Crypto Bot Trading: Part 2

If you haven’t seen part 1, please check out that first. Note, I am not a financial expert, these are just all my opinions, please do your own research.

Signals

One of the things you’ll see when setting up a bot is the Deal Start Condition. This is the signal that tells the bot to buy into your deal. There are a ton of signals and they are different depending on which exchange you are using. Binance, for example, has RSI or Relative Strength Index where Bittrex does not. 

If you use too many of these signals or configure them in certain ways, then you will often create a bot that very rarely opens any deals. You have to weigh the risk to reward, just like with any signal we are going to talk about.

Trading View

These are powerful signals that encapsulate other signals within them. It’s Trading View providing a green light at different time periods. You can select Buy or Strong Buy for 15 minutes and if that goes green from TradingView, your bot will open a deal. 

I recommend using these as extra security to stop yourself from buying a big dip. A Trading View 5 minute Buy signal can often stop your bot from buying into a crypto asset that is dropping in value quickly.

RSI

Relative Strength Index is a trailing indicator that shows, in a set period of time, how overbought or oversold an asset likely is. People often consider a score of over 70 to be overbought and a score of 30 to be oversold. You can use these at different timeframes to provide your bot with deal start conditions you might feel comfortable with. For example, you likely never want to start a deal when the RSI is over 70. 

QFL

Quickfingers Luc is a strategy and 3commas turns it into an indicator. QFL looks at recent price history and determines a value that it sets as the “base”. If the price of the asset drops by whatever percent from the base that you set, then it considers that a good entry point and jumps in on a deal. This is great for high volatility coins where, in an hour to a quarter of a day, it can drop more than a few percent.

CQS Scalping

I don’t understand this deal start condition very well, and so can’t speak to it here. I also haven’t been able to find great insight into it in my research. It might be the best signal ever, but I don’t know anything about it.

Manually/API

This is when you have someone coding a signal solution and telling your bots when to start/stop.

Trading View Custom Signal

This is when you’ve set up TradingView in such a way that you have a strategy that you can’t start/stop using the built-in signals that 3commas has. It requires some technical understanding and a paid membership on Trading View.

Open New Trade ASAP

Always starts deals right away. If one closes, it opens another, unless you have other settings that might delay it. 

Backtesting

There are people that have created tools and scripts that work with TradingView to be able to try out different settings and get an understanding of how it would have performed over the previous few days or few weeks. I highly recommend doing this from time to time as it has increased my profitability immensely. 

It doesn’t cost anything to do this, and there are a ton of 3commas backtester scripts available. I am currently using Backtesting 3commas DCA Bot v2 from rouxam

With it, I can set up all of my bot settings and see what kind of return it might have made previously. Of course, the past is not a perfect indicator for the future, but if you put in the settings and check them at different time scales and it says you’d have made consistent profit, then it might be worth copying these settings into a real bot and let it run for a bit and see what happens. 

Deals Per Day

One thing to note about trading bots is that you might have days where the bots do multiple trades in a day and days where the bots do no trades in a day, so no matter how you set things up, you need to think in averages, diversification and risk-mitigation. 

I have had a single bot do half a dozen deals in a day before. Unfortunately, it was only with the initial buy order and as such the profit per deal was quite low, but it is still worth noting that this can happen.

Expected Returns

The expected returns when “investing” in cryptocurrencies varies wildly. I put quotations around investing because it doesn’t really feel like investing as much as gambling or “gambvesting”… 

In August 2021, I saw returns of over 30%. This is not regular. In September, I saw 6%. I would expect to see returns averaging in the 5% – 15% per month range over the course of the year. So for every $1000 invested, I’d expect to see around $50 – $150 per month in profit averaged over the course of the year (or $600 – $1800 per year for every $1000 invested). 

MonthReturn Percent
September 20216.7%
October 202110.4%
November 20218.5%
December 20212.8%

*In December 2021, I was moving most of my funds out of Binance due to Ontario regulations.

You can then withdraw that profit or keep it invested to take advantage of compound interest growth. Some months will be lower, and potentially see no profits (unlikely but possible) and other months might see higher amounts like I experienced in August 2021. 

I don’t have an understanding yet of what can be expected over the course of a year, but I’ve seen and heard a great deal about receiving above 50% per year, which I think most people would be hard pressed to find in the traditional stock market. 

I think potentially, with the right settings on the right crypto currencies at the right time, people can see returns two to ten times that amount, but it depends on a number of factors. 

I typically look to see a return of around 0.2% per day or better as my indicator for a good day, or $2 in net profit (after fees) per $1000 invested per day, which amounts to an effective annual rate of around 100% year over year.  

This means a $1,000 investment would end a year with a value of $2,031. Even if I fall a bit short of that, and end up at $2,000 for the year, I consider that a win. Doubling my money every year sounds good to me!

Figuring out what YOU consider a win is an important part of this process, but be realistic!

Security

2FA

You should set-up some form of 2FA, two-factor authentication with any account you use that has access to your money. This means you’ll need to use your phone or email to receive a code that you then use on top of using a strong password. This reduces the chance of someone gaining access to your account. 

API Settings

Make sure when working with a tool like 3Commas or Cryptohopper that your API credentials don’t have the power to withdraw funds. If they do, this means if someone got your API details, they could drain your exchange funds from your account. You want your bots to trade, not deposit or withdraw. 

Withdrawal of Funds

When you want to move funds from your exchange of choice, they often have options to send it back to your bank account using a wire transfer or SWIFT bank transfer. It can be a bit of a pain to set-up, but you’ll likely only have to do the settings once. Make sure you get it correct! 

Taxes

I don’t know. I just don’t know. You should talk to your accountant/tax professional. I don’t have one of those yet, so I will have to do so at some point, but it really seems to depend on a number of factors. 

From what I’ve seen, the CRA considers it a digital asset, and as such you can treat the gains as capital gains on your personal income tax return. Every buy/sell is a taxable event according to what I’ve read. This makes things complex. I suggest looking at Koinly.io for tax tracking tooling but Crypto.com also has a tax tool that’s free though apparently not as good as Koinly. 

Fun Notes

Funds Deposited, Funds Used: Unfilled Safety Orders

When you have safety orders that don’t have enough funds to fulfill them, and you add more funds to your account. The bots will open safety orders using those funds and you don’t get to pick which deals use the added funds you put in. It’s a bit annoying, but bots will do what they do.

Patience with Red Bags

There is a difference between realized and unrealized losses. Your crypto bot might have made a bad trade. It happens. But if you are patient, sometimes it can turn around. Holding your capital in a bad deal is called a “red bag” and often if you wait, instead of converting your unrealized loss into a realized loss, you can turn around profit.

I had a bot that made what it thought was a good deal, and normally it would have been right but then the market dropped quickly and for weeks it sat there with a thousand dollars stuck in the deal. If I were to close it, I would have lost around two hundred dollars, sometimes more, sometimes less. But I held it for weeks and about two months after the deal opened, it closed for a profit of around ten dollars. Was having a thousand dollars stuck in a deal for two months worth ten dollars? No. But in my mind, it was better to wait and close with a profit than taking the risk to try to close the deal and take my seven-hundred and fifty dollars to try to get back to over one thousand dollars. 

Deciding when to hold a red bag open and be patient with it is a difficult call, and one of the few aspects of trading that can be difficult. Had the asset I was invested in dropped further and further, then I would have potentially lost most, if not all of my investment when closing the deal.

I recommend not having any one bot, or any one trade have access to the entirety of your funds. I had other deals running that made money while these funds were stuck in this deal. 

Manual Actions

Additional Safety Orders

Sometimes, your bot will use all of it’s safety orders, but the value of the asset will continue to go down. At this point, you can manually add funds to the deal which is like adding more safety orders, bringing down the take profit price that it needs to get to. 


When and why you do this is really a personal preference.

I often leave things alone versus adding additional funds. In rare instances where my bot has used up all of it’s safety orders and the price has gone down five percent or more beyond that, then I typically build out a spreadsheet with formulas to determine what the effect would be if I added more money to the trade at different prices and look at the history of the value of the asset to determine if it’s likely to help me exit the deal sooner.

To do this, I take the currently listed buy price, the total amount invested and it gives me how many of the assets I should have. If this is correct, I then move to another line where I put in an amount, the price I’d purchase it for, and have it calculate how much of the crypto I’d get. Then on a third line, I have a calculation that averages the two to show what my new average purchase price would be and on the final line, what the price would have to get to for the deal to close. It sounds more complicated than it is, but it helps me with valuing decisions.

If your goal is to be totally hands off, then doing all of this is unlikely to be necessary. 

Closing a Deal Manually

You notice a deal could make you a ton of profit for the day, and it looks like it’s starting to go the wrong way. You can, if you want, click Close at Market Price to have 3commas immediately sell your holdings and take profit.

I have done this before on a deal where a ton of my money was wrapped up in a deal with a crypto that wasn’t moving up or down after several days. I manually closed the deal, taking no profit or loss after turning off the bot so it wouldn’t make another trade. You can also do this for a deal in the negative, but that means you take an actual loss. Either way, closing a deal manually releases your capital for future deals. 


Please be aware that some crypto currency values move very fast, and the time between clicking close at market price, the close order going in, and the close order being fulfilled can mean you get more or less money than you expect.

Turning On and Off Bots Like a Light Switch

You can turn on and off your bots. If you feel they aren’t doing well, or you don’t want them to open another deal for some reason, you can click on DCA Bot, and see all of your bots. You can then change their status from green to grey. 

Turning a bot off while it has a deal open will not make it sell immediately. It will still complete it’s currently active trade, but it won’t attempt to open another deal if it is disabled. 

When you turn on a bot, it will then start to look for an entry position relating to the settings that it has, so unless it is set to start ASAP, it likely won’t open a deal right away, even if you think it should. 

Editing Bots for Next Time

If you edit a bot while it has an active deal open, it won’t affect the bot currently running. It will only affect the next deal or trade that the bot starts. If you edit a bot that isn’t currently active, then the next deal it takes will use the new settings you’ve saved. 


My recommendation is to be careful editing bots as it can be difficult to know if a strategy is working or failing if you are constantly editing bots. It can be helpful to use paper trading to create different bots to test different strategies and then copy the configuration of your best performing bot. 

Copying Bots

3commas has a tool to discover bots being used by other traders. It’s called Discover a bot. I don’t recommend these bots as they typically are sorted by gains made in the last month, oftentimes they use custom signals from third parties (which can cost money) and if you don’t know how they work, will they truly have long term value for your account?  

Positive Percent, Negative Value

3commas takes into consideration the fees of the original trading platform when determining the current value of the deal. So if you see a time when you have a positive value over your purchase price, but you would still be losing money in the trade, it’s because trading fees come out first. So sometimes you won’t see a profit on the trade until the bot has moved up beyond a few tenths of a percent.

A Typical Bot Set-up

Creating a bot that trades one pair with a long strategy, getting profit in your fiat or stablecoin, and has a reasonable take-profit exit and a good number of safety orders is the way to go. Below are some images from one of my better performing bots that also is focused on being ready for a big correction in the marketplace.

This bot below is focused on a low-fee provider. I wouldn’t recommend a high number of safety orders for a higher fee provider. It would be better to have fewer orders, more spread out in that situation.

The deal start conditions I’m using here are a combination of RSI, relative strength index in a 5 minute increment of a value under 35, and two Trading View general indicators of 5 and 15 minutes saying Buy or Strong Buy. Only when all of these conditions are true will it submit a buy order.

When it submits a buy order, we are looking to exit with 2.25% in profit, but I do have a trailing deviation of 1% which is huge. I don’t recommend this and suggest you not use trailing deviation! Trailing deviation means that once it goes above 2.25%, if it keeps going up, the deal won’t close, but if it drops by 1% at any point, then it’ll exit. This means that sometimes you can follow a big run-up to 5% or if it gets to 2.25% and then drops by a percent, you exit with a profit of 1.25% instead of 2.25%. 

I have my safety order set to $60 initially and the first safety order gets placed when the value of the crypto drops by less than 1%. From there, I have 14 more safety orders (for a total of 15) that grow in size by 12% each time and the distance between each safety order grows by 10% each time. 

The above chart shows that if the crypto dropped by 31.68%, I’d have $1952.77 invested in it, and it would have to go back up 23.84% for me to exit with my 2.25% profit. So you want to focus on the Deviation %, the Required change % and the Total Volume in your quote currency. 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.