In the first article about the Secrets of Algo Trading, we only started to lift the veil…
The behavior of any system is easier to predict if it has entered the realm of extreme, atypical values (in the absence of a general force majeure in the market)!
The artificial intelligence included in EDVI trade strategies also works with virtual losses. In this article, we will explain why virtual losses are not the Holy Grail, but at least a leg of it!
We consider losses to be virtual if they would have occurred systematically but fell within our waiting period (when the trader is “on the fence”). The waiting period lasts until the strategy generates a certain number of consecutive losing signals. After this, it is possible and necessary to enter real trades. The expected value has already started to work in favor of the trader.
Why does this work?
So, if you flip a coin and get 10 tails in a row, what is the most likely next outcome?
Heads? Of course not.
The outcome is always 50/50! This is where the well-known rule applies: THE COIN HAS NO MEMORY.
We are lucky.
Unlike the coin, THE MARKET HAS MEMORY.
If your strategy is based even on a breakout of the previous day’s high (an example of a near-zero strategy) and you get 10 consecutive losses, something important happens.
Market participants start to notice that the price falls after breaking the previous day’s high for 10 times in a row (triggering your virtual stop-losses 10 times in a row). They begin to trade (often with leverage!) this, as they see it, cool and fresh “inefficiency.”
In such abnormal situations, the market becomes “efficient” again and proves to those eager for easy money that there is no such thing here. The price starts to rise again after breaking highs, or at least does so once in the very near future. When a local inefficiency is noticed by the crowd, such inefficiency stops working (its liquidity, rarely large, gets exhausted).
This is why abnormal series are more prone to mean reversion than ordinary series.
