In this article, we show the procedure for calculating the performance of stock portfolios based on the AI module for stock valuation, or stock scores in the StockPicking Lab application.
For trading, we always use a report with a monthly prediction of stocks rated from the most undervalued to the most overvalued. This report is generated on the first day of the month, each month. So the mentioned portfolio/strategy is being rebalanced 12 times a year.
The initial capital was $100,000. Subsequently, a rebalancing application was used to prepare an order to create a portfolio in Interactive Brokers.
For trading, we leveraged the capital by 1.5 times and then bought the 20 most undervalued stocks for 60% of the capital and sold the 20 most overvalued shares for 40% of the capital. This strategy is based on the assumption that indices have been growing for a long time period and large market declines are hampered by the short-term component (speculating on a decline in the 20 most overvalued stocks).
- The results of the mentioned portfolio are marked as AP S&P 100.
- The results of only the 20 most overvalued stocks (without leverage) are marked as AP Shorts.
- The results of only the 20 most undervalued stocks (without leverage) are marked as AP Longs.
- The results of the entire index are referred to as the S&P 500 Index.
- Chart and results are not linearized = all yields had been reinvested.
- Compound annual growth rate, or CAGR, is the mean annual growth rate of an investment over a specified period of time longer than one year.
The tables and graphs below show the final value of the portfolio. Until May of 2020, the performance registered is based on a backtest. After June of 2020, we began trading on real accounts.
Backtest from 02/2008 till 05/2020 + Real Operations from After 06/2020 till 11/2021
Only Real Operations from After 06/2020
Stock picking and the strategies based on it are successful if stocks identified as undervalued outperform the market and stocks identified as overvalued underperform the market. As can be seen from the above results, StockPicking Lab’s stock score value for each stock, meet these assumptions and exceed the benchmark, which in our case is the S&P 500 index.
Speculating purely on growth (AP Longs), users would achieve the greatest appreciation, but also greater losses in the event of poor market entry timing.
Using the AP S&P 100 strategy, which is based on speculation for both growth and decline, losses are eliminated. For example, during the last significant decline in the stock index in the spring of 2020, this strategy overcame that decline with a profit.
If you are interested in results based on factor investing, do not hesitate to try it with the free Beginner Tier of our application.