Background

I am based at Guojin Securities, where I focus on quantitative research, strategy methodology, program-trading applications, AI-enabled investment research, and bringing quantitative use cases into real business settings. Prior to this, I had one year of experience as a quantitative researcher and also won a gold medal in the World Quantitative Trading Competition. These experiences allowed me to gain early exposure to strategy development, data analysis, backtesting validation, trading rule construction, and model optimization. They also led me to gradually realize that quantitative finance is not simply about replacing manual trading with code, but rather a complete system supported by financial logic, mathematical modeling, statistical validation, risk control, and engineering implementation.

In actual research and business practice, I have increasingly come to understand that the core of quantitative finance is not merely about "discovering an effective strategy," but rather about understanding the sources of return, risk structure, and market environment behind the strategy. A strategy that performs well in backtests does not necessarily have long-term viability, nor does a model that works in one market regime automatically remain robust across cycles. True quantitative capability lies in identifying patterns in complex markets while knowing exactly where those patterns cease to hold.

Career Development

My career path can broadly be divided into three stages.

The first stage was the quantitative research phase. As a quantitative researcher, my initial focus was on how to describe the market with data, express trading logic with rules, and validate strategy effectiveness through backtesting. At that time, I would focus on metrics such as returns, maximum drawdown, Sharpe ratio, win rate, profit/loss ratio, turnover rate, transaction costs, and factor stability. Winning a gold medal in the World Quantitative Trading Competition deepened my understanding of strategy construction and strategic interaction in the market. Competition constantly forced me to ask which signals were truly valuable in uncertain markets and which returns were merely chance outcomes of a particular sample.

The second stage was a re-understanding of the essence of quantitative finance. As my research deepened, I gradually discovered that many strategies, while seemingly effective on the surface, might hide issues such as sample bias, overfitting, insufficient liquidity, underestimation of transaction costs, and fragility in extreme market conditions. In real markets, strategies do not rely on static data – instead they depend upon constantly evolving capital structures, market sentiment, trading regimes, and risk appetites. This stage made me realize that quantitative research should not just focus on whether a strategy performs well, but rather delve deeper into the underlying drivers: what kind of returns does the strategy generate? Do these returns stem from risk premia, behavioral biases, liquidity provision, or short-term sentiment dislocations?

The third stage began after I joined the quantitative business department at the Guojin Securities head office, where I started to combine quantitative research capabilities with real financial business scenarios. Quantitative business in securities firms is not just about strategy research itself, but also includes the construction of quantitative methodologies, optimization of investment research processes, application of program trading, establishment of risk management frameworks, and exploration of new technologies in financial scenarios. This process helped me further understand that the value of quantitative finance is not just the model itself, but whether it can form a system that is explainable, verifiable, executable, and risk managed.

The true value of quantitative finance is not just about discovering returns, but about using more rigorous methods to understand where returns come from, where risks lie, when models are effective, and under what circumstances they will fail.

The CQF played a very important connecting role in this path. It elevated my past practical experience in strategies, data, and trading to the level of financial engineering and risk modeling. In the past, I focused more on "whether the strategy is effective"; now, I pay more attention to "why the strategy is effective, when it is effective, why it fails, how risks are exposed, and whether model assumptions hold true."
 

Industry Insights

I believe that China's quantitative finance industry is undergoing a significant transformation. In the past, quantitative finance was largely considered a tool for private fund managers, asset management institutions, and leading financial institutions, with high barriers to entry and far removed from ordinary investors. However, in recent years, with the development of quantitative platforms, cloud-based backtesting environments, large AI models, and program trading tools, quantitative capabilities are gradually moving from within institutions to a broader range of market participants.

However, the popularization of quantitative tools does not mean the popularization of quantitative capabilities. Tools can lower the barrier to entry, but they cannot replace professional judgment. Especially in the Chinese market, price fluctuations often involve not only fundamental changes but also the combined effects of capital structure, trading systems, market sentiment, risk preferences, and liquidity constraints. Truly valuable quantitative research is not simply fitting historical data into a model, but rather identifying, amidst seemingly chaotic market fluctuations, which signals come from real capital behavior, which are merely noise, which opportunities are statistically significant, and which are just temporary illusions.

I have always believed that market sentiment is not the opposite of quantitative analysis; on the contrary, sentiment itself is a variable that can be observed, decomposed, and modeled. Especially in the A-share market, sentiment cycles are often very intuitively reflected in short-term capital behavior. For example, when a new industry theme, policy catalyst, or event shock emerges in a certain period, capital often first concentrates on a few highly recognizable targets, forming the first batch of limit-up stocks; subsequently, if the profit-making effect is recognized by the market, capital will further spread to the same sector, same concept, and upstream/downstream industrial chains, manifested by an increase in the number of limit-up stocks, higher consecutive limit-up streaks, increased sector turnover, and stronger buying support for core targets' turnover. This process, superficially appearing as "theme speculation," is essentially, from a quantitative perspective, a process of rising risk appetite, concentrated liquidity, strengthened group behavior, and the formation of market consensus.

Taking the A-share short-term ecosystem as an example, indicators such as tiers of limit-up stocks, consecutive limit-up streaks, limit-up success rate, limit-up failure rate, continuation rate, increased trading volume, turnover structure, sector diffusion extent, number of limit-down stocks, and performance of yesterday's limit-up stocks can all serve as variables for observing market sentiment. When the market is in an upward sentiment phase, it typically exhibits characteristics such as continuous advancement of core targets, diffusion to follow-on stocks, less sector divergence, and willingness of capital to accept high-level turnover; when sentiment enters a divergence phase, one often sees an increase in high-level stock limit-up failures, a decrease in consecutive limit-up continuation rates, and internal sector divergence; if it further enters a cooling-off period, it will manifest as a spread of loss-making effects, lagged selloff of strong stocks, a decrease in yesterday's limit-up premium, and even consecutive limit-down stocks at high positions driving a rapid contraction in short-term risk appetite.

Therefore, understanding A-share sentiment cannot remain solely at the subjective judgment level. Excellent quantitative research should attempt to transform these trading experiences into an observable, statistically verifiable, and testable indicator system. For instance, the consecutive limit-up streak can reflect the upper limit of risk premium the market is willing to offer; the consecutive limit-up continuation rate can measure the willingness of short-term capital to take over; the limit-up failure rate can reflect the degree of capital divergence; the number of limit-up stocks within a sector and its turnover ratio can measure the intensity of theme diffusion; and the number of limit-down stocks and the drawdown of high-level stocks can depict whether loss-making effects are spreading. Through these indicators, we are not simply trying to predict the market but rather determine whether the current market is in a sentiment initiation, acceleration, divergence, recovery, or cooling-off phase, thereby understanding capital behavior and risk changes more rationally.

Market sentiment is not the opposite of quantitative analysis; truly mature quantitative research transforms seemingly emotional market phenomena into observable, statistically verifiable, and structured variables.
In the next few years, I believe three trends will emerge in quantitative finance. First, AI and quantitative research will be deeply integrated, with strategy generation, intelligent backtesting, automated risk control, and real-time data monitoring becoming important directions. Second, quantitative finance will no longer be a "black box tool" for a few institutions but will become a fundamental capability in more investment decision-making processes. Third, the market demand for cross-disciplinary talent will be increasingly high. Future excellent quantitative practitioners will not only need to understand code and models but also market structure, risk management, trade execution, and business implementation.

Daily Work and Core Skills

My current daily work involves more of a connection between research, strategy, tools, risk control, and business. On one hand, I pay attention to changes in market structure and trading environment, understanding how capital behavior, risk preferences, and trading opportunities change in different market stages; on the other hand, I research how quantitative methods can be applied to real investment scenarios, considering how to integrate strategy logic, data processing, backtesting validation, trade execution, and risk control into a complete process.

In the context of rapid AI development, I also focus on the possibilities of combining AI-assisted investment research with quantitative research. AI improves efficiency but also brings new questions: Are the generated strategies reasonable? Are the backtesting results credible? Is the model output stable? Are the risk boundaries clear? All these questions require practitioners to possess underlying judgment capabilities in financial engineering and risk management.

In my opinion, the most important aspect of doing quantitative work well is not a single skill, but a systemic capability: first, market understanding, as any model ultimately needs to be validated against market mechanisms, trading behavior, and risk structures; second, model thinking, the ability to decompose variables, assumptions, signal sources, risk exposures, and validation methods; third, data and engineering capabilities, as quantitative research must ultimately be implemented in data processing, code implementation, backtesting frameworks, and trading systems; fourth, risk awareness, as every model has its boundaries, and every strategy has its failure cycle; fifth, communication skills, as complex quantitative logic needs to be clearly expressed, and abstract risks need to be accurately explained.

The Value of the CQF

I chose to study the CQF not simply to obtain a certificate, but to further enhance my understanding of financial engineering, quantitative modeling, and risk management through systematic training. My previous quantitative research and competition experiences made me focus more on strategy performance, while the CQF prompted me to think from a more fundamental perspective: where do returns come from, how risks are exposed, are model assumptions valid, and is the strategy still effective when the market environment changes?

For me, the greatest value of the CQF is that it helped me complete the upgrade from a "strategy perspective" to a "financial engineering perspective." It made me realize that quantitative finance is not just about models, code, or backtesting results, but a systematic approach to uncertainty, risk pricing, asset distribution, statistical validation, and decision discipline. Regardless of how the market changes, this underlying framework can help practitioners understand the market more rationally, rather than being swayed by short-term fluctuations or singular outcomes.

For newcomers to the field of quantitative finance, my advice is: do not rush to pursue complex models; first, build a solid foundation in finance, mathematics, statistics, and programming; attach importance to data quality, transaction costs, liquidity, and risk control in real markets; and always remain humble. Models can help us understand the market, but they cannot replace the market itself. Truly excellent quantitative practitioners do not always pursue correct predictions, but rather continuously validate, continuously correct, continuously iterate, and build a more robust decision-making framework amidst uncertainty.

Reflecting on my own experience, I have increasingly come to realize that the true value of quantitative finance is not just about pursuing higher returns, but about using more rigorous methods to understand the market, manage risks, and improve decision quality. The significance of the CQF for me lies precisely in its help in connecting my practical experience, market understanding, and financial engineering system, enabling me to more clearly understand the logic, risks, and boundaries behind models.