I think 20% is balanced approach. Since HBD is pegged to USD, 20% would mitigate the loss of purchasing power in USD. I believe current USD inflation in terms of losing purchasing power is about 15%. Have HBD APR slightly above it, makes sense to me. But wouldn't go higher than the double of USD inflation rate.
At this time, the limit I would put would be 30%. Anything above that will get risky. Anything below is manageable. These risks can even be lowered by introducing longer lockup time as suggested by Taskmasters HBD Bonds idea.
What data are you using to suggest that USD is losing purchasing power at a rate of 15% per year?
US CPI is only 7.90% (exceptionally high at the moment, but way below 15%). Core inflation is 6.40%. Even producer price inflation is only 10%.
CPI is not a good indicator for overall inflation. It doesn't take into account everything. Now if you double it, it may represent more of a reality. :)
Better indicator of inflation was M2, which Fed discontinued after the massive printing during pandemic. Even M1 is, I believe around 13-14%.
More over, 15% is my personal observation.
M1 and M2 measure supply. This is tangentially related to price inflation/purchasing power and is absolutely not a superior metric to those that actually measure prices directly, regardless of the particular limitations of those metrics. It would be akin to saying that the change in supply of Hive from yesterday to today was a valid metric to measure the change in Hive price.