Stocks to Short: S&P 500 Stocks Forecast Performance Evaluation Report for Short Positions

In this stock market forecast evaluation report, we will examine the performance of the forecasts generated by the I Know First AI Algorithm for short positions for S&P 500 stocks which were daily sent to our customers. Our analysis covers the time period from January 1st, 2020, to March 25th, 2020. The report also demonstrates the performance since Coronavirus started affecting global economy (February 22nd). Below, we present our key takeaways of our stock market predictions.

stocks to short
Chart 1: Average returns for S&P 500 stocks to short forecasts for time horizons spanning from 3 days to 1 month against SPY (short position) since January 1st, 2020.
stocks to short
Chart 2: Hit Ratio for S&P 500 stocks to short forecasts for time horizons spanning from 3 days to 1 month since January 1st, 2020.
Chart 3: Average returns for S&P 500 stocks to short forecasts for time horizons spanning from 3 days to 1 month against SPY (short position) since February 22nd, 2020.
Chart 4: Hit Ratio for S&P 500 stocks to short forecasts for time horizons spanning from 3 days to 1 month since February 22nd, 2020.

S&P 500 Stocks to Short Evaluation Highlights:

  • Stock market forecasts for short positions that were generated achieved positive returns and outperformed SPY (short position) on both periods analyzed.
  • Signal filtering has a positive effect on returns for 3-,7- and 14 days time horizons.
  • I Know First algorithm performed impressive accuracy achieving hit ratios higher than 60% on every time horizon and reaching up to 98% on 14 days’ time frame over coronavirus times.

The above results were obtained based on the stock market forecast evaluation over the specific time period using a consecutive filtering approach for short positions – by predictability and then by signal, to give a general overview of the forecasting capabilities of the algorithm for specific stock universe. We will start with an introduction to our stock picking and benchmarking methods and then apply it to the S&P 500 stocks universe covered by I Know First. After introducing the evaluation methodology in detail, we develop the analysis and present the evaluation results with relevant conclusions. This stock market forecast evaluation is part of our continuous studies of live I Know First’s AI predictive algorithm performance.

About the I Know First Algorithm

The I Know First self-learning algorithm analyzes, models, and predicts the capital market, including stocks, bonds, currencies, commodities and interest rates markets. The algorithm is based on Artificial Intelligence (AI) and Machine Learning (ML) and incorporates elements of Artificial Neural Networks and Genetic Algorithms.

The I Know First Market Prediction System models and predicts the flow of money between the markets. It separates the predictable information from any “random noise”. It then creates a model that projects the future trajectory of the given market in the multidimensional space of other markets.

Since the model is 100% empirical, the results are based only on factual data, thereby avoiding any biases or emotions that may accompany human derived assumptions. The human factor is only involved in building the mathematical framework and providing the initial set of inputs and outputs to the system. The algorithm produces a stock market forecast with a signal and a predictability indicator. The signal is the number in the middle of the box. The predictability is the number at the bottom of the box. At the top, a specific asset is identified. This format is consistent across all predictions.

Our algorithm provides two independent indicators for each asset – signal and predictability. The signal is the predicted strength and direction of the asset’s movement. This is measured from -inf to +inf. The predictability indicates our confidence in that result. It is a Pearson correlation coefficient between past algorithmic performance and actual market movement. This is measured from -1 to 1.

A detailed description of our heatmap can be found here.

Stock Picking Method

We take the top 30 most predictable assets for short positions, and then we apply a set of signal-based filters: top 20, 10 and 5 based on predictability.

By doing so we focus on the most predictable assets for short positions on the one hand, while capturing the ones with the highest signal on the other.

We use absolute signals since these strategies are short ones.

For example, a top 30 predictability filter with a top 10 signal filter means that on each day we take only the 30 most predictable assets, and then we pick from them the top 10 assets with the highest absolute signals.

Stocks To Short Forecasts Performance Evaluation Method

We perform a stock market forecast evaluation on the individual forecast level. As the forecasts are short positions, if the price goes down the return is positive, this way we calculate the return of each forecast we have issued for each horizon in the testing period. We then take the average of those results by forecast horizon.

For example, to evaluate the performance of our 1-month forecast, we calculate the return of each trade by using this formula for short positions:

This simulates a client shorting the asset on the day we issue our prediction and selling it exactly 1 month in the future from that day, ignoring commissions involved in the transaction.

We iterate this calculation for all trading days in the analyzed period and average the results.

Note that this stock forecast evaluation does not take a set portfolio and follow it. This is a different stock forecast evaluation method at the individual forecast level.

The Hit Ratio Method

The hit ratio helps us to identify the accuracy of our algorithm’s predictions.

Using our daily forecast asset filtering, we predict the direction of movement of different assets. Our predictions are then compared against actual movements of these assets within the same time horizon.

The hit ratio is then calculated as follows:

For instance, a 90% hit ratio for a top 30 predictability filter with a top 10 signal filter would imply that the algorithm correctly predicted the price movements of 9 out of 10 assets within this particular set of assets.

The Benchmarking Method – SPY (Short position)

In order to evaluate our algorithm’s short forecasts performance in comparison to the US market, we used the S&P 500 ETF (SPY) as a benchmark, since the SPY allow investors to bet on a decline in the S&P 500 Index.

The SPY is one of the most popular funds that aims to track the S&P 500 Index, which measures the stock performance of the largest 500 companies by market cap listed on different stock exchanges in the United States. The S&P 500 is one of the most followed equity indices and is frequently used as the best indicator for the overall performance of US public companies, and the US market as a whole.

For each time horizon, we compare the SPY (shorted) performance with the performance of our forecasts after the filtering processes described above.

Stock Universe Under Consideration – S&P 500 Stocks

In this report, we conduct testing for S&P 500 stocks that I Know First cover by its algorithmic forecast in the “S&P 500 stocks” package. The period for evaluation and testing is from January 1st, 2020 to March 25th, 2020. During this period, we were providing our clients with daily forecasts for S&P 500 stocks. The time horizons which we evaluate in this report are 4 periods spanning from 3 days to 1 month.

Stocks To Short Performance Evaluation – Since January 2020

After filtering by the top 30 most predictable assets for short positions, we applied further filtering by signal strength to investigate potential improvement. The results of the testing showed that there is a significant positive effect on the assets average return, especially in the case of the 3-days, 7-days and 14-days investment horizon. We present our findings in the following table.

stocks to short
Table 1: Average returns for S&P 500 stocks to short forecasts for time horizons spanning from 3 days to 1 month against SPY (short position) since January 1st, 2020.

As can be seen in Table 1, by applying the Top 30 predictability filter (All signals) our algorithm provides not only positive returns, but also significant out-performance over SPY (shorted). The data also demonstrates significant positive signal effect on average return. Top 5 signal group has the highest average return for the 3-, 7- and 14-days’ time horizons. The algorithm was able to outperform the benchmark on every time frame since the beginning of the year.

stocks to short
Table 2: Hit Ratio for S&P 500 stocks to short forecasts for time horizons spanning from 3 days to 1 month since January 1st, 2020.

For all time horizons and filtering methods, hit ratios were higher than 62%. We can observe the hit ratio increases as the time horizon increases. In addition, the hit ratio increases when we apply the signal filtering except for the 3 days’ time horizon. In this case, even though the hit ratio is lower for the Top 5 signals subset, the average return is higher. These results show that the algorithm is able to identify short positions consistently and is capable of focusing in the best opportunities in the market.

Stocks To Short Performance Evaluation – Since Coronavirus Crisis

In this section, we will examine the results with the same evaluation process since the Coronavirus started affecting the global economy. Therefore, in this case the evaluation period is from February 22nd to March 25th. In this period, the results showed a positive signal effect on the assets average returns for the 3-days, 7-days and 14-days investment horizon.

Table 3: Average returns for S&P 500 stocks to short forecasts for time horizons spanning from 3 days to 1 month against SPY (short position) since February 22nd, 2020.

In Table 3, after applying the Top 30 predictability filter (All signals) we can observe higher returns in this period. The algorithm achieved impressive return and show significant positive signal effect on the average returns. The highest average return was reached by the All signals group for the 1-month time horizon with a return of 41.12%. The benchmark was outperformed by all the assets groups picked by the I Know First algorithm.

Table 4: Hit Ratio for S&P 500 stocks to short forecasts for time horizons spanning from 3 days to 1 month since February 22nd, 2020.

In this period of crisis, the algorithm identified short positions with an impressive accuracy. Hit ratios were higher than 80%, for all time horizons and filtering methods. For the 14 days’ time horizon, the forecasts achieved 98% hit ratio. The algorithm showed better performance for short positions in a bearish market (Coronavirus period), nevertheless the stock picking process was very successful in both periods showing a positive effect for the signal and predictability indicators.

Conclusions

This evaluation report presents the performance of our forecasts for short positions for S&P 500 stocks picked by I Know First’s AI Algorithm during the periods from January 1st, 2020 to March 25th, 2020 and from February 22nd, the beginning of the coronavirus crisis. The results of this analysis for both periods, demonstrated a strong overall performance for forecasts that are filtered by predictability, together with significant and consistent improvement that was obtained by the signal-based filtering method.

As we are analyzing short positions, when we compare both periods, we can observe a better performance over crisis times. Although, the benchmark was outperformed on every time horizon. The I Know First algorithm showed the ability to adjust in volatile times and pick the best opportunities to invest with high accuracy and effective indicators. We look forward to new market data in the following months and will monitor the changes in performance trends that are going to be communicated to our investors and subscribers in the follow-up reports.