minus the 10-year G-sec yield |
| Market cap / GDP | Total equity market value relative to the economy |
| Breadth | Share of stocks participating in a move |
| India VIX | Implied volatility index for the Nifty |
| FPI / DII | Foreign / domestic institutional investors |
| SIP stoppage ratio | SIPs discontinued as a share of new registrations |
| Grey-market premium (GMP) | Unofficial pre-listing price premium on an IPO |
| Pendulum (Marks) | The oscillation of market psychology between extremes |
| Flow regime | The prevailing marginal-buyer structure |
Check your understanding¶
- The Nifty trades at 19.5x trailing earnings, below its 5-year median. Give two reasons this may not be "cheap".
Answer
Trailing earnings may be at a cycle peak (FY25–26 margins), so forward P/E on normalised earnings is higher; and the earnings yield (~5.1%) is below the 7.07% G-sec yield, so equities are not cheap relative to bonds — the "cheapness" is only relative to the market's own recent history.- What would a halving of SIP flows for a year do, mechanically, and which holdings would feel it most?
Answer
Remove the marginal buyer that has absorbed FPI selling; small caps and high-multiple, DII-heavy names would de-rate most; liquidity would thin. Test each holding's ownership (MF share) and its multiple's dependence on the flow regime.- A watchlist name hits your trigger price during a small-cap crash. The thesis is intact but liquidity is thin. What do the rules say?
Answer
Buy, within the liquidity limit and in tranches (expect to be early); do not raise the size above the max-loss rule because "it's cheap"; the cash you held for this is the reason you held it.- Name three late-cycle markers in the primary market.
Answer
Subscription multiples in the hundreds (especially SME), OFS-heavy issues by promoters and PE sellers, and pricing arguments that reference no earnings (GMV, TAM); post-listing underperformance across a cohort confirms it.Go deeper¶
- Howard Marks, Mastering the Market Cycle (2018) and the memos "On the Couch" and "The Pendulum".
- SEBI annual reports and monthly bulletins — flows, ownership, primary-market data.
- PrimeInvestor / IndexPE / Trendlyne — free Nifty valuation-band histories (check their EPS methodology).
- Case I8 and I14 — a primary-market cycle and a small-cap cycle from the inside.
← Previous: 12.1 Macro for equity investors · Module index · Next: 12.3 Special situations →