Forecast Revision
Updating our model, the total number of US deaths from Covid-19 will most likely be closer to the high end of our initial forecast range (~80,000) than it will be to our initial base case (~46,000).
Our main error appears to have been underestimating the Mortality Rate. We had incorrectly assumed that rates would stabilize after a brief period of increasing, but have continued to push higher. This is not just in the US, but has occurred in most countries including those used in our model to forecast the US.
Interestingly, though our model’s figures are pointing higher, real world mortality rates are showing themselves to be lower than consensus expectations. According to various recent reports, the percent of the population that has been exposed to Covid-19 is much higher than consensus expectations, making the real world mortality rate much lower. The Mortality Rate that we refer to in our model is based off of confirmed cases and not estimated population exposure.
Our forecasts are based on deaths from Covid-19 that have been lab-confirmed and excludes repatriated cases. The data is consolidated from the state-level up. These figures seem to be slightly lower, by around 10%, in comparison to others that are at times quoted in the news. As of April 21st, the number of lab-confirmed Covid-19 deaths in the US was 40,040.
As long as another wave of Covid-19 cases does not surge, we do not believe marginal changes in models will significantly impact markets at this stage. Our initial forecasts were important at the time as they came in well under the 100,000 mark while others were forecasting multiples of this and some were still talking in millions.
Instead, indications we are seeing is that the focus has shifted towards practical measures of re-opening economies and the potential depth / permanency of the recession.