Defensive Quality Across 11 Markets: Where the Signal Works and Why
Defensive Quality Across 11 Markets: What Works, What Doesn't, and Why
We ran the same defensive quality screen across 16 global exchanges from 2000 to 2024. The signal: quality stocks in Consumer Defensive, Utilities, and Healthcare sectors, ranked by dividend yield.
Contents
- Method
- The Results
- Markets with strong defensive quality performance
- Markets with modest or mixed results
- Markets excluded from content
- Why the Signal Varies by Market
- What's Consistent Across Markets
- Down capture is negative everywhere it works
- Quality filter does real work
- Cash rule is protective
- Comparison Table (Full)
- Limitations
- Takeaway
The results split into three clear groups. A set of markets where defensive quality compounds reliably with low drawdown (US, India, UK, Hong Kong). A second group where the strategy produces modest returns without edge (Germany, Japan, Canada, Switzerland). And markets where the defensive sector universe is too thin to produce investable results (Taiwan, Sweden, South Africa).
The main finding: defensive quality isn't uniformly defensive across markets. Market structure, sector composition, and local dividend culture determine whether the signal works.
Method
- Data source: Ceta Research (FMP financial data warehouse)
- Universe: 16 global exchanges, market cap thresholds in local currency
- Period: 2000–2024 (25 annual periods)
- Signal: Quality screen (ROE > 6%, OPM > 8%, D/E < 2.5, DivYield > 0.5%), ranked by dividend yield
- Sectors: Consumer Defensive, Utilities, Healthcare
- Benchmark: S&P 500 Total Return (SPY, global reference)
- Portfolio: Top 30 by dividend yield, equal weight. Cash if < 10 qualify.
Full methodology: backtests/METHODOLOGY.md
The Results
Markets with strong defensive quality performance
| Exchange | CAGR | Excess vs SPY | Sharpe | Max DD | Cash | Avg Stocks |
|---|---|---|---|---|---|---|
| India (BSE+NSE) | 17.23% | +9.40% | 0.416 | -6.6% | 16% | 28.1 |
| UK (LSE) | 12.63% | +4.80% | 0.666 | -10.0% | 0% | 13.6 |
| US (NYSE+NASDAQ+AMEX) | 9.80% | +1.97% | 0.608 | -19.9% | 0% | 22.8 |
| Hong Kong (HKSE) | 8.62% | +0.79% | 0.280 | -19.2% | 12% | 22.3 |
India: best absolute returns, driven by pharma compounders and FMCG quality names. The defensive label here refers to real defensive businesses. Hindustan Unilever, Sun Pharma, Cipla, not sector-label coincidences.
UK: best Sharpe ratio globally. Thin universe (13.6 stocks) concentrated in Unilever, AstraZeneca, GSK, regulated utilities. These are some of the world's most durable consumer businesses, and the quality screen picks them reliably.
US: negative down capture (-9.5%), the most unusual metric in the entire study. When SPY falls on average, this portfolio gains. Consistent defensive protection with strong absolute returns.
Hong Kong: modest excess (+0.79%), but positive. HKSE defensive stocks are a mix of Chinese consumer multinationals listed in Hong Kong and local utilities. The strategy works but with higher drawdown (-19.2%) than UK or US.
Markets with modest or mixed results
| Exchange | CAGR | Excess vs SPY | Sharpe | Max DD | Cash | Avg Stocks |
|---|---|---|---|---|---|---|
| Germany (XETRA) | 7.66% | -0.17% | 0.424 | -35.8% | 0% | 21.9 |
| Canada (TSX) | 7.09% | -0.75% | 0.328 | -22.4% | 16% | 11.8 |
| Japan (JPX) | 6.89% | -0.94% | 0.440 | -22.8% | 16% | 28.1 |
| Switzerland (SIX) | 6.62% | -1.21% | 0.514 | -24.8% | 0% | 15.9 |
| Thailand (SET) | 5.24% | -2.59% | 0.170 | -26.7% | 24% | 22.1 |
| China (SHZ+SHH) | 5.14% | -2.69% | 0.072 | -55.6% | 0% | 26.6 |
Germany: near-SPY returns (-0.17% excess), but with a Sharpe of 0.424 and zero cash periods. The strategy holds its own but doesn't add edge over passive. Germany's defensive sector includes Henkel, Beiersdorf, and a handful of pharma names.
Japan: similar picture. Large qualifying universe (28.1 stocks) but modest underperformance (-0.94% excess). Japan's defensive sector is dominated by consumer goods exporters whose returns are tied to yen dynamics and global demand rather than domestic defensive characteristics.
Switzerland: negative excess (-1.21%) but solid Sharpe (0.514). Nestle, Novartis, and Roche dominate the universe. These are globally dominant businesses, but at Swiss valuations, dividend yields compress and the strategy's income-first ranking produces modest results.
China: the most volatile in the set. -55.6% max drawdown, driven by the 2007-08 crash after the Shanghai bull market (+110% in 2006, -54% drawdown from peak). The Chinese defensive sector doesn't behave defensively in A-share market dynamics, correlation with speculative market cycles overwhelms sector fundamentals.
Korea: 3.70% CAGR with 44% cash periods. The defensive sector universe is thin in Korea (12.8 avg stocks when invested). High cash drag and modest returns when invested.
Malaysia: 3.02% CAGR, 32% cash. Similar thin universe problem.
Indonesia: 2.66% CAGR, 44% cash. Low conviction signal.
Markets excluded from content
Taiwan (TAI+TWO): 76% cash periods. The defensive sector universe in Taiwan is too small to run the strategy effectively for 19 of 25 periods.
Sweden (STO): 88% cash. Only 3 invested periods. The Swedish exchange doesn't have enough qualifying defensive stocks in this universe size filter.
South Africa (JNB): 100% cash. The JNB defensive sector universe historically never reached 10 qualifying stocks. Excluded from all content, not a data quality issue, just a genuinely thin universe for this strategy.
Why the Signal Varies by Market
Three factors drive the cross-market differences:
1. Sector composition varies more than sector labels suggest. "Consumer Defensive" in India means FMCG giants with decades of pricing power. "Consumer Defensive" in Germany includes Henkel (adhesives, household cleaners) and Beiersdorf (NIVEA), quality businesses but with different competitive dynamics. The same quality filter applied to both produces very different underlying portfolios.
2. Dividend culture matters for ranking. The strategy ranks by dividend yield. In markets with strong dividend cultures (UK, India's quality tier, US utilities), the income signal is informative, high-yielding defensive companies are genuinely financially strong and undervalued by growth-focused investors. In markets with weak dividend cultures, the ranking becomes noisier.
3. Market structure affects "defensive" sector behavior. China's A-shares market has historically exhibited high retail participation and speculative dynamics. During bull markets, everything including defensive sectors inflates. During crashes, everything including defensive sectors falls hard. The sector label doesn't provide the same protection as in markets with more institutional participation.
What's Consistent Across Markets
Down capture is negative everywhere it works
| Exchange | Down Capture | Interpretation |
|---|---|---|
| India | -37.4% | Gains significantly when SPY falls |
| UK | -31.7% | Gains when SPY falls |
| US | -9.5% | Gains slightly when SPY falls |
| Hong Kong | ~0% | Near-flat when SPY falls |
The negative down capture shows up consistently in the markets where the strategy works. When global equity markets fall, quality defensive stocks in mature markets gain in relative terms, and often in absolute terms too.
Quality filter does real work
The ROE > 6%, OPM > 8%, D/E < 2.5 combination filters out yield traps, companies with high yields because they're financially distressed or because their stock price has fallen. Without quality filters, high-yield defensive screens tend to pick up financially weak utilities and consumer names that cut dividends in downturns.
Cash rule is protective
Periods with fewer than 10 qualifying stocks go to cash. This prevented the strategy from holding a handful of concentrated positions in thin markets (Korea, Taiwan, Malaysia during most periods). It also explains why high-cash markets (Taiwan 76%, Sweden 88%) underperform, the strategy correctly identified that those markets didn't have enough quality defensive stocks to invest.
Comparison Table (Full)
| Exchange | CAGR | Excess | Sharpe | Sortino | MaxDD | Cash% | AvgStocks |
|---|---|---|---|---|---|---|---|
| India (BSE+NSE) | 17.23% | +9.40% | 0.416 | 1.961 | -6.6% | 16% | 28.1 |
| UK (LSE) | 12.63% | +4.80% | 0.666 | 2.310 | -10.0% | 0% | 13.6 |
| US (NYSE+NASDAQ+AMEX) | 9.80% | +1.97% | 0.608 | 1.827 | -19.9% | 0% | 22.8 |
| Hong Kong (HKSE) | 8.62% | +0.79% | 0.280 | 0.780 | -19.2% | 12% | 22.3 |
| Germany (XETRA) | 7.66% | -0.17% | 0.424 | 0.802 | -35.8% | 0% | 21.9 |
| Canada (TSX) | 7.09% | -0.75% | 0.328 | — | -22.4% | 16% | 11.8 |
| Japan (JPX) | 6.89% | -0.94% | 0.440 | 1.847 | -22.8% | 16% | 28.1 |
| Switzerland (SIX) | 6.62% | -1.21% | 0.514 | — | -24.8% | 0% | 15.9 |
| Thailand (SET) | 5.24% | -2.59% | 0.170 | — | -26.7% | 24% | 22.1 |
| China (SHZ+SHH) | 5.14% | -2.69% | 0.072 | — | -55.6% | 0% | 26.6 |
| Korea (KSC) | 3.70% | -4.13% | 0.026 | — | -63.7% | 44% | 12.8 |
| Malaysia (KLS) | 3.02% | -4.82% | 0.100 | — | -14.1% | 32% | 13.6 |
| Indonesia (JKT) | 2.66% | -5.17% | 0.063 | — | -15.5% | 44% | 17.4 |
| Taiwan (TAI+TWO) | 0.97% | -6.86% | -0.006 | — | -11.8% | 76% | 13.5 |
| Sweden (STO) | -0.18% | -8.01% | -0.939 | — | -9.9% | 88% | 11.3 |
SPY reference: 7.83% CAGR, -36.27% MaxDD. South Africa (JNB): 100% cash, excluded.
Limitations
Currency. All returns are in local currency (INR for India, GBP for UK, etc.). Cross-market comparisons in USD would show different relative rankings depending on the period.
Benchmark. SPY is used as the global reference benchmark. Against local benchmarks (Nifty 50, FTSE 100), the excess returns would differ.
FMP data coverage varies by exchange. India data before 2004 was thin, 4 cash periods at the start. Japan and UK adjClose data became reliable in 2000. China early-period data had some quality issues filtered by the 200% single-period cap.
The strategy is sector-constrained. This is an intentional design choice, not a limitation. The result is a portfolio that genuinely reflects defensive sector behavior, not a multi-sector portfolio with a defensive tilt.
Takeaway
Defensive quality works best where defensive companies actually behave defensively: established markets with strong institutional participation, clear dividend culture, and genuinely consumer-defensive or healthcare businesses.
India stands out as the global winner on absolute terms. UK stands out on risk-adjusted terms. The US delivers the most unusual characteristic: negative down capture.
Markets with thin defensive universes (Taiwan, Sweden, South Africa) shouldn't run this strategy. Markets with speculative dynamics (China) get the sector label but not the sector behavior.
Full backtest code and SQL: github.com/ceta-research/backtests Run live screens: cetaresearch.com/data-explorer
Part of a Series: US | UK | India
Data: Ceta Research (FMP financial data warehouse). Returns in local currency per exchange. Past performance doesn't guarantee future results.
