Ethereum Flash Crashes — don’t panic
TL;DR — The Ethereum network was saturated with transactions from a particularly challenging ICO — exchanges had insufficient controls to slow a flash crash, and were isolated from rebalancing one another via arbitrage — a big ether holder in one exchange put in a huge sell order at-market that caused the price to drop precipitously on that exchange only— since then, normality has been achieved — there are a lot of reasons to believe the ecosystem will be able to deal with this type of situation better in the coming years.
If you hold much ether, or maybe if your friends do, you will spend some time wondering why the heck it’s so volatile. About once a month you will wake up in the morning and hear “Eth is crashing” from your friends and/or Google News. The sky will darken. Your cat will give you a look that says, “I knew you would run out of tuna.”
I’m going to assume you are like me — you know something about ethereum, and you are just dang interested in what the heck is going on. You saw the news headlines concerning ethereum on June 21, and you want to understand what happened and what this means going forward.
The first thing you might do, upon hearing news of an ethereum crash, is pop over to reddit.com/r/ethereum, and see if there is an explanation.
On June 21, a bunch of stories read something like: Ethereum Network is so Congested Exchanges Are Forced to Disable ETH Wallets
The explanation of the price dip you might get from that article is that it’s all market-psychology — holders of ether heard that there were delayed transactions on the network (because there were more transactions than the network can handle — saturation — all other transactions had to wait in queue). Some exchanges decided to halt trades (probably a good idea). Some users panicked, and sold everything.
It’s probably fair to say that the most likely people to panic and sell before they know what’s going on are people who don’t understand the product they are invested in and the market around it. Long term holders of ether who follow the Ethereum ecosystem more closely are not panicked by a day of network saturation — they know that multiple scaling efforts are well underway and that some of them will probably be going into production over the next twelve months. A day of network saturation is not going to fundamentally change anything.
Quick aside: If you ever want to see how saturated the network is, etherscan has a count of pending transactions. It’s around 1000 as of this writing and goes back about 2 minutes. That’s what normal looks like. The Status ICO absorbed the transaction capacity of the Ethereum network on the 21st, pending transactions went back for hours. That’s a headline, but not the one that most people saw that day.
Lets dig a little further.
There were a lot of stories like this one on google news: Ethereum flash-crashed to 13 dollars today. (FWIW, it has been pointed out to me that the price of ether on GDAX dropped all the way to 10 cents.)
When you read one of those articles, you find that it was just the exchange GDAX that flash-crashed. That didn’t make a good enough headline, so the author said “Ethereum” crashed. Mmm…K. What’s the explanation for the GDAX flash crash, then? Is GDAX buggy? Not really, but they are coming under attack for other reasons.
GDAX’s blog says that a multi-million dollar sell order came in. Right now I am theorizing that this is due to panic — GDAX will try to determine if the big seller was in cahoots with someone who subsequently bought at the bottom. So far, GDAX has not detected any behavior like that. Selling a giant amount of ether at-market-price gets you a terrible average price on an isolated market with no controls, so it certainly smells like panic to me.
That big sell order filled every buy order in the system down to 10 cents at-market-price, making the price of eth on GDAX go to…well, 10 cents, for a second or two, until cooler heads with FIAT cash on the exchange put in orders.
That is a “simplification” BTW. Margin calls played a role in that price drop, but the most of the selling was triggered by that one large sale. A margin call happens when an investor, who has been borrowing money from an exchange to buy ether, doesn’t have enough cash to cover their debt in the case of further drops in their ether. If the ether drops enough, the exchange can sell the ether to cover the debt, which further depresses the market price of ether.
In some markets, you would get a friendly call from the exchange warning you of the situation. You would then wire in more money to cover your debt, and nothing would be sold. AFAICT, GDAX just sold ether instantly when margins were exceeded. It’s my opinion that there is a market for a higher level of service than that.
The GDAX blog says there were 800 stop loss orders and margin funding liquidations. I assume that both margin trading and the giant at-market-price sale needed to be present to make this one happen.
That seems like a reasonable explanation — a big sale — possibly panic driven — triggering lots of automatic sales due to margin or stop-loss orders — but…it opens a couple more questions.
First, it makes you wonder how mature these exchanges are — shouldn’t they have some controls in this situation? Second, it makes you wonder how these exchanges work. They aren’t walled gardens, are they, unable to engage with buyers and sellers outside their userbase?
To some extent they probably are walled gardens. They are surely especially isolated when they can’t conduct transactions on ethereum mainnet. And ICOs filling up all available transaction slots will do that.
I’m not an expert on how exchanges are run — I’ve only just now done a couple hours of googling and reading about it. I might edit this as I learn more. However, AFAICT, there appear to be a few ways an exchange can maintain liquidity in times of crisis:
First and foremost, some exchanges establish limits and controls that prevent flash crashes and halt trades at certain levels — these can be complex, but some are mentioned here. These can slow down flash crashes. GDAX just didn’t have any of these, and their margin calls were instantaneous. It is only my opinion that controls of this kind should be in place — not everyone wants them.
Second, users will transfer currency in from other exchanges or a local wallet to take advantage of arbitrage— something that should take seconds or minutes. This mechanism is always rebalancing exchanges, all day — every day. This was not available due to network saturation.
Third, and AFAICT this is uncommon among forex exchanges, a liquidity swap could be established with another exchange or other organization. It seems like a good idea to me to have this capability for a rainy day — I’m sure it would have paid off for GDAX, who, not even considering the lawsuits, is in a bit of trouble with their customer base. This swap probably didn’t exist (good luck getting an answer from an exchange anytime soon) but would not have been available anyway due to network saturation.
We have a situation in which exchanges are immature(in my opinion), isolated, and queueing up transactions, making internal trades happen, waiting for confirmation on those transactions, making instant margin calls on everyone who was long ether and didn’t see it going to 10 cents that day (by which we mean — zeroing out the accounts of most of the people who trade on margin). No one could transfer in to arbitrage and support the price, and no other controls or options were in place.
We have a situation in which a day of good news — a hot mobile platform was added to the ecosystem, driving demand for it’s token and Ethereum as a whole, eating up 60 million dollars in ether supply — a great day in which Ethereum could reasonably be expected to rise sharply in price — turned into a day of carnage in the markets.
That was the situation on June 21, and it was not great. Inaccurate news stories that claimed “Ethereum” has flash-crashed to 13 dollars did not help the price of ether, as that level of volatility is something way beyond the expectations of the average investor.
But, all is not lost. Lets look hopefully at forces acting on the ethereum network that might change all this.
Over the next year or so, one or more scaling mechanisms will almost surely be put into place in the Ethereum network, and sidechains will begin to develop into useful tools for ICOs and other transaction-heavy systems. There may be setbacks, but the demand for scaling this thing is an irresistible force, and the brainpower set upon that task, considerable.
As the network scales, the best ICOs add value, and the market cap of the network grows, ICOs will have a smaller impact on the network and the price of ether. Once the harbingers of short-term high prices, currently the harbingers of short-term low prices, ICOs will have very little impact on the price of ether or the function of the network a few years from now.
Exchanges will have to adapt to the actual reliability level of the blockchains they support or die. They will have to design with the expectation that a blockchain could be saturated, and big fish might game the system. It is my opinion that they will need to address this to maintain their user base, otherwise they will be viewed as unreliable and competition will take care of the rest (like this exchange, perhaps).
Proof of stake will provide dividends for ether holders, giving them another reason to hold long-term other than price speculation. This shift in the makeup of ether holders may decrease price volatility in ether.
Finally, one could hope that the average ethereum investor will become more well educated, and see the long term vision of the ethereum network as an incredibly valuable decentralized app platform on which governance and financial systems of all kinds will flourish. An educated investment community is not absolutely necessary for a stable market, but the effort to create one is worth contributing to.
I would never speculate on price movement over any period, but I do believe that the entire ethereum ecosystem has the potential to become a much more functional, less volatile network and ecosystem in the coming years.
Not only that, but the volatility should matter less and less — ether is not primarily a store of value (SOV), but it makes a good medium of exchange and unit of account (MOE and UA). Volatility should matter less and less as long term stakers are rewarded, and short-term speculators continue to get punished, shifting the demographics.
News articles will still preach panic, but they will have much less basis in fact, and, hopefully, we won’t see them on the front page of google news.
At least, I think that’s what will happen. Anyway, the moral of the story is — your cat needs a home —better promise him right now that, next time, you will definitely avoid trading on margin — and you will read more than one news article deep and not panic sell — and that, yes, sparkle kitty, we are adding tuna to the shopping cart now. Click buy. Got it. Good.