The Hourly Average is based on a couple of days average, rather than a longer term benchmark. It would be useful if we could choose the timeframe in which the hourly average is calculated e.g. weekly hourly average, monthly hourly average, quarterly hourly average, annual hourly average, so that the benchmark is based on a bigger data sample than just a couple of days. Limitation with a couple of days is it could be disrupted too much in a highly volitile dataset