Depending on what you're optimizing for, that might be rational. The elderly will in large parts not be alive for long with or without covid. On the other side you have a group of people where you can save younger people's lives that have a much higher downside in terms of shortened life expectancy.
I.e. if you have an equal number of 40 and 80 year olds, you might expect them to die in 40 or 5 years respectively. So keeping the first group alive has an 8x multiplier.
Additionally if you consider quality of life concerns, that might tilt everything even more in favor of saving the young.
Also the young probably have more social contact, meaning that early vaccination there saves other lives by containing the spread.
It's not as cut and dry as one group having a higher risk.
The hypothesis being that young people were mostly responsible for the spread. Even though they weren't necessarily affected by the virus as much.