If I recall correctly, bitcoins from a coinbase (mining reward) transaction can only be transferred after 100 blocks confirmation (ie about 16.5 hours). I'd assume that they might keep a bit of a speculative long position, but otherwise liquidate - they have bills to pay.

As far as I know only 6 confirmations are required for a transaction to be considered practically irreversible. Is it different for coinbase (reward) transactions?

Yeah, it's different for coinbase transactions because those transactions and the reward from them vanishes if that block is ever reorganized out of the main chain, whereas ordinary transactions will generally be included in the new chain too. It's as much a protection against technical issues as it is malicious attacks. I think there have been a small handful of reorganisations that are larger than 6 blocks in practice, and this has generally held up well.

That’s my understanding.

The altcoin crowd insists that people with a lot of bitcoins can and do "diversify" their cryptocurrency holdings by exchanging them for altcoins, thus making bitcoin somewhat of a "reserve" cryptocurrency.

This would also give those mining altcoins a nice market to convert their coins to btc and subsequently cash out.

I guess the question is, if bitcoin drops, will everything drop? Or will spreading your stake around the altcoin minefiled be enough to protect against huge losses?

Large shops pretty much have to, either to pay for new ASICs and GPUs in fiat currency, or return capital to existing shareholders.

(not joking) - Do the hardware Bitcoin mining makers not accept Bitcoin?

Why would they with the insane transaction fees, instability of value and slow transaction time? It’s of very limited use as a currency.

Miners and mining hardware makers don’t actually care for bitcoin from a philosophical point of view, so they have no incentive to use it as a currency. It’s literally just a means to an end for creating value to them. If they were growing tulips instead, you’d see them investing in farming techniques just as happily.

I think bitmain (a large asic maker) only accepts bitcoin cash for their miners, or at least their latest version.

Yeah, mainly for political reasons - they were one of the main supporters of Bitcoin Cash and it re-enabled the secret mining optimization they'd promised they weren't using, honest, which was accidentally broken by Segwit. At the time they started requiring it as the sole method of payment it was basically unusable with block times measured in hours, little in the way of exchange and other support, and I think the exchanges may still have required so many confirmations it took days to get funds onto them. They could get away with this because they had no real competitors.

There were some rather unamused comments at the time from miners who got their first experience of Bitcoin Cash by having to use it to buy from Bitmain, questioning whether it was always this terrible given that all its very vocal boosters had been claiming it was the fast option compared to Bitcoin. (The usual response was to argue that the block speed varied between a minute per block and the hour+ one and they were just unlucky - which technically it did, it's just most of the time it was in slow mode and the brief fast intervals were days apart.)

This is correct. It's also important to note that bitmain is one of the main responsible parties for the creation of Bitcoin Cash and this move is mostly a way to boost it's value, because even though Bitcoin fees are high, they aren't that bad ($3 to $30 depending on the load) when you're selling devices for $2K+.

This is a good reason for these miners to short the bitcoin futures contract

If price shoots up, lots of new miners join, difficulty shoots up, they can no longer produce enough bitcoin to cover their shorts, they lose a ton of money on the shorts.

Commodity futures only work when you produce a consistent amount.

There's a worst-case for the difficulty adjustment, right? Just hedge your expected worst-case production through shorting the future, then.

Not really. If they release a new miner that's an order of magnitude more efficient the difficulty will shoot up as soon as enough units are online.

Hmm. I thought I remembered reading something about two weeks, but that's the time period required without enough blocks to make difficulty fall.

It adjusts every 2016 blocks, but if new miners come online that could in theory be mined in a day instead of two weeks.

I haven't looked too closely at how difficulty adjusted after previous order of magnitude efficiency gains, but the risk is certainly there.

they are cash settled Bitcoin futures not settled in BTC

That doesn't change the economics. Either way they have short exposure to BTC and no guarantee they will be able to produce BTC, which could hurt them if market moves against them too quickly.