Robustness Check of the Quantile Model: Is the Model Sensitive to Outliers? Why It Matters An important aspect of a model’s validity is its resilience to outliers. Outliers introduce meaningless noise. A good model should not be overly sensitive to noise or to any particular data point. This is why we avoid doing overly simplistic technical analysis (“TA”) or “dot-connecting” when modeling data. In technical terms, the model should be robust to outliers. Methodology To test the Quantile Model’s robustness, we will alter the data so that the last cycle peaks at 100K instead of 67K. To make this scenario more extreme, we will keep the price at 100K for two weeks—fully pushing the model to see if it is thrown off by the outlier. Results The predictions do not change. The model does not care! It is not dependent on how any particular cycle played out because it captures the general long-term behavior, making it very resilient to any specific unexpected event.