Indian financial markets have over the years seen different cutting-edge investment products. One of the emerging areas is quant funds. Quant is a short-form of quantitative. A quant fund is an investment fund that selects stocks by utilizing the capabilities of advanced quantitative analysis. This analysis is done by using software programs-driven customized models. Think of this as a way of rule-based investing, devoid of any emotional bias. With many investment funds unable to beat benchmarks like Nifty and Sensex, there is a growing feeling among investors that quant funds could provide a worthy alternative. Mutual fund houses have been quick to spot the growing interest in quant funds among sophisticated investors. Let us have a look at how quant mutual funds (MFs) work in India.
Quant process explained
A quant fund makes investment decisions based on the use of advanced quantitative analysis. The people behind quant funds design algorithms and custom-built computer models to pick investments. The popularity of quantitative analysis within funds has risen in recent years, thanks to the rising availability of market data. Market data is a key to the entire quant process. Because data is the fuel that runs the quant process.
The quant fund model uses systematic investment strategies. Statistical models used in quant funds are developed and managed by the fund-house by leveraging data on parameters that impact price movements of underlying securities. For instance, just a human being is attracted to stocks that are cheap in terms of price/earnings (PE) valuations, quant funds select stocks for their portfolio based on quantitative decision-making frameworks. We all know and understand that machines can do repetitive tasks much more easily, quicker and more efficiently.
Even the smartest investors find it difficult to remain unemotional, while investing and at times, end up taking irrational decisions. This can greatly affect long-term returns. The quant process has become popular in generating an initial list of stocks for fund managers to begin their portfolio building work. This is because the quant process is only number-focused. A quant model has no anchoring to past winners, has no recency regency bias, has no peer pressure, has no loss aversion, has no desire to time the market, has no intention to follow the herd mentality, has no ability to react to short-term noise, etc.
Quant fund landscape in India
There are three quant funds in the country that are open to mutual fund investors. Nippon India Quant Fund (formerly Reliance Quant Fund) was launched in April 2008. It is the oldest fund. DSP Quant Fund was launched in June 2019. The third fund is Tata Quant Fund, the newest entrant to the club.
Now, let us see how these three quant funds work.
The Nippon India Quant Fund is an actively managed fund that approaches stock selection process based on a proprietary system-based model. The model shortlists 30-35 S&P BSE 200 stocks through a screening mechanism at pre-determined intervals, i.e. on quarterly basis. Stocks are selected on basis of parameters like valuation, earnings, price, momentum and quality. The investments will be made in the selected stocks on weightage defined by the fund manager. The portfolio is reviewed on a quarterly basis and changes are made based on the data generated by the model and on the discretion of the fund manager. The change in the portfolio involves both sale and purchase, both partial and complete, of the existing stocks and purchase of new stocks. As you can understand, the quant strategies are used by the human fund manager to create and maintain a portfolio.
The DSP Quant Fund is a rule-based fund constructed on good investing principles. From the universe of BSE 200 companies, the fund eliminates bad stocks, chooses stocks based on growth, quality and value and then assigns weight to 40-50 stock portfolio. The fund is based on set principles and hence reduces any human bias, while managing the fund. The fund is re-balanced semi-annually to avoid excessive transaction costs and turnover. The DSP Quant Fund does not follow a high frequency/algorithmic trading style. It is designed to become a core equity allocation fund. The fund does not incorporate price momentum based signals in its strategy. The strategy is to be fully invested, with minimum cash for managing day to day portfolio flows and liquidity. The portfolio construction is based on the model, done by the risk and quantitative analysis team at every re-balance. The role of the fund manager is limited to best execution during rebalances, limit cash drag and minimize impact costs.
The Tata Quant Fund is a new fund offering (NFO) that has an active multi-factor investment model with embedded Artificial Intelligence (AI) modules. The AI modules dynamically change factor strategies once every month based on prevailing market conditions. The fund has machine learning algorithms as its core drivers for investment strategies. The fund would invest in a portfolio of stocks selected from the BSE 200 & equity derivative list. The fund would use factor strategies like value, quality, momentum, and market-cap for rule-based stock selection and portfolio allocation. The fund portfolio would be a 30-50 stock basket. The self-learning and realigning of the model will happen every six months. The machine would take market direction calls as well. In this fund, human expertise is largely limited to design, validation, and finalization of the framework used in the investment process. Portfolio creation and regular rebalancing are driven by machines with embedded machine learning algorithms. Of course, human expertise is involved in the execution of recommendations generated by the framework. The fund model has a high average monthly turnover ratio.
When quant models don't work
Most quant models rely on historical data. This means that quant models may not work in certain situations. One, quant models may not work in the event of sentiment-driven rallies/market euphoria (not backed by fundamentals). Two, quant models may not work when market reactions are based on actual or expected changes in policy/regulation or events. Three, quant models may not fully capture ‘hope trades’ or ‘turnaround stories’ where actual historical numbers are poor, but the market is pricing in a sharp future improvement. Four, quant models that rely on artificial intelligence (AI) and machine learning (ML) face another hurdle: if the machine has not learned something in the past or has not witnessed an event in the past, it will not be able to account for the same in its prediction.
(The writer is a journalist with 14 years of experience)