covid19-supplier-recovery 0.0.3

Modeling the recovery from covid19 crisis for suppliers to industries that have been severely impacted.

covid-19-supplier-economic-recovery

DISCLAIMER

We are not economists. We are not epidemiologists. All core assumptions that are related to each of these fields are taken from experts. This is not an academic paper and has not gone through a peer review process. As a pandemic in modern times is a unique event, it is not possible to model resulting economics events with any degree of certainty. The best we can likely hope for is to maintain multiple plausible scenarios that can be used to inform high level strategic planning.

Any examples provided in this paper or documentation are NOT to be taken as ready for use in any particular use case.

Install

# ALWAYS create a virtualenv first

# then pip install
pip install covid19-supplier-recovery

Step 1: Read the paper

It's here

Read it and understand it. It does not make sense to use this without understanding the assumptions and structure.

Step 2: See the API usage

Say that you have a 10 week sales forecast in which you will sell 100 euros each week. You would express this forecast using a simple python list:

forecast = [100] * 10

Assuming a single 3-week lockdown starting on the second week:

lockdowns = [
    {
        'start': 1,
        'length': 3,
        'immediate_loss': 0.9,
        'client_deaths': generic_linear_func(
            x_intercept=0.05, slope=0.05),
        'lockdown_growth': generic_linear_func(
            x_intercept=0.0, slope=0.01)
    }
]

And the corresponding recovery:

recoveries = [
    {
        'client_recovery': generic_linear_func(
            x_intercept=0.1, slope=0.01),
        'market_appetite': generic_linear_func(0.2, 0.02)
    }
]

You would model your forecast like so:

rm = RecoveryModel(
    normal_forecast=forecast,
    lockdowns=lockdowns,
    recoveries=recoveries
)

adjusted_forecasts = rm.model_recovery()

And your adjusted forecasts should look like so:

[
    # start business as usual
    100.0,
    # start lockdown
    10.0,
    11.0,
    12.0,
    # start recovery
    30.0,   # .12 + ( (.85 + .10) * .20 )
    32.0,   # .12 + ( (.85 + .11) * .22 )
    34.0,   # .12 + ( (.85 + .12) * .24 )
    36.0,   # .12 + ( (.85 + .13) * .26 )
    38.0,   # .12 + ( (.85 + .14) * .28 )
    41.0,   # forecast end
]

Step 3: Make your own forecast

This part will be hard. Selecting the parameters for the model is insanely difficult and requires a TON of industry knowledge. You should probably maintain several different scenarios and update them as more data, policy, and other information becomes available.