It’s starting to look like 2022 hit a peak at $4,766. The question is whether a model can be built to reliably predict whether if a recession or correction will happen.
Currently, the model is looking at short interest, inflation adjusted PE, sentiment data such as household sentiment on durable goods and business conditions and unemployment. There’s still lots of factors to test but the signal “looks” to be working.
However, a forecast is just that, just a prediction and it’s subject to overfitting. What’s troubling is that historical forecasts were predicting the downturn during the downturn, whereas in 2022, it seemed to be an leading indicator.
I’ll further refine the model in the upcoming week. An interesting view would be to look at how the signal does in historical recessions. Further analysis to come, stay tuned.