Thanks for the replies, all.

To be clear, this really is a "What If" scenario more than anything else. I agree that in most cases, such a person probably wouldn't try to dump all their coins all at once. With current order books, it'd be better to do the job in small batches; even if you're worried about people noticing a huge, old coin stockpile moving on the blockchain, and trying to front-run it onto the exchanges, there's ways to cover your butt - those cash-settled BTC futures that CBOE is going to launch soon, for instance.

I think if a huge holder like this did just do a market sell, the price would recover almost instantly. That kind of thing actually happens already from time to time, when someone with deep pockets fat-fingers a trade (see, for example, June 23rd 2016 on Bitstamp, or June 22nd 2017 on GDAX). The price almost always recovers instantly, because the order books on one exchange only represent a tiny fraction of the actual demand.

But a slow sell-off could also work out okay. About 10,000 BTC are traded daily on any major exchange, so 2-3kBTC daily might fly under the radar without tripping dedicated traders' "big money is cashing out for some reason" alarms. It does mean that you have to wait a few months to get all your money, but with withdrawl limits at most major exchanges (as @dothebeats mentioned), that was probably true already.

Paper is here, including methodology. The appendix includes transaction graphs, with amounts and datestamps included.

First, a little bit of background.

In 2012, Dorit Ron and Adi Shamir published a paper analyzing the Bitcoin transaction graph, in an attempt to understand how much could be determined about the largest BTC holders, just from the publicly available info.

They concluded, among other things, that there were a dozen or so "entities" who - even in 2012 - were holding several hundred thousand BTC each:

Entity ID	Address Count	Total Incoming BTC	Transaction Count
A	78,251	2,886,650	246,012
B (Mt.Gox)	156,722	2,206,170	477,526
C	13,289	941,013	77,525
D	12,520	867,996	48,347
E	191	692,864	1,353
F	12	660,000	23
G (Instawallet)	23,649	633,606	92,593
H	9	580,000	59
I	10,561	514,066	49,550
J	4	500,021	6
K	134	479,254	1,039
L (Deepbit)	2	452,929	814,044
M	9	442,000	10
N	128	432,161	137
O	10	432,286	14
P	1	432,078	3
Q	14	430,490	23
R	2,124	321,866	300,486

---

Last night, I dreamt that I checked BitcoinCharts on my phone, only to see that someone had dumped a huge number of BTC onto the exchanges, causing the price to crash to about $2.

As you can see from the above chart, this is not an impossible situation. The whole order book of the big exchanges is "only" about 200,000 BTC deep. If one of these Entities were to rediscover a lost wallet... or if the price were to reach their "exit price" that they'd decided upon long ago... it's easy to imagine that "get all of my money out before anyone can react" would become more important for them than getting the absolute best price for it.

So my question here is... if this were to happen, what would happen next? Would the price bounce back, bolstered by confidence that this sort of holder was less likely to still be lurking around? Or would people get scared, and try to sell off their own BTC too? How would miners act? How would exchanges act? Et cetera.

Lots of people have talked technicals (what seems to be "going up", price movements, etc), so I'm gonna talk fundamentals for a bit.

BTC is still the king as far as "what can you buy with it" and "how hard is it to 51% attack", and network effects mean that those advantages are hard to overcome.

However, Ethereum seems much better at getting technical improvements pushed through. Features continue to be added, bugs continue to be fixed, etc, without much drama, even when it requires a breaking change. Bitcoin, lately, has not been so good at that (18 months between the HKC and SegWit getting enabled, with multiple forks along the way); it's harder for me to be certain that, five years from now, we will see much progress on the longstanding bugs/issues that weren't fixed by SegWit.

So long story made short? Even though I've previously held only BTC as my cryptocoin of choice, I'm considering rebalancing into a 70%-30% split based on the above.

There's a donut hole of crisis.

If the situation is catastrophic enough to put national currencies and contract enforcement in doubt, "hard money" becomes important for trading with strangers - and Bitcoin is the hardest digital currency right now. So that might end up "good for Bitcoin" in a sense.

But if the situation is catastrophic enough to put the availability of an Internet connection in doubt, Bitcoin is useless, because the network can't gossip and converge (there was a thread a while back about how we might be able to failover to HAM radio, but I doubt that's been built yet).

So for a time of crisis to be good for cryptocoins, it has to land in that narrow landing zone - where the crisis threatens confidence in the governments, but doesn't bring down the infrastructure.

I mean, I wish them the best, but... *points down at my signature*

They're already up 2x if they pulled the trigger "several months ago", so I hope they at least have the good sense to cash back out their initial buy-in ASAP instead of staying in this leveraged-over-your-dead-body state for any longer than necessary. But I'm not optimistic that this sort of quixotic enthusiasm can comprehend the words "realize your gains".

Realpolitik (PoW is a plutocracy, big money decides, etc) aside, I have problems with this kind of attitude:Do you think most users are ideologues? Do you think punishing the ones who are ideologues is worth this? Maybe regular users don't have political power, but this level of not giving a shit about them on a personal level is baffling to me.

from article 5 days ago:
(hardware in question)

I get you. To a large extent I agree, which is why I think "Segwit2x becomes another BCH" - i.e., a semi-irrelevant satellite - is one of the two likely outcomes. But let me put forward a counterpoint.

What makes Bitcoin valuable in the first place? I'd argue that the most base, most meaningful factor is that transactions are atomic and irreversible.

Okay, what makes them atomic and irreversible? It's that, in order to reverse a transaction, someone would have to have more computing power than the rest of the network put together.

Now, I don't like that miners have taken the law into their own hands on this topic. Miner incentives are not really aligned properly vis block size questions. That's part of why, when the Hong Kong Consensus called for 2MB blocks (called for them! planned to implement them!) it gated their rollout behind "broad acceptance" by the ecosystem.

But the problem is, if 90% of the hashpower really does move to Segwit2x, staying on the 1MB+ blockchain now puts my atomicity and irreversibility at risk, because there now exist many, many pools with the hashpower to unilaterally 51% attack my preferred ledger. I'm not talking about Jihan Wu here; an operation the size of Slush or BTCC, with sufficient ideological motivation, would have enough power on their own to render the 1MB+ blockchain literally unusable.

Now, if that happens, would some of the other NYA signatories switch back to mining 1MB+ to outvote that asshole? Hopefully! That's what's happened when ideological splits have happened on other coins. But now we have to worry about it, when we didn't before. Isn't that a reason for businesses to switch to Segwit2x regardless of their ideological take on how it was rolled out?

That's what makes this fork different from the others, and why I feel like there's risk here where there wasn't in past splits.

I'm hoping for this fork to be short-lived, TBH.

The way I see it, either Segwit2x doesn't attract the hashpower we think, lands with a whimper, and becomes another BitcoinCash... or the 1MB(+witness) chain gets abandoned by miners, services follow, and eventually the bitcoin.org client adopts the size bump out of practical necessity.

So I'll wait to see which it is. No rush. If I'm wrong, then both sides of the fork will have value, which means I can reassess at that point.

• claims to be a BTC fork but has a 20000 BTC premine
• seems to fundamentally misunderstand how "longest chain" logic actually works
• supposed to fork at the same time as BitcoinABC but there's still no client 1 month later
• nor any details on the proof of work function other than "it's GPU only"
• in fact, none of the links work at all

tl;dr looks like someone was trying to get rich by forking, but got bored and left when it became clear that nobody cared

(Who am I talking about? Bulls? Bears? BitcoinCasheers? That's for you to decide )

Here's the thing about "zero".



Even Zimbabwe banknotes are worth a little money.

Not as currency, of course; Zimbabwe itself switched to USD years ago.

But people buy them on eBay regardless. Why? Well, because it's an interesting object, a funny gag gift, or a conversation starter.

So let's say, for some reason, that one day people stopped using bitcoins. Stopped accepting bitcoins as payment. Let's say the big mining pools all close up shop because it's impossible for an unused currency to pay their electric bills.

Then, even in that world, I imagine the system would probably still live on in the same capacity that Zimbabwe Dollars have survived. Folks buying mining ASICs on Craigslist for a pittance and running them out of nostalgia, or because it's funny, or just because it's something to do.

So we return to pre-2011 prices - sub-$1 prices - from back when BTC was just a funny crypto-nerd novelty. It won't be worth anything as an investment vehicle.

But that's not zero.

The very short version of the story is this:

• When people send their coins to someone else, that's a "transaction".
• They "sign" the transaction to prove to the network that they have the right to spend the coins.
• Before SegWit, transactions and signatures were sent out as part of the same document.
• After SegWit, they're sent out separately - like signing the back of a piece of paper instead of the front.

Why does this matter? It turns out it's useful for letting the Bitcoin network process more transactions per second. Right now there's a limit on how many transactions can be processed per second, because everyone has to double-check whether the transactions are 100% correct, and we want to make sure they can keep up.

It turns out, moving the signature out of the transaction itself helps that in two ways:

1) Partial Verification
Part of what makes double-checking transactions a slow process is that they have to be sent to you before you can start to check them. Moving the signatures means that Alice can check the transactions, and then start forwarding them on to Bob right away, even before she's had a chance to check the signatures (she can tell Bob later should the signatures turn out to be bogus). This means less latency in passing along transactions - they travel faster - and so a block can include about half again as many transactions as used to fit inside.

2) "Fixing Malleability"
The way Bitcoin has worked up to now, since the signature is part of the transaction, changing the signature changes the transaction's identity. Even though all the same coins are being spent in the same way, it's a "different" transaction than before. And it turns out anyone can "smudge" a signature slightly, such that it's still valid, but it's "changed" enough to trigger this situation.

This made it harder to write more complicated chains of transactions - where the transactions depend on each other.

One example of these "complicated transactions" is a "payment channel". The idea here is that if Alice and Bob send coins back and forth a lot, they can compress all of their sending back and forth into two transactions - one to "open" the channel, and one to "close" it and settle up. This saves everyone the hassle of double-checking all of Alice and Bob's payments individually; they can just check that the settling-up at the end was valid. (If you've heard the phrase "lightning network", it's basically just a much more advanced version of this same idea.)

With signatures stored inside their transactions, this was really hard to set up correctly, because if the transaction that "opens" the channel gets its signature smudged, the "settling-up" can't find that transaction anymore, so the coins will be left in limbo.

But with SegWit, the signatures aren't inside the transactions. So even if the signatures get smudged, the "settling-up" transaction can still find the "open" transaction, and the technique works.


Together, in theory, (1) and (2) can significantly increase the amount of Bitcoin transactions that can be sent per second. And that means transaction fees go down, and Bitcoin gets cheaper to use. And at the same time, (2) increases the number of interesting things you can do with your bitcoins.

Since both of those make Bitcoin more valuable, the price goes up. ^_^

Do we know how many of the full nodes have upgraded to a version that understands SegWit style scripts?

I did some websearching but couldn't find many details. Not even sure if this kind of info about a distributed system can be feasibly collected.

Joined the community in early 2012, have been sort of disengaged since early 2015. More watching-from-a-distance than anything. But now that a malleability fix (SegWit) is finally landing I've started to pay more attention again.

My BTC are still in a "long-term storage" mode. I could theoretically earn more profits by putting them on an exchange, but I don't want the counterparty risk; when shapeshift.io supports Tether, or when a proper distributed market appears (thus allowing BTC/USD market making without having to leave my coins in someone's webwallet), then maybe I'll get into it again. But even then, a big chunk of that stash is destined to sit quiet until I can spend them at the grocery store's NFC terminal. "Current money or bust", right?

That said, the BitcoinCash split has been a wonderful opportunity for me, since now I have what amounts to a 10% dividend to use for diversification. I'm planning to buy into Ethereum, mostly. My comment above about a "malleability fix" should clue you in that I'm extremely bullish on the potential of blockchain-enforced business logic, and there are a couple killer apps in that space that Ethereum contracts can support but Bitcoin scripts can't. That said, I've got no interest in direct ICO investment; why try to pick the best individual product, when you can invest in the platform they all use? ^_^

Usually, if someone likes a coin, that person will think it has a bright future.

That said, if you're worried that you might be getting off the rocket by selling your BCC, I'd suggest hedging! You can sell most of it, take profit immediately, but still hold the rest Just In Case it sees major appreciation after all.

Another proposal was to embrace the fact that people seem to be storing data "on-chain", and make the financial incentive system revolve around that. A certain coin balance would be allowed to have up to a certain UTXO size, but these balances would gradually drain to pay the people storing the UTXO set (who would "mine"/get paid by proving to the network that they have the blockchain saved and publicly available to download).

These proposals, I'll reiterate, are most useful when discussing possible altcoins. There is a lot of momentum behind Bitcoin as a pure SHA256 proof-of-work thing; I think changing the proof of work would be too big a difference for people to keep thinking of the new consensus mechanism as still being Bitcoin.

I'm trying to get a handle on the current state of various SegWit activation mechanisms. Please forgive me if these are stupid questions; I've been tuned out for a while and it's hard to sift through all the partisan jargon to find out what's going on.

• At this point, due to BIP141 activation windows, SegWit can't become active until about August 8th, right?
• At this point, assuming BIP91 miners don't change their minds and they continue to ignore non-BIP141 blocks, BIP148 is a no-op, right?

Ahhhh. Yeah, if some of Antpool's hashpower hops on, then the chances of it existing for long enough to unload some coin are a lot better than I was thinking.

Still, he'd have to commit pretty significantly to the project. Otherwise, if his name is on it, the project is vulnerable to any other big miner that dislikes him enough (and after the contentious year we've had, I imagine there's a few).

IMO, Bitcoin's main strength (compared to Dash, Ethereum, etc) is that it's chain with the biggest hashrate, and thus, the hardest to double-spend against.

This is an especially real risk when we start talking about ideological hardforks. When Ethereum Classic happened, one ideological ETH-loyal miner made public threats to use his hashpower to double-spend on ETC and render it useless; it was only an influx of ETC hashpower that stopped him.

So for BCC, BitcoinABC, etc to be useful, they need to win over a significant % of BTC's current hashpower.

Do you consider that likely, when the /NYA/ presents a less risky road to increased blocksize?

Edit: I just realized I didn't answer the original question.

Long story made short, I don't think I'll have the chance to sell my BitcoinABC, because I don't think there will be a market for it.

If there is a market, I'll probably sell most of it. Not all - might as well hedge a little. But most.