Creating a Quantitative Investment Strategy

1. Defining the investment thesis.
2. Data acquisition and processing.
3. Back testing the strategy.
4. Evaluate the strategy.
5. Construct the portfolio.

Defining the Investment Thesis

The initial step involves creating an investment thesis that identifies a potential market opportunity. Quantitative analysts (known as “Quants” or “Quant Analysts”) seek exploitable ideas that can generate alpha, aiming to outperform the market actively rather than passively tracking it. They explore publicly available information to uncover predictive factors that could anticipate securities prices.

Data Acquisition and Processing

Data management encompasses gathering, sorting, cleaning, processing, and analyzing data. This task requires advanced database knowledge and data analytics skills. Quant investors utilize various data types, including:

• Company Mapping: Tracking companies over time using different data vendors.
• Company Fundamentals: Involves financial statements, company demographics, etc. Third-party vendors like Capital IQ play a crucial role in collecting this data.
• Survey Data: Publicly available data about firms, markets, and economies. It may include flow of funds, analyst forecasts, and macro data like GDP.
• Unconventional Data: This unique data could hold significant value, especially if it has predictive power. It ranges from satellite imagery and web searches to physical visits to factories or stores.

Back testing the Strategy

Back testing involves testing a proposed strategy using historical data instead of real-time data. This allows us to assess how the strategy would have performed in the past. While past success doesn't guarantee future results, a strategy that performs well in back testing adds credibility to its potential effectiveness.

Information Coefficient measures the correlation between a predictive factor and subsequent returns. A higher correlation indicates a stronger predictive factor.

Earnings Yield, calculated as Earnings / Price, is often used to create a factor score.

Higher earnings yield signifies better returns for each invested dollar. Factor scores measure the deviation from the average earnings yield. When ranking these scores, similar to a z-score, analysts focus on the deviation from earnings yield. This ranking aids in assessing the predictability of stocks.

For example, a factor scores of 0.8, as shown below, may result in the following way:

$$ \begin{align*}
\text{Factor score} = & \frac { (\text{Earnings yield}_{(\text{stock i})} – \text{Average Earnings Yield}) }{ \text{Standard Deviation of earnings yields across stocks} } \\
& \frac {(7\%-3\%) }{ 5\% } = 0.80 \end{align*} $$

$$ \begin{array}{c|c|c|c|c}
\textbf{Stock} & \textbf{Factor} & \textbf{Subsequent Month} & \textbf{Rank of Factor} & \textbf{Rank of} \\
& \textbf{Score} & \textbf{Return (%) } & \textbf{Score} & \textbf{Return} \\ \hline
I & 0.80 & 4.01\% & 2 & 2 \\ \hline
H & 1.16 & 4.09\% & 1 & 1
\end{array} $$

The Pearson Information Coefficient (IC) is a correlation coefficient that measures the relationship between a factor's value in the current period and the subsequent period's stock returns. It ranges from -1 to +1, with higher values indicating stronger predictive power for future returns.

For US stocks, a monthly IC of around 5%–6% is generally considered strong. It's important to note that outliers can influence the coefficient. The Spearman rank IC is a variation that offers more versatility. It involves calculating correlation coefficients between ranked factor scores and ranked forward returns, similar to the Pearson IC.

Evaluating the Strategy

The next step involves critically assessing the strategy's performance using various metrics. This evaluation considers such factors as risk, returns, drawdowns, and other relevant performance indicators. It ensures that the strategy aligns with the desired investment goals and risk tolerance.

Portfolio Construction

When a strategy is prepared for live trading, it's integrated into a portfolio along with other strategies. Key portfolio construction considerations are even more important at this stage:

• Risk Models: Analyze correlations among strategies and assets within the portfolio. External risk model providers might be employed to implement these risk assessments.
• Trading Costs: Account for both implicit and explicit trading costs. High trading costs can undermine the viability of a profitable strategy.

Pitfalls in Quantitative Investing

• Survivorship bias: This involves excluding companies that no longer exist from analysis, which can skew results.
• Look-ahead bias: Using data that wasn't available during the period being analyzed, leading to inaccurate outcomes.
• Data mining/overfitting: Continuously searching for data that fits the model can result in over-optimization.
• Turnover: Constraints on trading frequency can limit the strategy's effectiveness.
• Hard-to-borrow stocks: Refers to expensive or unavailable stocks for short selling.
• Transaction costs: Unmanaged costs can undermine an otherwise profitable strategy.

Question

Which of the following is least likely an example of unconventional data Quant investors use?

1. Statement of cash flows.
2. Satellite data tracking shipping volume.
3. Web-site traffic.

Solution

The correct answer is A.

The statement of cash flows is a traditional financial statement that provides information about a company's cash inflows and outflows over a specific period. While it's an essential source of financial information, it is considered conventional and widely used in financial analysis. Quant investors may incorporate this data into their models, but it is not unconventional.

B is incorrect. This is an example of unconventional data. Quant investors can use satellite data to track shipping volumes, which can provide insights into economic activity, supply chain trends, and potential investment opportunities.

C is incorrect. Web-site traffic data is also unconventional and used by Quant investors. Analyzing web-site traffic can reveal consumer behavior, trends, and the popularity of certain products or services, which can be valuable for investment analysis.

 

Reading 25: Active Equity Investing: Strategies

Los 25 (h) Describe how quantitative active investment strategies are created