I Know First Evaluation Report for Top ETFs Forecast Package

Executive Summary

In this forecast evaluation report, we will examine the performance of the forecasts generated by the I Know First AI Algorithm for short-term and long-term ETFs which is sent to our customers on a daily basis. Our analysis covers the time period from 1 January 2019 to 19 August 2019. We will start with an introduction to our asset picking and benchmarking methods and then apply it to the ETFs universe as covered by the I Know First’s “Top ETFs Forecast” package. We will then compare returns based on our algorithm with the S&P500 performance over the same period. Below we present our key takeaways: 

ETFs Highlights:

  • The best returns were obtained in the 1 month time horizon, with the best performance being a return of 2.69% for the top 5 predictability and signal filtered assets during this time-horizon.
  • The top 5 predictability and signal filtered assets beat the benchmark in all time horizons, with the best outperformance being for the 2 weeks time horizon, where the top 5 assets beat the benchmark by over 87%.

Note that the above results were obtained as a result of evaluation conducted over the specific time period to give a general presentation of the forecast performance patterns for ETFs. The following report provides extensive explanation on our methodology and detailed analysis of the performance metrics that we obtained during the evaluation. This report is a new I Know First evaluation series illustrating the ability to provide successful long term and flexible forecasting for ETFs.

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 system outputs the predicted trend as a number, positive or negative, along with a wave chart that predicts how the waves will overlap the trend. This helps the trader to decide which direction to trade, at what point to enter the trade, and when to exit. 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 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 movement of the asset. 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.

You can find a detailed description of our heatmap here.

The Asset Picking Method

The method in this evaluation is as follows:

To fully utilise information provided by our forecast, we filter out the top X most predictable assets and rank them according to their predictability value. Thereafter, from them, we pick the top Y highest signals and re-adjust the rankings accordingly.

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

For example, a top 20 predictability filter with a top 10 signal filter means that on each day we take only the 20 most predictable assets from our asset universe, and then we pick from them the top 10 assets with the highest absolute signals. On the other hand, a top 20 predictability filter with a top 20 signal filter would imply that we are solely filtering based on predictability, since we are selecting all assets in this particular set which have already been filtered by predictability.

We use absolute signals since these strategies are long and short ones. If the signal is positive, then we buy assets, i.e. open long position and, if negative, we open short position on such asset. This is to help us to identify the assets with the maximum magnitude of change, which is indiscriminate as to whether one adopts a short or long position. 

The Performance Evaluation Method

We perform evaluations on the individual forecast level. This means that 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 based on our positions on different assets and forecast horizon.

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

This simulates a client purchasing the asset on the day we issue our prediction and selling it exactly 1 month in the future from that day.

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

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

The Hit Ratio Calculation

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

Using our asset filtering method based on predictability and signal, 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 20 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

The S&P 500 index is used as benchmark. Stocks should be bought (or shorted) when they have been identified to have high signal strength and high predictability. We compare our rate of return based on purchasing (or shorting) the top X assets after applying both the predictability and signal filters with the rate of return of the S&P 500 index in the same time horizon, from 1 January 2019 to 19 August 2019 in a long position. This helps us to determine the effectiveness of our algorithm by comparing the rate of return of the benchmark against the average investor returns.

Asset Universe Under Consideration: ETFs

In this report, we conduct testing for ETFs that I Know First covers in its algorithmic forecast in the Top ETFs Forecast package.

Performance Evaluation: Evaluating the Predictability Indicator

We conduct our research for the period from 1 January 2019 to 19 August 2019. Following the methodology as described in the previous sections, we start our analysis by computing the performance of the algorithm’s long and short signals for time horizons ranging from 3 days to 1 month, without first considering the signal indicator. We apply filtering by the predictability indicator to investigate its sole marginal contribution in terms of return, and observe how returns change as these filter is applied. Afterwards, we calculate the returns for the same time horizons for the S&P500 using the ETFs universe and compare it against the performance of the filtered sets of assets. 

In our evaluation of the predictability indicator, the benchmark that is used is our internal benchmark, which is the returns we would have if we did not filter ETFs by predictability, and merely purchased an equal amount of each of the ETFs in the ETFs universe. This is to determine the usefulness of filtering our results by our predictability indicator.

Our findings are summarized in the table below:

Table 1. Predictability Effect On Return for ETFs

From the above table, we can observe that generally, the top 20 assets underperformed the benchmark for all short-term and medium-term time horizons. We also note that the best performance for predictability-filtered assets is recorded for the 1 month horizon, which provided returns of 0.53%. 

Overall, by comparing between the rate of returns for the top 20 predictability assets and the benchmark, this suggests that it would be more valuable to filter out and invest into the top predictable assets for potentially greater returns, rather than indiscriminately invest into all ETFs in the ETFs universe. 

Based on the above analysis, we continue our study to identify whether the results could be improved when we filter by signal indicator.

Performance Evaluation: Evaluating the Signal Indicator

In this section, we will demonstrate how adding the signal indicator to our asset picking method improves the above performance even further. It is important to measure the performance of our strategy with respect to the benchmark, and for that we will apply the formula:

We further filtered the assets based on signal strength, which were previously already filtered by predictability. The results of the testing showed that there is a significant positive marginal effect on the assets’ return, especially in the case of the 3 days, 1 week and 1 month time horizon. We present our findings in the following table and charts.

Table 2. Signal Effect On Return for ETFs
Figure 1. Rate of return for all time horizons after predictability and signal filter
Table 3. Out-performance Delta for ETFs
Figure 2. Out-performance delta for all time horizons after predictability and signal filter

The data that is presented is an analysis of assets from the ETFs universe which have previously been filtered by predictability. These assets have already been ranked by predictability, and the top 10 and 5 assets are selected from the top 20 predictability-ranked assets after ranking all assets by signal. In doing so, we will be able to ascertain how investors can select and filter ETFs based on forecasts generated by the I Know First algorithm.

From the above set of charts, we can clearly see that in general, if we apply signal strength filtering to the ETFs universe, the best performance can be observed for the Top 5 signals ETFs. Most notably, these signals are more likely to produce much greater returns than the benchmark, as observed from the outperformance delta, which beat the benchmark by at least 29.03% for all analysed time horizons. 

In the short term, we see that the best performance is for the 1 week time horizon, where the filtered assets beat the benchmark by over 53%. This suggests that a stricter selection of assets from the ETFs universe, where we only choose the best assets ranked by both predictability and signal, has the potential to provide almost immediate returns.

When we look at medium term to long term performance, we observe something similar, where the top 5 predictability and signal filtered assets also produced the best results. The top 5 assets were able to produce returns that were consistently greater than the benchmark. For the 2 week time horizon, returns for the top 5 signals were more than 87% of the benchmark. For the 1 month time horizon, we also observe that returns were more than 31% of the benchmark as well. By applying the predictability and signal filters, investors can be assured of identifying assets with a higher probability of significant returns, regardless of the time horizon.

Hit ratios are important for the investor using I Know First’s proprietary AI algorithm. The investor is interested in understanding how an unadulterated portfolio would compare against one that uses the algorithm. If the hit ratio is 50%, it is merely as good as the flip of a fair coin.

Table 4. Hit Ratio for ETFs
Figure 3. Hit ratio for all time horizons after predictability and signal filter

When we consider the hit ratios for this particular set of assets, we see that in general, predictability and signal filtered ETFs were able to provide higher hit ratios, with top 10 and top 5 signals being at least 54% for all time horizons. In the case of the top 5 signals in the 1 month time horizon, we were able to obtain the best hit ratio of 65%, and also achieve the highest returns of 2.69%, which beat the benchmark by 31.86%. Given this excellent track record, this is a reflection of the legitimacy of the artificial intelligence algorithm in predicting the price movement of assets within the ETFs universe.

Conclusion

In this analysis, we demonstrated the out-performance of our forecasts for ETFs from the ETFs universe picked by I Know First’s AI Algorithm during the period from 1 January 2019 to 19 August 2019. Based on the presented observations, we record significant out-performance of Top 5 assets when our signal indicators are used as an investment criterion. As shown in the above diagram, the Top 5 assets filtered by signal yield significantly higher returns than any other asset subset on all considered time horizons spanning from 3 days to 1 month. Thus, an investor who wants to critically improve the structure of his investments into ETFs within his portfolio can do so by simultaneously utilizing the I Know First predictability and signal indicators as criteria for identifying the best performing ETFs from the ETFs universe.