More pairs, more profits!!!!

There is a belief in some users that if their pairlist contains more pairs, it would bring more profits. A simple analogy is that if you have $10,000 balance, trading 200 pairs with $50 stake each will bring more profits than trading 10 pairs with $1,000 stake each. Is it true though?

Low volume pairs

While it’s true that more pairs would bring more trade opportunities in theory, in practice most of them are low volume pairs. Read this to know why they are bad. If you are trading those low volume pairs, you are most likely to exit in loss than in profit. Remember that in backtest, FT will enter and exit trades successfully if the prices are reached despite the stake amount of those trades and the actual volume data of that coin. You can get the bot to sell 1000 BTC when the data shows only 0.01 BTC got traded in that candle.

One-for-all strategy

Based of my experience so far, I don’t believe you can have a strategy with global logics that can works for many pairs and give you a (relatively) big profit. Read this for explanation of why I don’t think such strategy exist. It’s already hard to predict the movement of one coin. Imagine having to do it for 200 coins, and have to correctly predict their movements frequently. If you want to try having such strategy, start small. Maybe 10 pairs. Then see how the strategy goes long-term. Then add few more pairs. See how they goes again.

If you still want to have one-for-all strategy, what you can do is try to have a very conservative logics that only enter trades on (what the past data says as) the best entries. It still won’t guarantee long-term profit, but in my experience, it lowers the probability of long-term big losses.

Be careful of missed entries in backtest

If you want to have such strategy, make sure to set unlimited max open trades and a very big wallet size for backtest. You want to make sure your strategy can work profitable for all of the pairs, which means you don’t want the backtest to ignore entry signals just because of full trading slots. If such cases happened a lot, that means the backtest don’t really test the strategy on all pairs, and the result must be taken with a big grain of doubt.