A Summary Of Forecasts Vs. Actuals

Resolved Forecasts:

January 2020 — When the model from China said COVID-19 would peak at 10,000 cases worldwide in February, I accurately foretasted 10,000 cases was far too low by orders of magnitude, that the Coronavirus could infect millions.


February 2020 — My model accurately projected we would hit 100,000 at the end of February. We crossed 100,000 within a day of the forecast.

Economic Forecast:

On February 10, at a time with DOW was over 29,000, I wrote, "My view is the market may not be adequately factoring in the significance of the Chinese economy to a wide range of businesses." Within a couple of weeks, the stock market executed a significant correction.

"Even though the stock market wasn’t showing it at the time, my forecast was projecting the Coronavirus was about to have significant financial impacts. China’s factory closing would affect supply chains and pricing. Commodity based economies like Australia would be hurt. Oil producing countries such as Iran, Saudia Arabia and Russia would be put under increased pressure as their revenues fell, and this could potentially have geopolitical implications. The US and other service-based economies would see significant declines in the first half of 2020. I noted that China’s decrease in travel would affect many tourist economies. I argued that we should expect major European economies like Germany, who were already on the brink of recession, to slide into recession in 2020. (Read the full story in Research World.)


March 2020 — I showed by COVID-19 would become the #1 Killer in the US, and the dire implications for NYC. COVID-19 become the #1 Killer for three weeks of April, and, as predicted, was horrific for NYC in particular.


April 2020 — When the US had less than 600,000 confirmed cases, I accurately foretasted we'd end April with 1,036,543 cases — a level we crossed on the last day of the month, ending slightly higher than foretasted.

The analysis showed Stay Home saved over 25,000 lives due to cases avoided.


To Be Determined:

November 2020, Forecast Briggs-3 Headline: Vaccine Ends Pandemic In Q3 2020.

You can read about the details of Briggs-3 in this Research World article.

September 2020, Forecast Briggs-2 Headline: Forecast raised from 15-20% infected and approximately 300,000 COVID-19 deaths in the US by end of the year to 20-25% infected and approximately 350,000 deaths (range: 330k to 390k).

You can hear about this forecast and my approach overall in this 12min ARF presentation.

Mid-March 2020, Forecast Briggs-1.1 Headline: Downgraded the Infection Fatality Rate (IFR) to 0.4% (range of 0.4 to 0.6) which lowers 2020 projection of deaths from 780k to approximately 300,000 dead by end of 2020

The lower IFR was presented at the ARF Town hall presentation, and is based on my Diamond Princess Cruise Analysis, which was completed after the 1.0 Forecast. At the time, with projections showing cases surging, and New York City likely to become overwhelmed, and the potential of other US regions to become overwhelmed, I noted a high level of uncertainty as to whether IFR would increase if hospitals became overwhelmed. With fewer than 200 deaths in the US, I was projecting more than three orders of magnitude increase in under ten months.

March 1 Forecast 1.0: 780,000 deaths by end of 2020 — On the first of the month, when there was 71 confirmed cases in the US, I released Forecast 1.0 where I evaluated "How Bad Could It Get?" I projected COVID-19 would be far worse than the Seasonal Flu, and would likely infect 15-20% of Americans, and with an IFR of 1.7%, could kill 780,000 by the end of 2020. I offered several scenarios of how we could avoid this level of detestation, and noted the uncertainty in the IFR, given limited closed-population analysis.

This Forecast of the overall infection rate was driven by a variety of factors including the way in which people would take precautions as the risk of infections increased (keeping it from infecting everyone it could potentially infect) and the lack of coordinated global and state level lock-downs, and data from China on IFR, which I lowered based on the early trending data I was collected from age weighted data from Diamond Princess Cruise fatalities.

In The ARF Town Hall, I explained that a key driver of the model was the unlikelihood that the US would be able to coordinate a global lock-down let alone a coordinate US lock-down. The US, I pointed out, had allowed the case count to pass 3,000 vs. around 500 when China implemented lock-downs.

Second Wave: In April, my analysis showed Stay Home saved over 25,000 lives due to cases avoided (measured with social distancing), but warned of a second wave. My analysis showed warmer weather/more sunlight would help slow the spread, but it could also lull the US into a false sense of risk such that in the Fall, with increasing case count, the US would be less likely to lock-down the population to stop the spread.

Herd Immunity Implausible in 2020: I estimated the US had less than 5% infected by end of April (likely closer to 2% actually infected nationwide), and therefore herd immunity would not be plausible in 2020 because it would require getting to around 70% infected — that would require repeating March & April more than 10 times over.

Economic Forecast:

In February and March I forecast that unemployment would be severe and recovery would not follow the V-shape Goldman Sachs, BofA and others initially expected in early March.

In April, Based on ZIPCODE analysis of new unemployment claims, I further projected the broad-based and skew toward middle and upper-incomes, which was later borne out in the government Jobs Report. It remains to be seen if the economy follows the four Phases I explained in my analysis, or follows the V-shape that others have projected.

Recent Analysis