Exactly, which is why I don't expect them to short immediately, & already kill what could be profitable for them.

Even if they do short immediately, it's still their best interest to make it go up again. But I don't think they will, as there's a risk it would scare everyone away.
I'd expect them to let it double, and then maybe manipulate it down.
But manipulating it down to 1k, buying everything & making it go up again? There's a high risk that it'd just stay at 1k & some alt (like.. BCH) would take over. And I do think they are smarter than that.


Besides, you're late with your story, everyone told the same when it reached 10k.

Hum, 1000 people is a lot. How would you even imagine 1000 *coordinated* people.

Not like they were in a sect or something. Perhaps just some of them are buddies.

I would be more scared by miners.

yeah, by the time I wrote that, on GDAX the BTC dropped by 2500, and then went up 1300. Crazy, considering these amounts were the price of the bitcoin a few months ago. And it's still 1500eur higher than on Bitfinex.

imagine when it'll be at 500k

What worries me the most is the difference between exchanges & currency.

Do you guys realize it's at 14500eur, which means $17000 on GDAX?

WHAAAAAT IS HAPPENING? I feel like the "it's over 9000" joke was a week ago. Wait, it WAS a week ago!

Arbitrage right now has to be very profitable from Bitfinex to GDAX where the BTC is

2500EUR HIGHER

W   T   F

Yeah, that was my question, how many times have you witnessed this in 15 years?

What I'm describing has happened 3x in a few weeks, and that's just those I saw on the 2 exchanges & few coins I'm checking. One made the news and amazed everyone, showing that it's not something many people expect to ever happen.

Moving fiat is slow but it's just that, there is nothing unsafe (it may cost, but you know how much in advance), you just have to move afterwards, during which time you can't trade anymore.

And in 15 years, how many times have you seen a flash crash down to -95% that did NOT rebound immediately?


The rebound totally depends on the true value, and that is exactly what you estimate to be that value that should be the limit price of a stop order.
Hey, I see a flash crash down to -95%, regardless of the reason, a hack or some chinese tweet, I buy!


On the ETH case, I hadn't seen this guy's screenshot btw. This shows that, as someone pointed out, it can be very true that some insider in the exchange could use the precious stop loss info to make big money. A million in a few seconds.... I wonder for how long that order had been sitting there.
https://stocktwits.com/jdemasie/message/86641301

ok, so we at least agree that in a fast market, chances are that the stop won't be processed.

We just don't agree on whether it's a good thing or not. I'll still maintain that there is always a price at which we don't wanna sell, a price that can only be reached because of technical limitations, nothing linked to the market itself, that is, nothing linked to demand & offer at all.

A stop loss final sell price should be based on the lowest demand, and it simply is not. It is based on the lowest demand in the current order book, which is only one part of the demand.

And you keep saying that a stop has its uses, when things like the mini-temporary crashes in the ETH example don't happen. But no one can predict when they happen, so the fear is always there.
Normally the guy who got his ETH sold for a few cents should have had his ETH sold to the highest bidder in a normal, "human time" auction, in which all of his ETH would hae been sold way way above 10 cents. Perhaps 75usd at the lowest. Which would have still been a bargain for the buyer, and would have been way better for the seller.
The books only represent those who are ready to have a lot of fiat stuck doing nothing in the exchange (which is already a risk in itself), not the ones who have some fiat & are ready to buy opportunities when they see them coming, which I believe is the majority.

There's actually arbitrage to do on BTC which is constantly higher on GDAX, especially in EUR, and these last days, generally 300eur above other exchanges (sure, it's only 3 or 4%, but it's easily done).

+ it looks like the 10k wall on GDAX is smaller, so it's quite likely that the 10keur barrier will be broken first on GDAX, while on other exchanges it will still be struggling. That will probably last for way more than the time it takes to transfer a BTC.
Probably enough people are waiting for that, though.

I'd do it if I could register GDAX, and if I wanted to bother with slow SEPA transfers afterwards.

You haven't explained anything, you just said it would never be filled. Perhaps you're talking of a stop order that would have its limit at the stop price, which isn't something that exists in any exchange, because that wouldn't make sense.

Stop limits have a price trigger and a limit price, they are separate.

of course it's possible, Bitfinex has a stop-limit, Kraken had (before becoming worthless), & a lot of other exchanges I'm sure.




yeah, I often have low orders to benefit from those cascades, but you need a lot of fiat sitting on the exchange for that.
That said, 1000 at 10 cents, that's just 100, quite doable to place one order for each currency.

Why wouldn't it be filled?

If you set your stop at 50, and your limit at 10, there, if it drops below 50 and you're near the top of the book, you will sell at 40, 30, 20 maybe. In any case, at prices you're ok to sell at.
If you're the last in the book and there aren't enough buyers, it will indeed NOT sell at 5. But that's the whole point!, you normally DO NOT WANT to sell at 5.

Let's face it, your value was 100, selling at 5 is pretty much the same as losing. 5 is no value. If it doesn't stop at 5 and goes down to 1, sure, you've lost even more, but you had pretty much lost everything anyway.
The reason you do not want to sell at 5 is that you are pretty much guaranteed (& you know it) that if a market falls from 100 down to 5 in a minute, it doesn't mean no one wanna buy, but it's a chain reaction of stops and no one IS GETTING A CHANCE TO BUY.

You are pretty much guaranteed that if your share goes from 100 down to 5 in a minute, it will go back to 10, 20 or 30 a few minutes later. So why the hell would you want your stop to sell at 5 and lose every chance to get back to a decent level?

The guy who got his ETH sold for a few cents, that's just not normal, it's a technical problem, not a human problem. Had the guy been given access to a stop limit, he would not have set his limit to a few cents. He would have set his limit to like 100usd (I mean come on, an ETH at 100, pretty much guaranteed to sell), thus his stop would indeed NOT have been processed, and he would have been happy with it.
Sure, he has to be aware of the technical problem behind stop orders, but he's not given an alternative. Which is why I advise to simply not use stops, unless you're certain that there won't be a massive drop, which just can't apply to crypto.
Bitfinex provides stop-limit, I really advise people to use that instead. This said, limits are normally processed after market orders, so it's not that ideal either, but still safer.




I gave examples that just happened, have happened a few weeks ago & made the news, and will happen again. The bitcoin seems to get its daily instadrop of 10% these days. So for alts it's even worse.

You will only make easy money from it by daily trading. Which is probably why you have pending buy orders most likely at, let me guess, 1300 or even 1000, to which it always gets back?

Because you know a lot of big money was wasted during the parabolic rise, and there are now huge walls up there from people who wanna get their money back. But be my guest & go buy their debt, they are waiting for that.

Because it seems to be way easier & too tempting to manipulate market downwards than upwards.
Hell it even happened because of a simple tweet last time. That was just a rumor - ok, which got later confirmed, but still.

Of course shorting is legal, it's the ways that people shorting push the market down that are illegal, but how do you even prove such manipulations..


Was it a normal mini-crash, it depends which chart you look at.
If I watch BTC-EUR on Bitstamp, I think the 8000 to 7000 candle would have been avoided with better handling of stops. And then there would probably have a bit less panic afterwards. Instead of a 25% dip it could have been 15 or 20, maybe? And I do think the majority was a cascade of loss - that doesn't mean people wouldn't have sold manually if stops didn't exist. But they would probably have sold less. It's as much the very abrupt dip as the depth of the dip that causes panic IMHO.
But it's still ok. The cases of ETH/NEO/OMG are not ok at all.

Let me explain in depth.
-when you place a limit sell, say at 100, that's the price you wanna sell at, and either it will sell at 100 (in one or more chunks), or won't sell at all.
-when you place a market sell, you sell your share to the highest price of whoever is ready to buy it (in one or more chunks).

You can understand that if you have like 1000 BTC to sell, and you place a market sell, some will be sold at 100, some at 90, some at 85, etc.. and if you have many and the order books are small, some will end up sold at 1 cent.

So that's where the problem of a stop loss is, the price of your stop loss is the price that will trigger a *market* order. You placed a stop loss at 100, it will then sell everything to whoever wanna buy it, at 100 or under.

That happens instantly, thus it only applies to the existing order book, it won't leave a chance to anyone to buy manually.
And when many people have stop orders, the falling prices resulting from market sells will produce a chain reaction. There will be a huge lot of sells that the order books aren't ready to cope with, and your share for which you placed a stop loss at 100, may end up being sold at 1 cent. And some lucky guy who had placed a limit buy at 1 cent, managed to exploit your stop order & made a huge margin in no time.


But the problem you had was probably due to the volatility. You see the market price at 130, you place your stop at 100 and you think you're ok, but the high volatility can make the market price fluctuate heavily between say 90 & 150, fast enough that you don't even see it. You then think that the exchange triggered your stop order by mistake, but it didn't, it just went under 100 & bounced back so quickly that you didn't see it happening.
Even a stop+limit wouldn't have helped here. To place a safe stop you have to watch the order book chart, to see if there is a hole around the mid price, to see if there are chances that the mid price is likely to go up & down very quickly.

Let me take, again, that example of that guy who sold his NEO at $4. But we could take the other example of that guy who got his ETH sold at 10 cents, if it makes it clearer:
https://www.reddit.com/r/ethtrader/comments/6iokzy/gdax_just_sold_a_good_chunk_of_my_ether_at_10/
https://www.cnbc.com/2017/06/22/ethereum-price-crash-10-cents-gdax-exchange-after-multimillion-dollar-trade.html
That made the news some time ago, and people didn't learn from it.

What did that guy do wrong? HE USED A STOP ORDER. HE DID NOT DO ANYTHING ELSE WRONG.

Wouldn't YOU buy an ETH at 10 cents? Who wouldn't? Do you think that if you had asked the guy "hey dude, you put a stop order there, we're gonna sell your ETH at 10cents now, are you ok with that?" he would have replied YES? Yet, that is exactly the risk you take with a stop order, it shouldn't be there, you should be responsible for the amount of risk you wanna take.

People who place stop orders are idiots who believe that if they place it at 100, they are garanteed to sell at 100.
The only real stop order is a stop limit. If you wanna place a stop at 100, you will probably place your limit at 50, or whatever price you're still confortable selling at. But I SERIOUSLY doubt that you will place your limit at zero, or I don't understand what's wrong with you.
Everyone understands that if something falls from 100 down to 0.1 in a few seconds, it's not normal market behavior, it's just a large dump, or a cascade of stops. Then you'd rather keep your coin, hoping it will go up again, rather than simply losing it.
The guy who placed his stop order and saw all of his ETH sold at 0.1, pretty much lost everything, quite ironic for something called "stop loss", no?


Obviously professionnal traders will tell you that stop orders are useful, because they don't trade things THIS volatile for a living.

I really think all exchanges should at least implement stop+limit, and put limit by default to 50% of the stop price.

There are things that could be done, even though I can't imagine any exchange doing it. Like if an exchange gets into a cascade of stops, it enters a phase of "you can buy, but you can not sell for now", spreading the cascade over time.
For ex, the other day's dip would have been spread over like 10min. For 10min, all the stops would have been processed slowly. Any manual sell order would have been scheduled for after those 10min, but any buy order would have been processed immediately. It's quite sure that the dip wouldn't have been this large. Not only it would have dipped less, but it would have rised back faster because there would have been less fear. And things like the ETH or NEO case would simply have been avoided.
It'd be quite easy for an exchange to implement this "special mode" happening whenever a market falls over x% per second, or something.
That's also what whoever is not an idiot would do, if he had to sell a lot of shares. He would spread it over time, selling little amounts, letting the market go up again. Quite possible that the ETH case wasn't a cascade but just 1 stop on a way too large amount.

An alternate solution would be, whenever a cascade of losses is detected, to average all stop orders. That would be less useful, but it wouldn't produce random big losers.

No, no way. Tell me just *1* advantage of a stop order over a stop limit. You'd just set your limit to zero, and a stop limit would become a classic stop (yeah sure, limit orders are normally scheduled after market, but at zero it would just mean market. The thing is that if there are that many market orders in a chain reaction, the last thing you want is to be the last of them and still be forced to sell).
But please don't tell me you would ever set your limit to zero, you would not. You would probably set it to 50%, 30% at worst. You well know that nothing drops 70% instantly without recovering at least a little, seconds later. And 70%, anyway, isn't a "stop loss", it's just a loss.

Yeah, but that's exactly the point.

When you place a stop order at 1000, do you agree to sell at 10? Of course not! No one does. But the exchange can end up selling at 10 and you are screwed.
Stop orders should not exist, they should always be stop limit, and then you are responsible for the limit. If you want a normal stop, then you simply put your limit at zero, which, admit it, you would NEVER do.

If a huge BTC megacrash was causing my BTC to sell at 100, I would be pretty pissed. There is no way the BTC would instantly drop to 100, without EVERYONE buying.