"HCP's ongoing experiment of the BotHIVE.io Trading Platform"™

[HCP]

UPDATE: 2021-05-09

REKT!

Who would have guessed?


UPDATE: 2020-04-21

Ok, so it's been fairly quiet since Arakne failed in spectactular fashion at the beginning of Feb... but you can now follow along as I try to lose the rest of my ~0.035BTC out the new Trading Platform that has risen up from the ashes of Arakne... Say hello to BotHIVE.io...


EDIT: 2019-12-30
So the "month" has ended... and after a nice little gain to start, the bot got #rekt.

Firstly, by "manipulation" ... resulting in a stoploss being triggered and a 30% loss of balance (ending in a -18% overall). Then more recently, an early Christmas "present"... A giant lump of coal! a 2nd massive loss, this one due to some pretty broken bot mechanics!

Anyway, as per my explanation here, I figure I may as well just let it run... and will continue to update once (if? ) the bot is back online.


Backstory:

Ok, so I somewhat stumbled across this service due to the bunfight between TMAN and the Service Owner. I think I originally followed a link from the "shit post" thread? Anyway, I digress... so after some (possibly ill advised) alcohol consumption, the old gambler in me was itching for something new. Rather than focus on somewhat ludicrous claims from backtests (I've never been a real fan of such things) and what I perceive to be possible language difficulties and personality conflicts (and with LoyceV's 10 month Altcoin experiment on hiatus)... I decided the "best"(??!?) way forward was to just "roll the dice" and see if it works

So, without further ado, welcome to the first month of "HCP's (at least) one month long ongoing experiment of the Arakne Trading Bot"™


Initial Setup:

The original service thread has been filled with claims and counter-claims (and what I would deem to be "BadMaths"™), so I did some napkin math of my own and decided that the "minimum" investment amount of $100 wasn't going to cut it, as (should be obvious to anyone that can do basic math) the returns would quickly be eaten by the 30% commission and $10/month VPS fee... you'd need be making at least 15% a month in GROSS profit just to cover both of those... for the math impaired: 15% of $100 = $15... -30% ($4.50) commission  = $10.50 - $10 VPS = 0.50c NET profit... Hardly what one would consider a "stellar" return on investment!

So, partly because I like round numbers and partly because that was what I had laying around and partly because I figured I could afford to lose that much without being tooooo upset about it, I decided that I would commit 0.1BTC (around US$735 at the time of deposit) as the initial investment... and paid the US$10 VPS (0.00133645BTC) up front. Setup was relatively easy... I just needed to signup for a BitMEX account, deposit the initial amount, create an "order only" API key and then pay the VPS fee (via Coinify I believe). There was a slight delay in getting setup because the bot was already in "trade mode" but as it turns out, that actually worked out in my favour

Basically, because I was fucking about getting all my stuff sorted and asking questions, it seems like I got "lucky" and the bot was already in the middle of the "loss" trade that the other users encountered a few days ago, so my account wasn't setup until after that had finished. #blindDumbLuck

The purpose of this thread (similar to what guigui371 has done in the service thread) is an effort to document the performance of the bot, from a "basic" starting position of 0.1BTC... I'll be looking at posting semi-regular updates, with little if anything in the way of judgement. I'm happy for the numbers to speak for themselves... "warts and all", wins and losses etc.


Current Results:

Daily Results:



Overall Results:


So, after 3 days of trading my account is 17% up... 0.01772045BTC ~= US$134 GROSS (~= US$84 ~= 11% NET profit with commission and VPS accounted for). A positive start. Will be interesting to see where I am at the end of my month #fingersCrossed

NOTES:

- Deposit was made on Nov 29, but Bot was only setup and running on Nov 30th in my timezone UTC+13.

- It seems there is something wierd in the display of dates by BitMEX... today is Dec 2nd UTC+13, so not sure how BitMEX is already in Dec 3rd??!?

- I am aware that some users DO NOT have to pay the VPS fee due to special promo's and deals... I am NOT one of those users, nor does it appear that this deal is currently available (I asked when signing up), so the VPS fee will be included in all my calculations until such time that the "$50 monthly commission == Free VPS" rule kicks in.

- I have been advised that there are likely to be a few minor 0.000000xx BTC "losses" incurred from time to time due to a problem with the BitMEX API. My understanding of this problem is that if you don't trigger the API at least once in a 24hr period, it can become unresponsive and "freeze", meaning that the bot will no longer be able to function properly. The end result is that on days when the Bot is not trading, there needs to be a small instant "buy/sell" type trade just to keep the API "active". Again, this is my interpretation of the issue and probably a bit oversimplified, so I apologise for any inaccuracies.


As I said, going forward, I shall try to regularly post the results so there is some more "independent" data available to complement what guigui371 has already provided in the main service thread.

Also, I should point out, that I am a complete novice idiot when it comes to trading... I have no idea about margins, positions, leverage etc... The limit of my knowledge and trading has been at the level of "lets buy some BTC at $X... lets sell some BTC at $Y"... hopefully with X < Y! So, don't be expecting any in-depth market analysis or pretty charts with lines drawn all over them claiming to predict the next pump or dump... it'll basically be some screenshots showing daily and overall results with the basic calculations of GROSS and NET profit as above


<insert disclaimer here about this being an independent project and should NOT be seen as an endorsement of either the service or the service owner. It is simply documenting the performance of the bot and I am not receiving any kickbacks or payments for doing so>

[1714577960]

1714577960

[1714577960]

1714577960

[1714577960]

1714577960

[1714577960]

1714577960

[1714577960]

1714577960

[guigui371]

Hey HCP,
This is an awsome thread.
I was wondering if I should do one as well.
As you may have read in the original thread I have posted the results of my journey too.

My intent was not to defend the OP, nor to enroll people, but to see by myself if what he says is remotely true or not.
It will be great to be able to cross reference our results.  Now with 2 "independent" observers none of us will be flagged as a scammer. or OP's alt. 


The layout of your thread is great (bb code and / font ...).
Do you mind if I use the same layout ?   (with due credits, and link toward your original post).

Cheers

[HCP]

Do you mind if I use the same layout ?   (with due credits, and link toward your original post).

Sure... for a one off "licensing fee" of 0.1BTC

Seriously tho... Do what you like... I'd hardly call it a layout! I typed some stuff up, chucked in a couple of screenshots and used some bold, underline and some horizontal lines If you do create a new thread, either post a link here or PM me or something so I can add it in.

[CryptoSparks]

Finally someone trying the service before accusing me of scam, ponzi and all that good stuff i've been fighting for months.
THANKS

EDIT: Hilarious seeing suchmoon meriting this thread, oooh lady, i've been enjoying so much our fights  

[TMAN]

~

Awesome, this is what I wanted.

~

2 Gentlemen/women who's data I will trust implicitly, really looking forward to this and I do wish you both the best luck with the profits here.

All it will take now is for CryptoSparks to shill the hell out of your real tests and he will turn himself into a profitable business as long as the profits come in for you both.

[buwaytress]

HCP, been a while since I saw you pop up on my screen like this. We're both gamblers, but likely in very different spheres of gambling that we don't end up meeting much;)

I've had that trading bot itch on occasion, just never scratched any of those itches for btc. Probably would have, if it didn't take too much of my time. Looking forward to your dailies (yes, please do dailies if possible, would like to see the entry/exit, durations of the trade too).

GL

P.S. 11% nett is pretty nifty so far, but I almost would always avoid something with such a seemingly high commission and cost.

P.P.S. Hope people realise this is, after all, a gamble! Nothing less;)

[HCP]

Looking forward to your dailies (yes, please do dailies if possible, would like to see the entry/exit, durations of the trade too).

Daily's aren't necessarily worthwhile... There are some days where the bot remains in "stasis" (as the dev calls it)... Basically, the bot is looking for favourable conditions before activating and starting to trade. It could be several days before more action. In the meantime, all you'd see would be the 0.000000xx BTC "insta-trades" to prevent the "API death" bug as mentioned in the OP.

P.S. 11% nett is pretty nifty so far, but I almost would always avoid something with such a seemingly high commission and cost.

It's certainly at the higher end of what I am "comfortable" with commission-wise... But it literally is ZERO effort/knowledge required. People talk about "longs" and "shorts"... To me those are different length pants!

Who knows... Maybe if he gets enough users, the commission rates might come down?

P.P.S. Hope people realise this is, after all, a gamble! Nothing less;)

Exactly... Nothing is guaranteed...But hoping for the best obviously!

[HCP]

Ok... so today didn't start well (a wakeup call at 0550 from crewing calling me in for a standby duty)... and it got even worse when I looked at the notifications on my phone  

Seems the bot decided that a "short" was the way to go... just before some whales decided to have some fun and the market spiked like a mofo and jumped like US$600-800 which triggered the stop-loss at "max 30% loss" The net result is that the 17% gains from the first 5 days of trading... just got crushed and are now an overall 20% loss #bugger

The Trade History from Today:



The Order History from Today:



Which gives us the daily PNL:



And the Overall:



So, there we are... -19.96% overall... that's going to take a while to turn around. I know some of the bot users are taking this opportunity to "re-fund" their accounts and deposit more... I guess this is an attempt to recover their loses by making more $$$ on any % gains that the bot makes over the next few weeks. I basically need to make 25% gains from my present position (0.08BTC) to get back to break even (0.1BTC). On previous bot history, that is going to take ~2-3 months.

I guess this little "experiment" is going to run a little longer than I had initially hoped!



PS. It's not all bad... Liverpool stomped Everton 5-2 in the Merseyside Derby today

[mk4]

I guess this little "experiment" is going to run a little longer than I had initially hoped!

Well, to be fair, it's a one month long experiment in the first place.  Most probably going to take more than a month to make up the losses though, and that's assuming almost everything goes well.

[pooya87]

could you please also add an ELI5 of the strategy that the bot is using to make this profit. i checked the main thread, it only had a simple 2 line explanation inside a picture that didn't have that much details.

it seems like the bot is using some sort of martingale "gambling" strategy and it also seems to me like it only performs well when the market is moving in a fixed direction (eg. from Nov. 30 to Dec. 4) and makes small profits during that time but if there is any change in the direction or any kind of volatility (eg. Dec. 4) it fails and leads to a big loss.

[HCP]

could you please also add an ELI5 of the strategy that the bot is using to make this profit.

Honestly... I don't really know... so can't really give you any real details of the strategy. The owner keeps all of that confidential, as is his prerogative I suppose.

My understanding is that it supposedly has some custom coded "algorithms" that search for various trading patterns and attempts to jump in and try to make some small gains. It does both "shorts" and "longs" depending on what it thinks the market is about to do. It is setup to "stop-loss" at max 30%... as we witnessed today

it seems like the bot is using some sort of martingale "gambling" strategy

This seems to be what people are saying... and what it looks like with the increasing "bets"

... and it also seems to me like it only performs well when the market is moving in a fixed direction (eg. from Nov. 30 to Dec. 4) and makes small profits during that time but if there is any change in the direction or any kind of volatility (eg. Dec. 4) it fails and leads to a big loss.

I guess it depends. The bot doesn't constantly trade or necessarily trade on any given day. As I said, my understanding is that there are some certain "triggers" that it uses to determine whether or not conditions are favourable to start trading.

It would seem to be a case of "bad timing"™ today with the pump and dump that occurred that screwed things up. The bot initially saw this 331m volume drop:



and then got rekt by an 803m volume pump:



As you say, it's been quite volatile today:



I'll keep posting the trades and results... and we'll see where we end up. But if anyone has any requests for specific data/screenshots etc, feel free to ask and I'll do my best to accommodate your requests.

[suchmoon]

could you please also add an ELI5 of the strategy that the bot is using to make this profit. i checked the main thread, it only had a simple 2 line explanation inside a picture that didn't have that much details.

it seems like the bot is using some sort of martingale "gambling" strategy and it also seems to me like it only performs well when the market is moving in a fixed direction (eg. from Nov. 30 to Dec. 4) and makes small profits during that time but if there is any change in the direction or any kind of volatility (eg. Dec. 4) it fails and leads to a big loss.

That's pretty much it. There is no magic and unlike dice gambling this uses 20x leverage so it only needs a very small move (like ~$200 in this example) in the wrong direction to fuck it up. Given BTC volatility, do you really want to gamble against it moving $200 in the wrong direction?

This also shows that CryptoSparks basically faked the backtest (at best made the "strategy" fit historical data) to make the line look so smooth up-and-up.

[Vispilio]

...

This also shows that CryptoSparks basically faked the backtest (at best made the "strategy" fit historical data) to make the line look so smooth up-and-up.

Unfortunately for Arakne (and its current subscribers), this is a good point.

One of the fundamental weaknesses of any backtest is curve fitting and over optimization bias, where the strategy producer adjusts the parameters to produce the absolute best results for a given set of data points, which does not necessarily provide any real trading edge, but simply provides a coincidental configuration for excellent results over a given past time period...

Contrary to popular opinion, however, Cryptosparks is a smart guy and has some strong insights into algorithmic trading  , so I hope these live experiments by veteran BTT users can be a good opportunity for him to enhance his trading and developing skills.

[LoyceV]

30% commission

If the bot would be as good as the owner claims, he wouldn't need to ask for commission, he'd just keep the bot for himself and get his many-times-return each year.
He now makes a risk-free 30%, while the bot-users risk their funds.

[HCP]

He now makes a risk-free 30%, while the bot-users risk their funds.

Well, he only gets commission if the users actually make a (monthly) profit... if it's a loss, he gets nothing. So, theoretically at least, it is in the bot creators best interests to make sure that the bot works and generates profits.

Also, just as an update, the bot has been quiet the last few days... no trading to speak of since the big loss (excluding the daily API bug workaround instatrades for 0.000000xx btc).

[PrimeNumber7]

- I have been advised that there are likely to be a few minor 0.000000xx BTC "losses" incurred from time to time due to a problem with the BitMEX API. My understanding of this problem is that if you don't trigger the API at least once in a 24hr period, it can become unresponsive and "freeze", meaning that the bot will no longer be able to function properly. The end result is that on days when the Bot is not trading, there needs to be a small instant "buy/sell" type trade just to keep the API "active". Again, this is my interpretation of the issue and probably a bit oversimplified, so I apologise for any inaccuracies.

This seems like a pretty big problem on either the part of Bitmex, or the bot creator. There shouldn't be any reason why a VPS shouldn't be able to connect to an API endpoint given it has a valid API key/can authenticate.

could you please also add an ELI5 of the strategy that the bot is using to make this profit.

Honestly... I don't really know... so can't really give you any real details of the strategy. The owner keeps all of that confidential, as is his prerogative I suppose.

My understanding is that it supposedly has some custom coded "algorithms" that search for various trading patterns and attempts to jump in and try to make some small gains. It does both "shorts" and "longs" depending on what it thinks the market is about to do. It is setup to "stop-loss" at max 30%... as we witnessed today

If you are running a program on a VPS you have access to, you should be able to review the code to see what triggers the bot to make various trades, and have an idea as to what strategy(ies) it is using. There is no 'guaranteed' way to make money trading, and different strategies will have different success in various trading environments.

[HCP]

This seems like a pretty big problem on either the part of Bitmex, or the bot creator. There shouldn't be any reason why a VPS shouldn't be able to connect to an API endpoint given it has a valid API key/can authenticate.

I believe it is an issue with a particular endpoint of the API and the fact that the bot doesn't trade every single day... bot owner explained it here:

Note that the negative PNLs of 0.0001 mBTC ( 0.001$ loss) are created by instant 1 contracts trades necessary to escape a bug in the bitmex websocket's position endpoint. After several days without trades, the position value restores and the endpoint returns error when trying to access it. This causes the bots to crash. What's the solution until bitmex fixes the problem? Make instant 1 contract trades to unlock the endpoint. That 0.001$ loss is the fee paid as taker(necessary to close the "unlock trade" in less than 1 second) and the mathematical loss that come from the 0.5$ of spread between the ask and the bid.

I'm unsure as to why BitMEX have not fixed this... or if there is anything else that can be done to work around the problem... in any case, the "losses" from this are ~20 sats per non-trading day... it's not what you would call "huge"... even an entire year of non-trading would only incur around 7300 sats of losses...

If you are running a program on a VPS you have access to, you should be able to review the code to see what triggers the bot to make various trades, and have an idea as to what strategy(ies) it is using. There is no 'guaranteed' way to make money trading, and different strategies will have different success in various trading environments.

You pay for the VPS, but you don't have access to it... that is how the owner is protecting his "IP".

[buwaytress]

OK, so I finally understood that it is a mini-martingale tactic going on in there under the hood. Which I suppose could work out to be a brilliant strategy if you have a big bankroll at high leverage -- precisely what BitMEX allows. But if you're not redepositing (I take it you're not), then the experiment isn't similar to what the owner would otherwise recommend right?

P.S. Did not realise you were a Reds supporter. Funny how all our betting/trading/gambling (they're not all the same are they? ) all seem much better when we win eh?

[suchmoon]

Well, he only gets commission if the users actually make a (monthly) profit... if it's a loss, he gets nothing. So, theoretically at least, it is in the bot creators best interests to make sure that the bot works and generates profits.

If the bot person would refund 30% on losses - that would be a decent incentive. Otherwise he doesn't risk anything. And he certainly doesn't care about customer relations, so you can't say the incentive would be retention or anything like that.

[squatter]

He now makes a risk-free 30%, while the bot-users risk their funds.

Well, he only gets commission if the users actually make a (monthly) profit... if it's a loss, he gets nothing. So, theoretically at least, it is in the bot creators best interests to make sure that the bot works and generates profits.

The pricing model suggests that he expects short-term profitable periods that he can extract rent from, but that he isn't confident about long term profitability. If the bot were consistently profitable, he wouldn't want to give away 70% of the profits by licensing it out. He could get rich by compounding gains instead.