📈 True Profitable Cryptocurrency Trading 💰 - [RESEARCH PT 2.]
I have done another Monte Carlo simulation, this time I have tightened the belt and included the micro-chunks too. So if the strategy is profitable from a tiny sample, then it shows that it’s super adaptive because it can reasonably guess and make profitable trades from a small datasize similarly as it would from a larger datasize.
Of course we have demonstrated in the past article that the more data we add, the more profit we get, but here we see that it’s profitable even with small data, it’s just that with more data it’s even better.
So it’s a very robust strategy, now let’s see what we have now:
The Strategy Overview
Now the strategy itself is for Direct Trading, this means that it is meant for physical buying and selling cryptocurrencies, not CFD’s. If you don’t know what a CFD is, then research it. Most brokers provide you with CFD trading platforms, not direct trading. Which is bad, since it’s a derivative not directly moving and owning coins.
This strategy is for direct buying and selling of cryptocurrencies, so it can be done even through Shapeshift or Changelly basically. Or through exchanges like Bittrex or Poloniex if you want tigher spreads.
This means that there is no SHORT option, so the strategy is basically for just the BUY side. Now the SHORT side could be simulated or if you’d like to trade it through CFD’s, it should just be as profitable as the BUY side.
So we only take half of the profits because of this, since the SHORT side is excluded, but since most exchanges don’t offer CFD’s, it’s more practical, and if you fear that a centralized exchange might run away with your money then, again it can be even done through Shapeshift or Changelly if their spread is not too big, which it shouldn’t.
Strategy
The strategy currently is very simple, it’s no big secret:
- BUY: ForecastN+1 > PriceN
- Close: N+1
So basically we buy when the next datapoint’s forecast is higher than the current actual price, and we close the trade at the next datapoint, and we take whatever we get.
There are no stoplosses or takeprofits, and the SHORT side is excluded, since it’s direct trading, so we have to rely on a liquid market of course to avoid huge drawdowns.
I choose ETH/BTC for now, but it might work later on other markets once I research those too.
This is no secret, this is the elementary strategy, buying and closing per 1 datapoint. It’s my forecasting method that is the actual secret.
TimeFrame
For simplicity I currently use Daily data. I haven’t tested the performance on other type of datasets like Hourly, Monthly or Tick data.
But the calculations already take a lot of time, so for now I’ll stick to daily data. Besides the 1 step trading strategy needs room for the price to move so I doubt it would work for a smaller dataset than that.
Of course the strategy could be for 2 or or more datapoints, so we could be holding a trade for a week or so, but for now it’s daily trade.
So taking the trade at 00:00 UTC and closing it at 23:59 UTC the same day basically.
Taxes & Fees
Yes taxes unfortunately. So this strategy is based on the raw data from Poloniex and Bittrex. But of course there are many other factors that determine profitability.
Certainly if the expectancy is small, then if it’s further decreased by fees and taxes it could become negative.
For example if we physically trade coins through Shapeshift or Changelly, then Sales Tax might apply and of course Income Tax (or Capital Gains) and since it’s direct blockchain trade, the transaction fee has to be considered too.
So all of this has to be factored into the expectancy since most of these expenses are subtracted from the revenue not the profit, so it might make a normally profitable strategy unprofitable, as it draws down the profit into the negative.
Sales Taxes are the worst, because you have to pay 20-30% extra, so if you make 15% revenue, that will actually be a -15% profit. Sorry, thank the Government for that.
So even if we have a profitable strategy in theory, in practice, there is a lot of theft going on with your money, so this factor is very important to be considered.
Now on CFD’s the Sales Tax usually doesn’t apply and the Income Tax is levied on the profit not the revenue so, in that case using CFD’s might be advantageous.
Again, this is not legal advice, and it depends on your jurisdiction, so make sure you research thes e, because it can determine whether you make money in the end or not.
Conclusion
Okay so that’s that, now you know my thought process. Again, we use an elementary strategy, but it might be tweaked, I am not sure if it can be improved, but it may be.
Now if the fees and taxes drag down our profit into the negative then that is still not a lost cause since, if we were to extend the timeframe, we could still be profitable.
Also see this Risk Management Article so you can grasp my the basics behind my trading theory better:
So even with a fucking 30% Sales Tax, you could end up profitable in the end if the volatility outperforms that, which it could, since cryptocoins are very volatile.
So then instead of daily trading, monthly trading might be better. So in 1 month if the coin makes 50-100% movements, then a 30% Sales Tax might pale in comparison, you can still buck a 70% Profit (or 35% after Income Tax).
Or if a more sophisticated trading strategy could be used, say filtering out days when there is low volatility, or other types of filters, like news releases, and so on....
Again all of this has to be tested, and I haven’t had the time yet to test all of this, as you can see it takes a lot of time to test and research a strategy, and I have just been working on the forecasting part, not the actual trading part.
Sources:
https://pixabay.com
