Building CetaResearch. Data analysis for policy, research, and startups.
equity growth
We backtested the Compounding Equity Screen on NSE stocks from 2000 to 2025. 5.15% CAGR vs 12.06% for the Sensex. The strategy underperforms the local benchmark by 6.90% annually, with 8 of 25 years in cash missing India''s strongest bull market.
backtests
We ran the same cyclical timing signal across 15 exchanges over 24 years. India: +4.55% annual excess. China: -17.97% annual excess. Germany and Sweden also positive. The same revenue growth signal produced opposite outcomes in markets with different economic structures.
etf
We backtested a portfolio of the 30 least ETF-owned quality stocks on US markets over 20 years. Result: 4.56% CAGR vs 10.61% for SPY. Under-owned stocks don't outperform. Here's why.
etf
Germany is the one market where quality stocks ignored by ETFs actually work. 9.23% CAGR, 0.460 Sharpe, -29.26% max drawdown over 20 years. The Mittelstand's structural ETF blind spot creates a genuine opportunity.
etf
We backtested a portfolio of the 30 least ETF-owned quality stocks on US markets over 20 years. Result: 6.56% CAGR vs 10.61% for SPY. Avoiding passive-owned stocks doesn't help. Here's why.
etf
We backtested a portfolio of quality stocks held by fewer than 10 ETFs over 20 years. Result: 0.90% CAGR vs 10.61% for SPY. The ETF orphan thesis fails. Quality stocks outside ETFs are underowned for a reason.
earnings
India shows the highest beat-streak returns globally at +5.05% T+63. But 95% of events are from 2022-2025. We publish the results with full transparency about the data coverage. The signal is real for that period. Whether it continues is an open question.
income quality
We tested income quality (OCF/NI) on 25 years of Stockholm exchange data. High IQ returned 8.7% CAGR, beating medium (3.8%) and low (4.2%). Clean monotonic ranking, +4.5% spread. But 130% down capture and 1.057 beta mean higher volatility than SPY.
income quality
We tested income quality (OCF/NI) on 25 years of KSC data. The medium quality basket (7.1% CAGR) beat the high quality basket (5.5%). The signal works as a filter against low quality (1.4% CAGR) but the top tier underperforms, likely due to chaebol cash flow structures.
income quality
We tested income quality (OCF/NI) on 25 years of SAO data. High IQ returned 12.5% CAGR with a 72% win rate vs SPY and 44.6% down capture. But medium IQ (17.4%) beat high IQ by 5% per year. The spread vs low quality is +8.1%. Cash conversion matters in Brazil, just not the way you'd expect.
income quality
We tested income quality (OCF/NI) on 25 years of JSE data. High IQ returned 16.6% CAGR with +8.91% alpha. Low IQ lost 5.6% per year. The 22.1% spread is the widest globally. But the high IQ basket averaged just 10 stocks, so interpret with caution.
income quality
We tested income quality (OCF/NI) on 25 years of LSE data. The spread between high and low quality is just 0.5% per year. All three buckets returned nearly identical CAGRs. The UK market is too cash-generative for this signal to separate.