Author

Topic: [OFFICIAL]Bitfinex.com first Bitcoin P2P lending platform for leverage trading - page 175. (Read 723903 times)

sr. member
Activity: 288
Merit: 250
ManualMiner
but to do like bitcoimdudeidiot describes you need a huge margin. as i understand he says that closing swaps used in a margin position doesnt trigger the one-hour-min-fee.  but the reserved margin does,  doesnt it?

i suggest putting cluttered swap offers at increasing  rates to catch those spikes,  although swaps would be out max 2 hours

... aha,  i think i got it... is it that the min one hour fee isnt triggered by taking a swap,  but by closing it (but not by closing it when its used in a margin position)...?!


well, so whats the problem? margin traders have to check their swap rates and close them if they catched a spike, lenders could catch a spike, but the swaps would be closed very soon. so all that just results just in some swap-market action but sums up to zero (except the margin traders, that dont regularily check their swap costs..)


obviously theres more about that and apparently bfx dit make some changes and did make some changes and did make some changes  (was there any notification about that?) - and it seems kinda fixed now, which raises the question, what did happen today?:
http://www.reddit.com/user/BitcoinIdiotDude

seems reddit is the channel for news from raphael, here´s mainly giancarlos hangout
full member
Activity: 172
Merit: 100
And done. They started at 9:11 and on time everything is returned at 11:11. Took them 2 hours this time. It really couldn't be more obvious. 2 hours and no FRR to be seen and now suddenly 450k on FRR.
I was so glad we were done with this shit. Gratz to the guys who had their offers in the 0.6... region and hopefully the swap payment is actually credited.
sr. member
Activity: 288
Merit: 250
ManualMiner
have a link to that BitcoinIdiotDude  post you mention?
i am not sure we have proof that no interest is payed, do we?
full member
Activity: 172
Merit: 100
so we have a spike in lending rates, but no ones going long?

Yep. That's what's so amazing about it. These spikes always happen during periods of low volume without any price gains/high volume trades. We can rise $100 and more in a single day and the rate doesn't jump to more than 0.25 but once in a while some player wakes up and thinks to himself "You know what, I should probably borrow myself a cool million dollars and put it in my pocket for an hour without buying a single bitcoin with it"

I would not care too much about this kind of manipulation had I not learned from BitcoinIdiotDude that these guys found a way to not even pay their mandatory 1 hour swap rate. This makes it very cheap to clear out the order book like that. This has probably cost me several hundred dollars of potential swap interest.

Amazingly enough the money hasn't been returned yet. So either they have not been able to lend their funds out at the inflated rates, they are reading this thread and want to prove us wrong or they have found a way to hold the swap for even more than one hour without paying interest.
full member
Activity: 224
Merit: 100
so we have a spike in lending rates, but no ones going long?
Yes, the whole bunch of money borrowed with fantastic interest rate and there is no any movement at the market.
sr. member
Activity: 288
Merit: 250
ManualMiner
so we have a spike in lending rates, but no ones going long?





high rates ahead



high rates seem to come from short time lenders



full member
Activity: 172
Merit: 100
Not sudden?
Eh, I mean it's predictable by those volume spikes.
Moreover, such thing happened a week ago, look at my ugly chart.

It's the same exact behavior that has happened many, many times in the past and was explained well by BitcoinIdioitDude. Then it disappeared, now it's back. Let's see how long this lasts. In the past these spikes almost always lasted pretty much exactly one hour. We'll see.
Bitfinex guy said they fixed that bug and a something about shoe size. Do you think we should check it again?
Yes, as I said this has happened many times in the past. In this very thread I've mentioned these spikes numerous times. Not many cared.
From what BitcoinIdiotDude has written BFX had not found this bug for a long times before so they will likely not find this one (if there's a new one) as well. Maybe an intelligent hobby investigator here can find what's up with this again.
full member
Activity: 224
Merit: 100
Not sudden?
Eh, I mean it's predictable by those volume spikes.
Moreover, such thing happened a week ago, look at my ugly chart.

It's the same exact behavior that has happened many, many times in the past and was explained well by BitcoinIdioitDude. Then it disappeared, now it's back. Let's see how long this lasts. In the past these spikes almost always lasted pretty much exactly one hour. We'll see.
Bitfinex guy said they fixed that bug and a something about shoe size. Do you think we should check it again?
full member
Activity: 172
Merit: 100
Seems like the manipulators have found a new way to rig the interest rate. Sudden spike to 0.69 %!
Not so sudden:


Not sudden?
Switch the view to 6/12/24 hours and you see how sudden it is.
Over 1.3 m taken out in three tranches clearing out the order book. This is completely uncalled for if you look at the corresponding trading volume. There is nothing in BTCUSD or LTCUSB that would even remotely explain such a move. It's the same exact behavior that has happened many, many times in the past and was explained well by BitcoinIdioitDude. Then it disappeared, now it's back. Let's see how long this lasts. In the past these spikes almost always lasted pretty much exactly one hour. We'll see.
full member
Activity: 224
Merit: 100
Seems like the manipulators have found a new way to rig the interest rate. Sudden spike to 0.69 %!
Not so sudden:

full member
Activity: 172
Merit: 100
Seems like the manipulators have found a new way to rig the interest rate. Sudden spike to 0.69 %!
full member
Activity: 136
Merit: 100
First, it's nice to see that you can confirm that there is variation and it's not exactly the amount it should be, so I'm not making this up. In defense of BFXs good intentions and what you stated I'm assuming that most likely some parts of the daily swap payment are not included this day for whatever reason and are then included in the next day so this difference might balance itself out.

I see some small variations, but typically less than a 1% difference from my predicted value. I'm chalking it up to rounding errors that average out over a few days (i.e. if they're only calculating fees owed every 10 minutes, or 1 hour or w/e the time period is, then I'd have to model that to get the same figure exactly, which I'm not).

I did have a period of maybe a week where I was consistently getting paid more than I expected each day, by a slightly larger margin than the normal +/-1%, and my accumulated error is still showing a small surplus because of that. I really doubt there's anything deliberately untoward going on.
full member
Activity: 172
Merit: 100
I try to calculate my expected swap payment for every day and almost everyday it's either up to 14 % lower or 15 % higher than expected.

I do the same in a moderately automated way (copy/paste swap details into a spreadsheet, it takes care of the rest), and while it does vary a bit around the actual figure each day, I get closer than 14-15%.

All I'm doing is amount*duration*rate*0.85 for each swap though, so... not sure where errors would be creeping in for you.

How do you handle accumulating errors between days? For mine, I keep a running total of how much I've been wrong by, and add the error for each day onto that total so that I can adjust for it and, in effect, start each day from zero. Some days it's a little over, some days it's a little under, but the overall total error has mostly corrected back towards zero over long enough spans of time.
First I start with an upper ceiling ball park estimate which even then sometimes reveals mistakes.
(Highest Swap Rate)*(0.85)*(lend out funds). This works for me when the swap rates are sufficiently similar and you have nearly all your funds lend out. For example even this upper ceiling estimate showed that yesterday's swap rate was about 8% higher than it could have possibly been. If there is nearly to no difference like there sometimes is, I stop right here.
If the difference is large enough I go for more detailed calculations similar to yours which often reveal and even larger difference (11 % in this case).

I'm doing this long enough that when I see the swap payment rate I immidiately have an estimate in my head (yep, fits, pretty good! or nah, this can't be) which almost always turns out to be correct.

First, it's nice to see that you can confirm that there is variation and it's not exactly the amount it should be, so I'm not making this up. In defense of BFXs good intentions and what you stated I'm assuming that most likely some parts of the daily swap payment are not included this day for whatever reason and are then included in the next day so this difference might balance itself out.

The P2P lending is surely a complex feature and we have hear about numerous problems in the past (the time of the swap payment was always fluctuating, now it's almost always at 2:40, the incident where some hours were not credited, the one time where we got a paymen 2 days later because the original one was too low, etc. etc. I'm still wondering about the very interesting intricacies BitcoinIdioitDude has laid out (the possibility (that now seems to be fixed) to churn aways loans by cycling them through would directly implicate that loans were taken out without the lender receiving a swap payment for this etc.)

That being said I still consider this feature an amazing feat of BFX and just hope it's in their best intent to not screw us over because they would probably have far more effective means to do so. For reasons of transparency I would be happy for some kind of counter (like the unrealized swap cost one in the margin trading tab) that shows the accrued amount of due swap payment and is then resetted with the daily pay out.
full member
Activity: 140
Merit: 100
@unclescrooge:

Good to hear that this has been resolved.
Have there been problems with the API today too?
Third party sites didn't receive any Bitfinex data for some time..

The orderbook on bitfinex.com doesn't update nearly half as fast as http://bitcointicker.co/ and the others. A live, fast orderbook would be great. The site mentioned removed the feature to expand the orderbook depth so I only use bitcoinwisdom right now, but I think it was really useful when it worked.
->a bigger, really fast live orderbook
->fast tickchart
->beep on trade (just hearing when activity increases instead of setting artificial price alarms is quite nice)

Could that have been due to DDoS attacks again? It happened at the same time as todays dump...
hero member
Activity: 868
Merit: 1000
Hey guys,

The orders can now be cancelled, they couldn't for a few moments because of a broken link with the trading engine.

Thank you
Raphael
Bitfinex team
full member
Activity: 224
Merit: 100
Oh jeez.
Some time ago I decided to collect exchange statistics by my own, but now I'm thinking about starting yet another exchange.
newbie
Activity: 42
Merit: 0
^ "cancel all" still works.

We need a statement of the BFX team very soon!!

This caused a tiny panic in our tradingroom already and it doesn't feel very trustworthy if your exchange behaves weird all of a sudden...


Thanks!!!
full member
Activity: 140
Merit: 100
^ "cancel all" still works.

We need a statement of the BFX team very soon!!

This caused a tiny panic in our tradingroom already and it doesn't feel very trustworthy if your exchange behaves weird all of a sudden...
newbie
Activity: 42
Merit: 0
I cannot cancel my orders!

When I click on cancel, the error message is "The order could not be canceled". One of the order ID is: 62432676

Nick.
Jump to: