Investors often face a perplexing dilemma when dividend-focused portfolios experience sharp downturns, particularly when those declines occur during periods of seemingly attractive yields. Two seasoned fund managers recently dissected this phenomenon, sharing their journeys through market turbulence and offering a playbook for both active and quantitative dividend strategies. Their combined insights provide a comprehensive roadmap for understanding and navigating the complexities of dividend investing in today's evolving markets.
Zhang Xueming, a quantitative investment expert, described his personal journey as moving from conviction to doubt and back to conviction. He noted that historical research on dividend strategies showed significant retreats only when portfolio dividend yields were relatively low, but this time was distinctly different. The drawdown occurred when dividend yields were already high, a scenario rarely seen in A-share history or even in overseas markets like the US and Japan. Initially, he viewed this as an excellent opportunity, but as the decline persisted, he began to question whether past data had become invalid. By revisiting longer-term and broader market samples, he ultimately reaffirmed his trust in the fundamental principles underlying dividend strategies.
Liu Yong, an active fund manager, recalled May and June as the darkest hour for the dividend style, with relevant indices dropping over 10% in a short period. Concurrently, tech stocks kept soaring, amplifying investor disappointment through stark contrast. He emphasized that a deep understanding of their portfolio companies' operational quality and valuation levels reassured them, as continued declines on already reasonable prices actually enhanced long-term buying value. Through rigorous research into their holdings, they maintained conviction in holding quality assets.
Regarding whether investors should abandon dividend strategies due to discomfort, Zhang firmly believes they should not. Historical annualized volatility and maximum drawdown data show that dividends perform relatively better compared to broad-based indices and most equity funds. Expanding the sample to the US, Japan, the UK, and Europe yields consistent conclusions: dividend strategies remain among the most stable equity approaches. He attributes the current perceptual bias to two key factors. First, dividends have only recently gained mainstream popularity; large drawdowns occurred in 2015 and 2018, but fewer investors held or paid attention then, making this the first real experience with dividend sector retreats for most. Second, this year's drawdown happened during rapid tech growth and an overall strong market. In contrast, during 2018, when dividends also fell, broad indices and growth sectors declined even more, making dividend stability intuitive. This time, however, cross-sector comparisons have magnified the pain of relative underperformance.
Explaining the underlying logic of dividend strategies, Zhang introduced the dividend discount model. While individual companies rarely achieve perpetual operation and dividends, indices can sustain dividend continuity by periodically rebalancing into a basket of high-yield stocks. The CSI Dividend index directly selects companies with higher dividend yields based on average yields over the past three years. In contrast, the dividend low-volatility index first filters high-dividend companies, then picks those with lower volatility, sometimes further excluding companies with unstable earnings. Adding the low-volatility factor typically reduces exposure to cyclical sectors like coal and non-ferrous metals, shifting toward banks, utilities, and some consumer industries where dividend sustainability is often stronger. From a quantitative perspective, the low-volatility factor pairs exceptionally well with the dividend factor, creating a synergistic effect greater than the sum of its parts.
Liu Yong differentiated active dividend investing as not simply selecting the highest dividend yield stocks. Instead, it involves long-term tracking and research to identify companies with high dividend certainty, reasonable valuations, and room for future growth. He focuses on several key aspects: clear business models with barriers from resources, licenses, or competitive landscapes; corporate governance demonstrating management's willingness to protect minority shareholders and sustain dividends; earnings quality metrics like ROE and free cash flow; and whether capital expenditure is entering a downward phase, freeing cash for shareholder returns.
Liu further elaborated that dividend investing fundamentally aims to buy quality assets at reasonable or even low prices. While markets typically price dividend assets based on dividend yields, different assets have varying operational lifespans and certainty, requiring distinct valuation methods. For instance, hydropower assets with perpetual attributes can reference long-term government bond yields with a risk premium, whereas expressways with limited operational periods warrant higher discounts. For cyclical dividend stocks centered on upstream resources, earnings fluctuate more significantly. If a company is at a cyclical peak, current profits and dividends may look excellent, making static dividend yields appear high, but future price declines could reduce payouts. Cyclical dividends require analyzing supply-demand balance sheets over the next one to three years based on commodity cycle logic.
Zhang highlighted that static dividend yields most easily distort on cyclical products. Industries included in indices at peak profits and dividends, followed by cyclical downturns, may face lagged adjustments based on three-year historical data, potentially purchasing at cycle tops. The space for quantitative dividends lies in further screening within high-dividend stock pools for companies capable of maintaining or increasing payouts. He categorizes companies into distinct types. Natural dividend assets like hydropower and expressways have relatively stable operations, making future earnings and dividends easier to estimate, combined with internal and external analyst forecasts. Market-oriented dividend assets in mature phases, such as home appliances, consumer goods, and pharmaceuticals, can be modeled using income statements, balance sheets, and cash flow statements, focusing on operational cash flow stability, capital expenditure contraction, and whether expansion still demands significant capital. Textual information from announcements and reports also helps assess dividend willingness changes.
Identifying pitfalls, Zhang warned of two red flags. First, companies with high static dividend yields but impending earnings and dividend declines, seen in real estate, shipping, and some cyclical industries that entered indices at payout peaks only to see earnings fall later. Second, financial and governance risks where superficially high yields mask poor financial statement quality. Quantitative research cannot conduct deep dives like active managers, so diversification and multidimensional verification from financial, textual, and price-volume information mitigate single-stock risks to portfolios.
From an active perspective, Liu noted that earnings trends can often be judged through long-term industry and company tracking, but governance and dividend willingness are harder to grasp. Companies that suddenly cut dividends without adequate communication often face very negative market reactions. Active managers must spend considerable time tracking governance realities, including whether state-owned enterprises face market value management constraints and how private enterprise shareholders view company value.
Zhang elaborated on the complementary nature of active and quantitative approaches. Active investing excels at identifying management quality, governance, and dividend willingness through research, while quant translates these experiences into rules for broader validation and execution. Historically, much active investment language was difficult to encode, but with AI large model development, unstructured information from announcements, research, and management statements increasingly converts into quantitative signals. Conversely, quant can surface anomalies from market-wide prices, volumes, and capital flows for active teams to investigate deeply. The final product is not a simple combination of two portfolios but a dynamic allocation based on product goals, benchmarks, and tracking error, refined through continuous interaction.
Liu emphasized that active managers feed opportunities and risks from research to quant teams, while quant backtesting helps verify subjective judgments. For active investing, quant provides a broader market perspective; for quant, active research supplies information that pure data cannot cover.
Zhang observed that dividend investing across nearly all markets evolves from traditional to market-oriented dividends. Early dividend assets concentrated in easily identifiable traditional sectors like banks and utilities, but as economies and industries mature, more companies once considered growth stocks transition to stable operations, lower capital expenditure, and higher payouts. Dividend investing itself is a product of an economy reaching a certain stage. During high growth, industries need capital expenditure, favoring profit reinvestment; as industries mature, competition stabilizes, capital spending falls, free cash flow rises, and high-dividend companies naturally multiply. Seven to eight years ago, few Chinese companies had three consecutive years of dividends with yields above 3%, but that universe has expanded significantly. Looking ahead, consumer, pharmaceutical, home appliance, and niche manufacturing leaders could become new dividend sources. Compared to the US market, China's industries boast more segment leaders, potentially allowing for more diversified dividend asset distribution, offering greater opportunities for both active and quantitative stock selection.
For investors choosing products, Zhang advised that for active dividend funds, one should examine whether managers have long-term dividend investing experience and genuinely prioritize dividend yields rather than deviating into non-dividend styles. For quantitative dividend funds, attention should focus on differences in investment philosophies and yield prediction methods. While long-term return gaps between different dividend strategies may not be enormous, their interim paths can differ significantly. When allocating across dividend products, investors should first understand why a product has outperformed or underperformed, then supplement with products that have temporarily lagged.
Liu suggested focusing on long-term risk-return metrics like Sharpe and Calmar ratios, as well as drawdown control and recovery speed, for better holding experiences. For index and quantitative products, tracking error and style deviation matter; for active products, a deeper understanding of the fund manager is essential. He also echoed that investment involves risk, requiring caution, and that all content serves as reference only, not investment advice, with past performance not indicating future results.
Comments