Continuous Trading (CT) is a new feature included in latest Gunbot versions. This feature can help Spot Grid DCA strategy to keep trading even when it was supposed to be stuck.
How does it work?
Imagine you are “bagged”: the strategy bought five times in a row already and break even is unreachable for now.
Normally, we would wait until we can get out, or for prices to drop more so we can DCA more.
When continuous trading kicks in after a configurable number of buy orders, the sell target would simply be set above the last buy rate.
If price hits that target then it sells a part of your bag at that rate, which is a guaranteed profit compared to the last buy rate.
If price then drops again, the strategy buys again and the next sell target will be based on this buy order rate.
If after such a sell order, price rises considerably, the strategy will place another buy order in roughly the same size as the last sell order.
In both cases, it would continue with regular grid DCA orders if price drops further.
Continuous trading orders are made with a smaller size than your regular trading limit, to keep a balance between active trading and possible negative effects on the break even price of the overall position. The setting for trading limit must be at least 2x the actual minimum order size of your exchange for continuous trading to work.
Let’s see an example:
When trading moves from normal spotGrid trading to a phase of continuous trading:
– no extra base currency gets invested compared to the last spotGrid DCA order
– the aim is to make small gains in base currency, but also work on slowly lowering the break even price of the overall position (which will not always work, but often does)
When a continuous trading phase ends by reaching the rate for the next DCA order, this can be either at a lower or a higher rate than the last DCA order. Because of this, it is probably a good idea to use a slightly lower trading limit than you normally would with spotGrid, and to expect situations where a bit more DCA trades happen in the same price range then normal spotGrid would fire.
Don’t even try running this on futures: it will lose money because futures positions have a single avg entry price used to calculate PNL for each individual order.
Let’s see some examples of CT in action:
In the first yellow rectagle ww can see a very good move:
1- CT starts in DCA 4, and it sells 50% of DCA 4 amount with profit. Then, using the 50% amount sold, it buys again at slighly better price and sells it with profit again. Once again it buys again at slighly better price and sells it with profit. Finally it buys again that 50%. Summarizing, it made 3 profitable trades and improved DCA 4 purchase as it bought 50% at a better price that the original buy.
2- In the second rectable we can see CT in action again. It sells with profit (only taking latest DCA into account, of course) and buys at simular price again.
Then it keeps doing more profitable trades instead of waiting for the price to correct.
In the pic above, in the first yellow rectangle, we can see CT making a profitable trade. In the second rectangle it buys again using the amount sold (50% of latest DCA), and sells again with profit and buys again with slightly better price. As you can see it keeps DCAing and getting slightly profitable trades that helps lowering the BE point
Until now we have seen positive effects of running CT, but is there any negative effect. Yes, there is, and the orange rectangle you can see in the pic above can help us to explain it. Ok, suppose you buy a coin for 300$, then you buy it for 200$ and finally you buy it for 100$. The average price in this case is 200. Now let’s suppose we sold 50% of of latest purchase for 110. It is a partial profit (taking into account only the 100$ purchase). In this scenario our Break Even point is higher than it was before appying CT. Why? Because now we have a purchase for 300, other one for 200 and 50% purchase for 100. In other words, our BE point is worse because we have got rid of 50% of the coins we purchased with cheapest price.
I made the numbers in one case and it needed 3 CT sells to improve “normal” BE point with no CT. Obviously this could vary, and depending on the trades, it could be 2 or 4 trades the amount needed to “compensate”.
Getting back to the latest image, CT made 2 trades and BE point was slightly higher that the original BE, so it wasn’t able to sell, get rid of the bag, and start a new cycle.
So, under some circumstances CT is better, but in cases like CT making 1 trade and then raising again, could lead to the bot not being able to average down so fast.
In my opinion, overall positive effects are higher than negative effects, so I keep running CT. On another note, it can be configured. By default value is 0.5, and this means that if CT starts, and you set it to 0,5 the bot will use 50% of the latest DCA sell, and in the case you set it to 0.8, it will use 80% of latest DCA.