Pros and cons of different forecasting methods

ARIMA

Pros:

Cons:

• Assumes normality
• Fails to capture seasonality (you might have to resort to SARIMA)

Harmonic regression

Pros:

• You can incorporate multiple seasonalities
• You can add covariates
  Cons:
• Assumes that the seasonal shape is fixed year to year
• Fixed amplitude
• Assumes normality

Holt-Winters

Pros:
Cons:

Prophet

Pros:
Cons:

• Assumes normality