Net Debt/EBITDA Screen in India: 25 Years, 21.6% CAGR, and Why It Works
The Net Debt/EBITDA screen that underperforms on the full US market produced 21.55% CAGR on Indian stocks over 25 years, 10.43 percentage points above the Sensex annually. We tested it across BSE and NSE from 2000 to 2025, with full transaction costs and next-day execution. Here's what the data shows and why the same signal behaves so differently depending on which market you're in.
Contents
- Method
- What We Found
- Year-by-Year
- Why the Signal Works in India
- The Sensex Benchmark Context
- Run It Yourself
- Limitations
Method
Signal: Net Debt/EBITDA < 2x and > -5x, ROE > 10%, Market Cap > ₹20B (~$240M USD). Top 30 stocks by lowest ratio, equal weight. Universe: BSE + NSE combined (India). All exchange-listed stocks meeting the filters. Period: 2000-2025 (103 quarterly periods). Rebalancing: Quarterly (January, April, July, October). 45-day filing lag for point-in-time data. Execution: Market-on-close next trading day (signal generated at close, executed at next close). Transaction costs: Size-tiered model applied. Benchmark: BSE Sensex. The correct local-currency benchmark for measuring alpha in Indian equities. SPY comparison is provided in the global comparison post. Data: Ceta Research (FMP financial data warehouse).
What We Found
Strategy: 21.55% CAGR. Sensex: 11.12% CAGR. Excess return: +10.43% per year over 25 years.
A $10,000 investment grew to $1,522,314. The same money tracking the Sensex grew to $151,080.
| Metric | Strategy | Sensex |
|---|---|---|
| CAGR | 21.55% | 11.12% |
| Total Return | 15,123% | 1,410.8% |
| Max Drawdown | -50.2% | -51.3% |
| Annualized Volatility | 29.9% | — |
| Sharpe Ratio | 0.504 | — |
| Sortino Ratio | 1.042 | — |
| Beta (vs Sensex) | 1.057 | 1.0 |
| Alpha (annualized) | +10.17% | — |
| Up Capture | 129.4% | — |
| Down Capture | 73.3% | — |
The up/down capture ratio tells the story: the strategy captures 129% of the Sensex's gains and only 73% of its losses. That asymmetry is strong. It means the portfolio tends to amplify market upswings while limiting participation in downturns. The max drawdown of -50.2% is slightly better than the Sensex's own -51.3%.
Year-by-Year
The annual return pattern shows the character of this market: large positive years, sharp but contained drawdown years. The first two years (2000-2001) the screen held cash, fewer than 10 Indian stocks with ₹20B+ market cap passed the combined leverage and profitability filters in those early years.
| Year | Strategy | Sensex | Excess |
|---|---|---|---|
| 2000 | 0.0% | -25.2% | +25.2% |
| 2001 | 0.0% | -18.6% | +18.6% |
| 2002 | -5.5% | +2.9% | -8.5% |
| 2003 | +127.0% | +79.1% | +47.9% |
| 2004 | +25.0% | +10.8% | +14.2% |
| 2005 | +48.4% | +40.6% | +7.8% |
| 2006 | +47.5% | +48.5% | -1.0% |
| 2007 | +66.7% | +46.8% | +19.9% |
| 2008 | -46.6% | -51.3% | +4.7% |
| 2009 | +104.5% | +76.3% | +28.2% |
| 2010 | +45.2% | +17.1% | +28.1% |
| 2014 | +65.5% | +33.5% | +32.0% |
| 2017 | +49.1% | +27.1% | +22.0% |
| 2022 | +10.2% | +3.4% | +6.8% |
| 2023 | +47.6% | +17.5% | +30.1% |
| 2024 | +30.5% | +11.2% | +19.3% |
2003 stands out: +127.0% in a single year vs the Sensex's +79.1%. India's equity market re-rated sharply from 2003 onward as the economy opened up, FII inflows accelerated, and the "India growth story" gained momentum. Low-leverage companies with strong fundamentals were among the first to be re-priced.
2014 shows +65.5% vs Sensex +33.5%, coinciding with the Narendra Modi election landslide. Low-leverage quality companies re-priced sharply as institutional capital flowed in.
2008 is notable: the strategy fell 46.6% vs the Sensex's 51.3%. During the global financial crisis, conservatively leveraged Indian companies held up slightly better than the broad market.
Why the Signal Works in India
1. The Indian market hasn't been dominated by capital-light tech megacaps.
The US underperformance story is essentially about mega-cap tech companies (Nvidia, Meta, Alphabet, Apple) moving SPY dramatically faster than any fundamentals-based screen can match. India's BSE/NSE don't have that dynamic to the same degree. The top performers in India over this period include Reliance Industries, HDFC Bank, TCS, Infosys, and pharma companies, many of which qualify on the Net Debt/EBITDA screen.
2. Financial conservatism matters more in an emerging market.
India has had multiple episodes of credit stress, currency depreciation, and sector-specific leverage crises (IL&FS in 2018, infrastructure companies in 2011-2012). During those periods, companies with clean balance sheets saw dramatically less equity damage than their leveraged peers. The signal captures this protective characteristic more reliably than in the US, where the corporate bond market provides refinancing lifelines more easily.
3. The quality filter (ROE > 10%) is more selective in India.
In the US, thousands of large-cap companies pass ROE > 10%, it's a weak filter. In India, consistently earning above 10% ROE across market cycles requires genuine competitive advantage. The combined filter (low leverage + real profitability) is more discriminating.
4. Cash periods reflect a disciplined filter, not data gaps.
The strategy held cash in 10 of 103 quarters (10%). These mostly cluster in the early 2000s (2000-2001) when fewer Indian large-caps passed the ₹20B market cap threshold, and occasionally in market stress periods. The cash discipline helped: holding cash in 2000-2001 avoided two consecutive negative years.
The Sensex Benchmark Context
A 21.55% CAGR on BSE+NSE deserves honest framing. Indian equities as a whole performed exceptionally from 2000 to 2025. The BSE Sensex compounded at 11.12% annually over this period, driven by rapid economic growth, demographic expansion, and re-rating from emerging-market discount to premium.
The strategy adds 10.43% above the Sensex annually, which is genuinely strong alpha. The Sortino Ratio of 1.042 confirms the downside-risk-adjusted performance is excellent. The down capture of 73.3%, capturing less than three-quarters of the Sensex's drawdowns, shows this is not purely riding India's bull market. The screen adds value within the Indian market, not just benefiting from India-level beta.
The practical question isn't "does this beat SPY", it's "does the signal add value within the Indian market?" The down capture ratio of 73.3% and the 10.43% excess over the Sensex says it does.
Run It Yourself
Current qualifying stocks (India, TTM data):
SELECT
k.symbol,
p.companyName,
ROUND(k.netDebtToEBITDATTM, 2) AS net_debt_ebitda,
ROUND(k.returnOnEquityTTM * 100, 1) AS roe_pct,
ROUND(k.marketCap / 1e9, 1) AS mktcap_bn
FROM key_metrics_ttm k
JOIN profile p ON k.symbol = p.symbol
WHERE k.netDebtToEBITDATTM < 2.0
AND k.netDebtToEBITDATTM > -5.0
AND k.returnOnEquityTTM > 0.10
AND k.marketCap > 20000000000 -- ₹20B (~$240M USD)
AND p.exchange IN ('BSE', 'NSE')
ORDER BY k.netDebtToEBITDATTM ASC
LIMIT 30
Run this query on Ceta Research
Full backtest:
cd backtests
python3 net-debt-ebitda/backtest.py --preset india --verbose
Limitations
Benchmark change. Previous versions of this analysis benchmarked against SPY. We now use the Sensex as the correct local-currency benchmark for measuring alpha. The +10.43% excess vs the Sensex is lower than the +13.24% vs SPY reported in earlier versions, because India's equity market itself outperformed US equities over this period.
Survivorship and data coverage. The FMP data includes delisted Indian stocks (survivorship bias is partially addressed), but coverage before 2005 is thinner.
Concentration. A 30-stock equal-weight portfolio in India has meaningful sector concentration, historically heavy in industrials, banks, and pharma. This isn't a diversified country-fund equivalent.
The 2003 return. +127.0% in a single year is a large outlier. Without it, the long-run CAGR would be roughly 16-17% rather than 21.55%. The signal is real, but some of the headline number comes from one extraordinary year.
Part of a series. This is one of 22 exchanges we tested. The comparison post shows where the signal works and where it doesn't, including the underperforming markets.
Data: Ceta Research (FMP financial data warehouse), 2000-2025. Full methodology: backtests/METHODOLOGY.md. Backtest code: backtests/net-debt-ebitda/.
