Excellent work - a couple of quick questions if I may:

1. Are the median and mean projections of pre-2020 peaks shown in the first chart drawn from curves of the same thing? I.e. are they all CO + NGL, or similar? I think this is addressed in the forecast performance section, but it isn't clear (to me at least - apologies if I have missed an obvious comment)

2. How does the statistical significance tests from the 'Are we there yet' section look if one uses developed world GDP growth rather than population growth? Without wishing to defend the CERA models, the obvious rebuttal of the two statistical outliers in the M0 example is that world growth has been falling recently.

Obviously, this last point gets into a deeper topic of whether GDP growth drives oil demand, or whether oil supply drives GDP growth - but it would be interesting to see the analysis nonetheless.

Great stuff though.

1. Yes, it is only including forecasts for CO+NGL.

2. Looking stricly at the significance levels, there are nothing unusual in the current production drop. Lower significance levels were observed during past recessions. Having said that, I'm just looking at volume supplied in isolation, other market indicators, such as oil prices, have been significantly high.

There is clearly a positive feedback between economic activity and oil demand and a negative feedback between oil prices and economic activity.