RuiLian Founder Xu Zhongxiang: Quantitative Multi-Asset Strategies are a philosophy of "fear of death" and a science of "probability."

Ask AI · How Does Xu Zhongxiang’s Fear-of-Death Philosophy Guide Investment Practice?

Caixin Securities News (April 7) (Reporter Feng Qijuan) Today, wealth management is moving from the era of single-asset thinking to the era of multi-asset allocation. As the founder and Chief Investment Officer of Ruilian Caizhi, a co-inventor of the RAFI fundamental factor-based quantitative strategy, and a pioneer of Smart Beta strategies, Dr. Xu Zhongxiang combines years of academic grounding with hands-on experience in global asset and wealth management markets to reveal the core理念 and practice paths of multi-asset allocation.

On the front page of Dr. Xu Zhongxiang’s collected papers on multi-asset allocation, the line “Fear of death is the best strategy” stands out. Although it sounds straightforward, it distills the investment philosophy he has honed over years in the industry, and it also sets a tone of reverence for risk for the practice of multi-asset allocation.

He points out that the essence of investing is to share in the growth dividends of companies and to capture long-term value returns—certainly not short-term games of chasing asset price fluctuations or fantasies of getting rich. True wealth accumulation comes from long-term, repeatable, verifiable probability advantages. In a market full of uncertainty, the greatest risk in investing is not volatility itself, but blind confidence that comes from insufficient understanding of risk—mistaking a sense of familiarity with an asset for a safety net, misreading short-term noise as a long-term trend, and ultimately falling into the trap of irrational investing.

Xu Zhongxiang often says, “Buying broadly and allocating well is, in itself, the most scientific response plan,” but he also emphasizes that the key to scientific diversification is reducing the correlations among assets, and building an “all-weather” resilient investment portfolio. When facing market uncertainty, only by looking through the surface of assets, focusing on underlying factors, relying on long-term cycles, and sticking to systematic discipline can we cut through the fog of short-term noise and grasp the true essence of investing.

In his view, quantification is the scientific path to achieving multi-asset allocation. As a rational probability game based on the “law of large numbers,” quantification—through precise risk measurement, effective separation from emotions, and continuous iteration of strategies—upgrades investing from reliance on subjective judgment into a rigorous, verifiable, systematic process. Its core value is not creating short-term performance miracles, but accumulating long-term, repeatable competitive advantages. He frankly acknowledges that the market will not provide immediate, correct feedback, and in the short term it is difficult to distinguish ability from luck.

Xu Zhongxiang believes that the key to crossing cycles is to adhere to long-term investment discipline that runs counter to human nature, and to abandon “pseudo-diversification” and stop gambling on one-way bets. In today’s environment, where there is a perception lag between advanced ideas and market realities, the core mission of professional asset and wealth management institutions is to help investors build low-correlation, volatility-resistant portfolios that can be held long term—embracing certainty amid the unknowable, and achieving steady wealth growth through long-term compounding.

Wealth Management Enters the Era of Multi-Asset Allocation

Xu Zhongxiang directly states that the wealth management landscape is undergoing profound change; a single asset or strategy can no longer meet investors’ needs, making multi-asset allocation an inevitable direction.

Practical experience from mature markets shows that the essence of multi-asset allocation is risk management, not precise prediction. The focus is not on picking the next “champion” among many assets and going overweight, but on fundamentally avoiding a “gambling mindset” and avoiding fatal risks such as redemptions triggered by short-term volatility and leverage blowups.

For asset management institutions, the key is to deeply understand the logic of price movements and long-term profitability foundations of different assets under various macroeconomic environments, and then build a resilient portfolio composed of high-quality assets accordingly. The goal is not to win every round of market moves, but to ensure the ability to stay in the market for the long run. By sharing in the long-term power of corporate growth and compounding, investors can capture long-term risk premia.

He points out that the profound changes in the global landscape are driving traditional allocation and hedging logic to evolve in parallel:

First, the dollar’s safe-haven function is weakening; gold has moved from a traditional tool for geopolitical hedging to a core asset for hedging against fiat currency credit and the risk of dollar weaponization;

Second, the global trade order is being reshaped; overreliance on a single market highlights risk. While positioning across multiple markets may face short-term pressure, in the long run it will bring greater pricing power and strategic initiative;

Third, domestic market recognition needs correction: China’s A-share domestic asset base is solid and residents’ savings have a broad migration space; in the long run, development relies more on endogenous drivers and institutional improvement.

Fourth, the value of globalized multi-asset allocation is becoming more prominent. Markets such as Vietnam, India, Indonesia, and Saudi Arabia have very low correlations with domestic assets. Allocating to such assets is key to building anti-volatility, low-correlation portfolios, effectively diversifying risk and capturing global growth opportunities.

However, there is a gap between advanced concepts and market practice. Xu Zhongxiang admits that many investors’ allocation needs stem from passive choices after high-yield products disappear—not from a proactive understanding of the logic of “diversifying risk and earning long-term compounding returns.” This kind of cognition is extremely unstable and can easily be overturned by short-term market hotspots, making it hard to maintain long-term investment discipline.

Therefore, he frankly says that the most difficult part of investing is to stick to long-term discipline under the persistent interference of short-term noise. This also defines the core value of professional asset and wealth management institutions: not to provide a password for short-term upside, but to help investors build and maintain an asset portfolio that can pass through cycles, control volatility, and enable long-term compounding to be realized.

Reject “Pseudo-Diversification”—How Do Multi-Assets Really Get Allocated?

To truly implement this philosophy, Xu Zhongxiang emphasizes that first you need to clarify your understanding and avoid misconceptions, and then build an effective multi-asset allocation system through scientific methods.

In practice, investors often fall into three major misconceptions, deviating from the original intent of multi-asset allocation:

Misconception 1: Thinking that “buying more” equals diversification. The core of scientific diversification is reducing correlations among assets, not simply increasing the number of holdings. For example, concentrating on technology stocks or technology-themed funds, or buying multiple funds that rank highly in different years, can also fail to achieve true risk diversification—often because the positions cluster in the same tracks and styles converge.

Misconception 2: Equating “multi-asset” with “macroeconomic market timing.” The name “multi-asset” does not match the reality: by over-relying on macro forecasts and re-pressing one single asset, it ultimately turns into a big gamble of “single-asset” trading, which contradicts the original purpose of diversifying risk.

Misconception 3: Treating “familiarity level” as the same as “risk level.” The true risk of an asset comes from price volatility and exposure to macro factors, and has nothing to do with how familiar an individual feels. This “familiarity illusion” severely blocks effective multi-asset allocation. Investors may give up allocating to “unfamiliar assets” because they try to understand every kind of asset, or they may bear a concentrated single-asset risk by investing heavily in “familiar assets.”

Xu Zhongxiang summarizes that building an effective multi-asset allocation requires following four major principles:

First, look through the surface appearance of assets and focus on the underlying driving factors. Correlation cannot be judged only by industry classifications or historical prices; instead, you should identify the macro factors, style factors, industry factors, and so on behind the assets.

Second, extend the time horizon to objectively identify risk-return characteristics. Short-term data is full of noise and easily leads to misjudgment. Only by standing on the long-term dimension can we accurately characterize an asset’s true risk-return characteristics statistically.

Third, use scenario analysis instead of short-term forecasting to construct an all-weather portfolio. Simulate how various assets perform under different macro scenarios, test the portfolio’s ability to hedge and control drawdowns under extreme conditions, and ensure that under important macro scenarios the portfolio has assets that can either fall less or benefit.

Fourth, embrace uncertainty and rely on systems and discipline. Give up the obsession of “understanding everything,” accept that the future is unknowable, and resist market volatility by building rigorous risk-control models and systematic discipline.

Quantitative Multi-Asset: Based on Probability, Emphasizing Discipline

In Xu Zhongxiang’s view, when constructing a scientific and robust multi-asset allocation framework, quantification provides a verifiable systematic path. Its core is to raise investment management from reliance on personal experience and subjective judgment to a rigorous process based on probability and emphasizing discipline, through systematic data processing and model building.

The primary goal of multi-asset allocation is risk management. Quantification has a natural advantage here because “risk” is itself a statistical concept. Human brains are not good at processing complex probability distributions and correlation calculations. Quantitative models can precisely measure dynamic correlations among assets and systematically assess the possible performance of different assets under different macroeconomic contexts, thereby providing objective support for portfolio construction.

By aggregating and analyzing massive historical data, quantitative models can systematically summarize and iterate investment experience. This approach is usually better than individual summaries in terms of analytical breadth and verifiability, helping to overcome sample biases and emotional interference that are hard to avoid in personal experience. It does not rely on narratives or short-term performance. Instead, through rigorous backtesting and validation, it verifies whether an investment logic can continuously generate measurable excess returns across different market environments—providing an objective benchmark for distinguishing “ability” from “luck.”

In specific practice, quantification understands the market by focusing on multiple categories of driving factors, including macro policy variables, market style factors, and characteristics of industry cycles, among others.

The core philosophy of quantitative modeling is to accumulate probability advantages by relying on the “law of large numbers.” No single factor is a “holy grail.” The model’s role is to integrate multiple low-correlation factors and never rely on heavy weighting based on a single signal. By making joint decisions through diversified factors, probability advantages can be realized over the long term and across repeated decision-making. Therefore, quantification is a sober probability game; the essence lies in continuously searching for factors that improve overall win rates, and兑现 system advantages through long-term execution with discipline.

However, turning this philosophy into practice faces significant challenges. Xu Zhongxiang points out that the fundamental pitfall in the investment industry is over-interpreting short-term noise as useful information. Short-term up and down moves are often random noise. If strategies are adjusted frequently based on that, it is easy to misjudge direction and fall into a low-efficiency cycle of repeated error-correction.

He also notes that another deeper challenge is not that people know their mistakes but refuse to change—rather, “attribution” itself is extremely difficult. The market will not immediately judge whether a successful outcome comes from ability or luck. A correct strategy may show temporary losses due to short-term volatility, while an incorrect decision may still earn profits at a stage due to luck. If short-term gains and losses are treated as the standard for right or wrong, it is easy to form wrong attributions, leading to systematic decision errors.

Market feedback has significant delay and is misleading—this is one of the core difficulties in strategy iteration. The market will not provide the right/wrong answer in time. The real success or failure of an investment often needs to be verified over a long period before it becomes clear.

Therefore, Xu Zhongxiang believes that true iteration must be grounded in long-term logic, test strategy effectiveness using long-cycle data and deeper logic, avoid overreacting to short-term noise, and patiently wait for reliable feedback that has been validated by time. Xu Zhongxiang emphasizes that before long-term results are revealed, you must stay clear-headed and steady, and not be led off course by accidental “rewards” or “punishments.”

Given the complexity above, quantitative strategies are often labeled as “black boxes.” In his view, the core value of a quantitative model is not to provide a simple story, but to build an analytical framework that can systematically process complex relationships and help reduce individual cognitive biases. For investors, the focus of evaluation should be placed on a strategy’s long-term risk logic and its performance under extreme market pressure, rather than digging into every technical parameter.

However, the biggest execution risk for quantitative strategies is that managers intervene in quantitative models with subjective will, deviating from the established system discipline. Regardless of whether short-term performance is good or bad, it can easily trigger human adjustments: when trading is profitable, managers may proactively raise risk exposure; when there are drawdowns, they may blindly add positions in an attempt to quickly break even. Such subjective actions essentially undermine the core logic that quantitative investing relies on—discipline, emotion-neutrality, and systematicization—causing a scientific systematic strategy to degenerate into another form of subjective decision-making, which in turn amplifies risk.

(Caixin Securities reporter Feng Qijuan)