Lots of things there. I wanted to make a big post about HBD but I have an important deadline Sunday so it will have to wait for Monday.
First of all, I want to start by saying that cycling in and out of haircut is not the only way to get more than 10% inflation. It is counter-intuitive but showing an example would take too much time so i'll just do it Monday in the post.
The short intuition is: we buy and burn HBD when the price of hive is high, and then print Hive when the price of Hive is low, so even without haircut, if the price difference of hive at the highs and lows is large enough, this can lead to massive inflation. I have been repeating this many times in this comment thread and it wasn't getting through to you so maybe an actual example with numbers will be more useful Monday.
In your example, if half the HBD supply were converted during haircut, that would only be 5% inflation.
Yes, but then there is the normal 10% from HBD that is always possible. That's why I said 15%.
If the price falls and half is converted again, that would be another 5%
Yes, so in this case the maximum becomes 20% because there is again, the 10% that HBD allows without haircut.
On the other hand, if some were converted before reaching 0.50 backing, which is likely (and historically consistent, as often conversions start happening before reaching even 1.00 backing), then the total would be less, and it wouldn't add another 5% each time.
Of course, however this would happen a lot more consistently than a 50% crash. Let's say HBD trades multiple times between $1 and $0.9, you get 10% every time. This can happen multiple times in a day. So the inflation is always infinite.
Do you know where I could find the historical inflation data? I found a few posts but nothing very helpful. One was based on coinmarketcap which I don't trust and another had a graph but no source for the actual numbers. (I estimated about 18% peak rate from the graph November 2018 to November 2019, but this is very rough.)
Sadly, not. It was part of my post to ask 
But even we didn't get 20%+ inflation on Steem, the risk is definitely there.
BTW, the above situation (dropping 50% in value adding 5% inflation) is exactly the leverage effect. As the price goes down, your holdings are diluted a bit more. As the price goes up, your holdings (share of supply) increase through deflation.
This also illustrates why a 10% cap and a 1% cap, as we were discussing earlier, are NOT the same.
Absolutely, in this case, you are right. However, as I mentioned, the haircut is not the only way to generate massive inflation. The haircut is the only way to generate infinite inflation but the market timing issue can also generate massive inflation. But, again, the haircut is useful here, because it constrains the max supply for HBD. For instance, if we start printing massive amount of HBD when hive is at 10$ (blockchain printing resulting from post rewards, not conversions by arbitrageurs), These are extremely expensive HBD, and the only thing saving us is the fact that we hit a maximum limit on printing.
This is exactly why I am now a huge fan of HF25 and the hbdstabilizer because they allow us to print all the HBD very early in the bull run when the cost in hive is still low. They bring us closer to the haircut limit and then when hive pumps, less HBD will be printed at an expensive price.
HOWEVER, the annoying thing is that as hive pumps, the marketcap increases exponentially, therefore the HBD max supply increases exponentially too. So we are never safe from expensive printing at the top with any haircut ratio. Whether it's 50% or 1%, in the end hive will pump and make room for more HBD, and market timing is more important than total supply or percentage of HBD. This is very complex to work through, as I always say HBD hurts the brain. I will be writing a big post giving the full intuition HBD which is critical for all witnesses to understand it as it relates to matters of monetary policy if we do not make changes to the way HBD works.
Yes, with a lower haircut ratio, there is less room to do expensive printing. But again, the price swings in crypto are so huge that the price swings will be bad no matter what.
The reason I said 1% haircut ratio might be worse is because it gives very little wiggle room for HBD supply, making it very prone to bee off its peg or in haircut mode. It will not work as a stablecoin effectively and will also lead more often to the infinite inflation case because of market volatility.
This brings up an important point. Calculating HBD inflation/deflation depends partly on the parameters of the system, but the main inputs into that function are
1- future price action
2- market liquidity
3- timing of conversions (or printing)
In most cases everything is fine, but when you get bubbles it can lead to high inflation. And sadly, the typical chart pattern in crypto is a bubble. One could argue the probability is even 100% of a bubble happening. Look at all the altcoin charts one by one. Regardless of fundamentals every single altcoin goes through bubbles. Of course it could stop now, but it doesn't seem like it has yet. The very price action of HBD and SBD at the moment prove that.
Also, as I noted earlier, the haircut means that the debt is self extinguishing, so the effect on your holdings is actually a bit less than true debt leverage at the same ratio (equity doesn't ever get wiped out, only progressively diluted, regardless of the drop).
Yes, you're right. I am not worried about HBD directly causing a hive blowup. I am worried about HBD causing very high inflation, such as 50% in a year, which is definitely achievable and leads to very weird things on hive.
Hive is very complicated. By doing that it affects governance and resource credits allocation. What happens to Splinterlands? They are powering up to "own" a percentage of blockchain resources. They expect inflation, but how much? 50% in a single year?
Similarly, what happens to anyone willing to participate in governance? they just get diluted massively.
The only way HBD could kill hive or lead to a blowup as I was referring too, is if market participants like me realize what HBD is doing and begin to sell in advance. Honestly hive is a toxic asset to hold in a bear market because of these tokenomics. And since we all like powering up hive and locking it for 3 months, we are only hurting ourselves. We are the ultimate hodlers in those downtrends.
One question that remains open is how much will the inflation be mitigated by the fact that we are burning at cheap prices right now, given the conversions and hbdproposal.
If we keep burning hive at these cheap prices, we might never have a problem.
Through all my comments you already got a big percentage of the post I'll write so there you go.
A solution for HBD (which I was going to write about in the post too):
Use HF26 to convert all HBD into HIVE instantly. Put all those hive in a locked contract. This is now collateral for HBD. Issue the corresponding amount of HBD (which is the original amount).
This will cause a one time inflation spike of 10% on hive, but it's much better than the current system we have, which risks of leading to a lot of inflation.
Once we have that, every HBD printed must have the equivalent hive deposited in the collateral. Every HBD destroyed allow people to withdraw their hive from the collateral. Maker DAO style.
Price fluctuations in hive lead to liquidations of the collateral, no more infinite inflation.
RE: Almost 10 million HIVE withdrawn from the exchanges in just one week!