Market Crash Projection using Yield Curve Cycles

Using the 10Y and 2Y US yield curves, I wanted to attempt a potential market crash timing on the Dow Jones Index. I used all available past yield curve data and found the timings of past 3 crossovers that correlated with the market. On average [(574+455+497)/3] it took 508 days after the first 10Y-2Y crossover for the DJI to reach it's maximum value between the crossover and the bottom of the market crash that followed. Since the most recent crossover occurred on roughly Aug 26, 2019, that would put the estimate for the maximum value of the DJI to be on January 15, 2021. Following that max peak, my goal was to ultimately time the market crash, so I ran a quick logistic regression on the bottoms of each subsequent market crash from each crash's 10Y-2Y crossover. I decided on a log regression as i saw that the time that passed between of each of the past three crashes that were predicted by the yield curves was increasing, and I wanted to maintain that trend in my analysis. I found that the next BOTTOM of the market crash should be roughly 1000 days after the curve crossover, which puts the time frame around May 22, 2022 (also happens to be a Monday #BlackMonday1987). This is by no means thorough and mainly just to put a different perspective on how one could use the yield curves to forecast a financial slowdown using the past as an example.