Before we dive in, let's set the stage.
Our trading system is built on a foundation of two powerful machine learning models (you can learn more about our machine learning market forecasts here: https://andywallsq.com/pages/machine-learning-market-forecast).
- Model 1: Busted Tree Classifier: This algorithm predicts the probability of the next hour's close being up or down, giving us a percentage probability.
- Model 2: Booster Tree Regressor (Price-Based): This model determines the price trend (up or down). Using conformal predictions, it establishes a confidence range based on its algorithm.
When both models align in the same direction (either up or down), we assign a "High Confidence" score. This typically happens about once a day, and these trades have a historically high win-to-loss ratio. (For a deeper understanding of High Confidence predictions, see: https://andywallsq.com/blogs/news/understanding-high-confidence-predictions-in-our-system).
This leads to the question we often get:
"Should I only use the High Confidence model?"
Our usual answer is: "It depends."
Let's break down the data and statistics to help you decide which model might be right for you at any given time.
All Trades Strategy (Risk: $400)
(Reference Image 1 and 2)
As you can see, this model comes with a significant drawdown of $4400. This could easily wipe out a smaller funded account. You'd likely need an account with at least $5000 to weather that kind of storm. However, this strategy produced a return of $8100 over 10 weeks, after recovering from the drawdown. The win rate for this strategy in this period was 54%, with a total of 134 trades.
High Confidence Strategy (Risk: $400)
(Reference Image 3 and 4)
This strategy uses the same parameters as the All Trades Strategy (Images 1 and 2), but we've selected the "High Confidence" filter. This means the system only enters trades when the probability parameters, recommended time window, and High Confidence criteria are met.
The big difference? While the gains are lower at $6300, the drawdown is significantly reduced to just $825. As mentioned, High Confidence trades occur less frequently, resulting in only 41 trades over the 10-week period. This strategy had a higher win rate of 65%.
High Confidence Strategy (Risk: $600)
(Reference Image 5 and 6)
The higher win rate of the High Confidence strategy allows us to take on more risk. By increasing the risk to $600 per trade, we achieve the same $8100 profit as the All Trades strategy, but with a more manageable drawdown of $1200. The win rate for this strategy during this period was 64%.
The Importance of Timing
Timing is absolutely crucial when it comes to trading strategies. (You can explore our market predictions dashboard for a deeper dive into timing and potential: https://andywallsq.com/blogs/news/unlock-your-trading-potential-market-predictions-dashboard-es-501)
(Reference Image 1)
Notice in Image 1, there are three consecutive losing weeks. These losses contribute to the drawdown, erasing previous gains. However, the strategy then rebounds strongly. It's not a coincidence that this period overlaps with high market volatility due to trade wars, tariff announcements, etc.
Now, what if you started your funded account near the end of February?
(Reference Image 6 and 7 )
As shown in Images 6 and 7, the All Trades Strategy performs exceptionally well during this period, generating nearly $8000 in profit with a drawdown of around $1600. What's significant is that it only takes 5 weeks to reach that $8000, half the time of the previous scenario.
Given the massive potential reward versus the cost of a funded account, and considering the use of our AI/ML model strategy templates (https://andywallsq.com/blogs/news/strategy-templates-for-ai-ml-model), it's reasonable to assume that with this strategy and timing, some accounts will blow up. However, by sequentially activating new accounts every 1-2 weeks, some of them will coincide with those periods of exponential gains and hit the payout target.
Should You Use Only One Method? High Confidence or All Trades Exclusively?
My personal approach is to use a mix of both.
Approach 1: You could have one account (Account A) with a specific risk setting for the All Trades strategy, and another account (Account B) with a higher risk setting for the High Confidence strategy.
Approach 2: Another option is to use one account (Account A) with a certain risk for All Trades, and the same account (Account A) with the same risk for the High Confidence strategy. This way, when a High Confidence trade occurs, you're effectively doubling your bet.
Ultimately, the best strategy depends on your risk tolerance, capital, and market conditions. By understanding the statistics and potential drawdowns of each approach, you can make more informed trading decisions.
Try It Yourself!
We encourage you to explore these strategies further and make your own conclusions. To help you in your analysis, we invite you to use our Free End of Day Subscription to backtest the data. Visit https://andywallsq.com/pages/plans-pricing to get started and unlock your trading potential!