forex-scalping.com

Edges decay when they are crowded. A pattern that works because few notice it stops working when thousands exploit it.

Classic examples exist across markets. Simple moving average crossovers, once profitable, lost much of their edge as they became widely known. Some news reactions that were reliable years ago have changed, as algorithms now process the data faster than any human.

For retail scalpers, the implication is that strategies need periodic re-evaluation. A method that worked last year may produce thinner results now.

Signs of decay include a falling win rate at the same parameters, a rise in average slippage and shrinking average win size.

Adaptation does not mean constant change. It means monitoring metrics, testing small improvements and retiring approaches that no longer pay.

Diversification across a couple of uncorrelated strategies reduces dependence on any one edge.

Which signs of edge decay have shown up in the records here, if any?

I had one that faded slowly enough that I almost missed it. It was a news fade on the first spike after UK data, which worked nicely for me for about two years. Then over roughly six months the average winner went from around 7 pips to 4, while the losers stayed the same size, and the win rate only dipped slightly.

Slippage was the other clue. My entries on the spike started filling a pip or more worse than in my earlier records, which I took to mean faster participants were getting there first and the move I was fading was already done by the time I acted.

I retired it once its rolling 100-trade expectancy went below zero, which was a rule I'd written down beforehand, so I didn't have to argue with myself. The lesson for me was to watch average win size more than win rate, because the win rate made the setup look healthy for months after it had stopped paying.