Darkstone2 (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 25, 2014, 10:05:33 AM |
|
My and a friend have made a successful trading bot. My friend is joking that we should sell this trading bot to customers. I'm not overly convinced that is the best decision on the long term. So... i would like some community input. Lets make the following ballpark assumptions about this product: - We've started with 1 week worth of CPU mining scrypt coins (Digitalcoin
) during the november bubble. So high amounts of cash is not required to make profit. - Bot gains an average of 5 BTC per month.
- For the last 6 months.
- We have pretty graphs (we maintain a github project that is dedicated to generating automatic reports from bitcoin exchanges) from this 6 months.
- We can provide sufficient evidence that these numbers are really accurate. For example, by releasing API keys of these exchanges to the public so that everybody can see if the amount of bitcoins deposited/withdrawal match up to our graphs.
- We are a registered company. Which means we pay taxes, and you have our names.
Given these circumstances. How much would you pay for unlimited access to this bot? And how about 1 month access?
|
|
|
|
xybersurfer
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 72
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 25, 2014, 11:06:10 AM |
|
i think it would be better to talk in terms of percentage gain. (no one has any i what 5 BTC would mean)
you could post the graphs here and save people the time of having to access the data with your keys.
can you atleast give a vague explanation of how your bot works?
how does it compare to buy and hold?
|
|
|
|
Darkstone2 (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 25, 2014, 12:09:31 PM Last edit: May 25, 2014, 10:45:04 PM by Darkstone2 |
|
i think it would be better to talk in terms of percentage gain. (no one has any i what 5 BTC would mean)
The ballpark of 5 BTC per month is unclear? In terms of percentages, since we started with so few bitcoins, the percentage gain is measured in the millions. I omitted that figure on purpose since it is so useless. The folks at https://bitcointalk.org/index.php?topic=551698.0 gave only percentage gains. And i came to the conclusion that i can do nothing with that data. you could post the graphs here and save people the time of having to access the data with your keys.
Of course i will be posting graphs. But graphs can be false, which is why i wanted to provide customers with the ability to check if the graphs are not falsified via the api keys. However, at this point i'm not sure if we're really going to sell this bot, so i'd rather not provide exact specifics (read: graphs) right now. Lets just assume that the ballparks i gave are correct, if so, how much would you be willing to pay? 1 BTC? 5 BTC? 25 BTC? Up front? every month? I need that input before i can make a decision to sell. can you atleast give a vague explanation of how your bot works?
It buys and sells stuff. Thats.. about as much information as i'm willing to give. how does it compare to buy and hold?
We've started with coins worth a few days CPU mining. And we are now [ballpark] gaining 5 BTC a month. Do the math.
|
|
|
|
agent725
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 25, 2014, 12:20:20 PM |
|
That sounds great, however, if your bot is capable of making so much money, why do you even consider selling it?
In the past, I have often analysed seemingly profitable algorithms on years worth of market data to measure if the trading bot could keep up profitable trading in the long term. Sadly, no bot could trade succesfully without extreme losses in terms of years of trading. In that sense they were no more than complex Martingale algorithms.
It would be convincing to set out and prove that your bot could stand up to the test of price and time. How about downloading a large amount of historic charts and letting your algorithm crunch that data?
|
▄██▄ ▐████▄▄▄▄▄▄▄ ▄████ ▄▄ ▀██▀ ▀▀▀▀▀▀▀█████▌ ▄████▌ ▐█ ███▀ ▄██████▀██▄▄▄▄█▌ ▄██▄ ██ ▄▄▄ ▄██▀ ▀██▀ ██▀▀▀▀██▌█████▌ ▐██ ██████▀▀ ▐▀ ▄▄██████ ██ ▐██████ ▄█████▄▄▄███▀ ▀███▀ ▐██ █████▀ ███████▐█▀ ██████ ▀▀▄█▄ ███████ ▐███████ ██▀▄███▄ ▀███▀██▄ ▄▄▄▄ ▐██████▌ ▀███████ ████▄ ▄██████▄ ▄███▀████▀ ███████▌███▄▄▄▄▄▄▄████▐███████████▀ ███████ ███████████████████████▌▀ ▀███▀ ▀▀▀▀▀▀███▐████████ ▀████▀
| | ◉ ◉ ◉ ◉ | | ◉ ◉ ◉ ◉ | |
|
|
|
Darkstone2 (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 25, 2014, 01:32:32 PM Last edit: May 25, 2014, 04:12:51 PM by Darkstone2 |
|
That sounds great, however, if your bot is capable of making so much money, why do you even consider selling it?
Two reasons: - If my bot makes 5 BTC/month, and i find 5 people willing to pay 1 BTC/month for it. That means 10 BTC/month.
- My bot makes money now, but i can't predict the future. Will it still make it's 5 BTC next month? Next year? Selling can act as an insurance for future profits.
In the past, I have often analysed seemingly profitable algorithms on years worth of market data to measure if the trading bot could keep up profitable trading in the long term. Sadly, no bot could trade succesfully without extreme losses in terms of years of trading. In that sense they were no more than complex Martingale algorithms.
It would be convincing to set out and prove that your bot could stand up to the test of price and time. How about downloading a large amount of historic charts and letting your algorithm crunch that data?
There isn't enough historic data available to crunch. My algorithm relies om the orderbooks of many differed markets on exchanges. Such data is not available in the bitcoin space. I have tried different algorithms that rely only on past trades so that it can be backtested, but only a small percentage of profit (less than 1 BTC all-time) comes from such algorithms.
|
|
|
|
xybersurfer
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 72
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 26, 2014, 08:35:25 AM Last edit: May 26, 2014, 12:05:31 PM by xybersurfer |
|
That sounds great, however, if your bot is capable of making so much money, why do you even consider selling it?
Two reasons: - If my bot makes 5 BTC/month, and i find 5 people willing to pay 1 BTC/month for it. That means 10 BTC/month.
- My bot makes money now, but i can't predict the future. Will it still make it's 5 BTC next month? Next year? Selling can act as an insurance for future profits.
There isn't enough historic data available to crunch. My algorithm relies om the orderbooks of many differed markets on exchanges. Such data is not available in the bitcoin space. I have tried different algorithms that rely only on past trades so that it can be backtested, but only a small percentage of profit (less than 1 BTC all-time) comes from such algorithms. if your bot works using order book then it makes more sense that it makes a fixed volume. (i was expecting something based on trades). honestly, it sounds more to me like you need to do some money management if you want to have a better guarantee of profits (since you already seem to have a strategy). you could probably program that into the bot as well. EDIT: one problem with your calculation might be that 5 people using your bot would be cutting into the volume. if there is only the opportunity for the bot to make 5 BTC per month then 5 people would consume it all, including your share of 5 BTC per month. so i also don't see a reason for you to sell your bot.
|
|
|
|
Darkstone2 (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 26, 2014, 10:47:49 PM |
|
So you don't think that selling this bot in a model similar to CAT will have a positive effect on my total profits?
About running multiple bots. I'm not really sure what would happen. I run multiple bots on most exchanges, but it is quite hard (read: impossible) to find out what the effect on the profits is. The basic idea's behind the bot are not really complex, but the lack of back testing possibility makes it really hard to see what effect the bot has on the market. I can prove that running multiple bots has a positive effect on the total profits (with some simple theories about network latency's and the amount of requests you can submit before you're being rate limited). But beyond that i have no idea.
I really appreciate your feedback, but it doesn't really help me answer the questions i still have. At this point i'm just trying to figure out how many people would be interested into buying a bot. So i can decide if it is worth my time (our time) to create an company around that business model. The technology is good, but are there customers for it?
|
|
|
|
blankia
Newbie
Offline
Activity: 27
Merit: 0
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 27, 2014, 08:53:38 AM |
|
You should definetely give more information about the strategy it is trading on. If it is an arbitrage strategy then selling the bot to multiple people will immeadiatly eliminate any potential profits, so the bot will become useless. Also it sounds like a big scam if you claim you make 5 BTC a month with almost no capital.
It would be better if you claimed the outperformance of the bot. So the bot makes 10%, the market did 5% for that period, hence the outperformance is 5%.
|
|
|
|
daybyter
Legendary
Offline
Activity: 965
Merit: 1000
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 27, 2014, 01:45:16 PM |
|
Thanks for the thread anywhere, since I'm in a similar situation...looking forward to the final answer to your question... ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif)
|
|
|
|
Darkstone2 (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 27, 2014, 04:48:23 PM |
|
You should definetely give more information about the strategy it is trading on. If it is an arbitrage strategy then selling the bot to multiple people will immeadiatly eliminate any potential profits, so the bot will become useless. Also it sounds like a big scam if you claim you make 5 BTC a month with almost no capital. I'm selling a product, not a strategy. What that product does to make profits is not important right now. About it being a scam. I'd rather not go into that now. I have the evidence to support my claims, but that has to wait until i have made a decision. It would be better if you claimed the outperformance of the bot. So the bot makes 10%, the market did 5% for that period, hence the outperformance is 5%.
I'm not sure if outperformance is a word. But i've already commented on this. The performance, relative to the market, is in the million % range because we've started somewhere close the minimum order quantity on some exchanges.
|
|
|
|
Darkstone2 (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 27, 2014, 08:08:18 PM |
|
I've discussed some details with my partner.
On a high level, there are 2 different outcomes: Our customers make more money than they pay is. Our customers make less money than they pay us.
My partner noted that in case of #2, we're actually scamming the customer. Which i agreed with him. Neither of us want to create an company with primarily unhappy customers.
That got me thinking. I realized that i've tested multiple times what the impact is of the amount of capital on the account relative to the profits. Doubling the amount of cash leads to about 50 tot 100% more profits. So it is not a zero sum game... we think. Selling this bot will lead to customers, and customers will lead to more capital, which will rise the total profits making both us and the customers happy.
However on the practical side, we're not sure yet how we can make that happen and what the best business model is for what we seek.
|
|
|
|
xybersurfer
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 72
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 27, 2014, 08:50:52 PM |
|
I've discussed some details with my partner.
On a high level, there are 2 different outcomes: Our customers make more money than they pay is. Our customers make less money than they pay us.
My partner noted that in case of #2, we're actually scamming the customer. Which i agreed with him. Neither of us want to create an company with primarily unhappy customers.
That got me thinking. I realized that i've tested multiple times what the impact is of the amount of capital on the account relative to the profits. Doubling the amount of cash leads to about 50 tot 100% more profits. So it is not a zero sum game... we think. Selling this bot will lead to customers, and customers will lead to more capital, which will rise the total profits making both us and the customers happy.
However on the practical side, we're not sure yet how we can make that happen and what the best business model is for what we seek.
if the percentage profit is so high, then why bother with customers taking part of it for a fixed fee? it just doesn't make sense to me
|
|
|
|
Darkstone2 (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
May 28, 2014, 09:14:02 PM |
|
A part of that is risk management. One of our (not so transparant) exchanges has around 8 BTC on it. sometimes i get pretty concerned that one of those exchanges shuts down. Aside from that, i still think there's an high chance that simultaneously selling the and running the bot yields more profits than only running the bot.
|
|
|
|
Darkstone2 (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
June 03, 2014, 08:31:27 PM |
|
Slight update. I think that this is.. not going to happen. For multiple reasons, including taxes, time spent on customers that could've been better spent on a job, and our doubts to gain profits and keep all customers happy at the same time. Mostly the last one.
Our bot is still making nice profits, perhaps in a year or so i will re-visit this decision.
|
|
|
|
|