Moving Average Cross Over, Position Sizing

Hello Forexfactory,

What do you think about adding to the position on EVERY Moving Average Cross Over, in the direction of the moving average?
Regardless of whether it is a winner or a looser.

E.g.
1 st Cross Over UP: Long 1 Lot
2nd Cross Over DOWN: Short 2 Lots (All previous positions closed)
3rd Cross Over UP: Long 3 Lots (All previous positions closed)
4th Cross Over Down: Short 4 Lots (All previous positions closed)
5th Cross Over UP: Long 5 Lots (All previous positions closed)

... and so on. Eventually one position will catch a big move.

P.S. This is not Martingale. Because here we do not add on loosing positions. We add irrespective of whether it is a winner or a looser.

Regards,
Xpertize

Interesting, but I wouldn't do it personally. It is not martingale per se, but I wouldn't say it is sound risk management either. It carries similar risk to martingale.

Until when? Why would you increase lot sizes?
If you want to trade a MA cross over strategy then do so. As you say, there's strong evidence that eventually you'll hit a big move, at least if you believe Michael Covel. No need to screw it up with crazy MM...IMHO

Thanks SpyderMan, Will study Micheal Covel and his book(s)

Regards,
Xpertize

Use a ma line as entry price to accumulate more positions when price cross the ma every time, like TWAP algo buy when a up trend is going on higher time frame (H4 or daily).

You will end up with A LOT of money in the market, hedging each direction until you find the swing that pays. That eventual swing net profit is "1" lot since you are countering "yourself". If you simply exit on reverses, and reenter, it is the same thing, but you have less cash sitting in the market. This will save your $$$$ if there is a spike, or divergence. What if you have to keep adding until you are out of money to cover the margin? Then you just lose...all of your money. Exiting and entering on direction changes is the exact same thing without having your entire account at risk.

Ideally though, you will learn how to trade and find the good swings and stop gambling.

But u don't mention what MA & what period u used.
thanks

To make things clearer let's drop the "hedging" view of the set of trades and let's think in term of net exposure. Let's use the notation of a negative lot size to mean a short position. The net exposure is the sum of the (signed) position sizes. At the first cross you are long 1 lot. At the second cross you open 2 lots short, -2 lots, so you're 1 lot net short (1 + -2 = -1). Next cross you're net long 2 lots, then -2, then +3, -3, +4, -4, +5, -5... That's equivalent to stop and reverse with one more lot when returning in the initial direction. The progression is linear in the number of crosses. That's not as frightening as the exponential progression of a martingale.

The problem with the MA crossover is the whipsaw in the ranges. For simplicity, let's say that each cross in this range will make you lose L pips (or an average of L pips). The monetary loss between two crosses is proportionnal L times the absolute net exposure. At the second cross you lose L. At the 3rd cross you lose L more (2L). At the 4th cross you lose 2L more (4L), then 2L again (6L), then 3L and so on.

After 6 crosses you lose 12L and you're net 4 lots. The breakout swing needs to be 3L in size to breakeven.
After 10 crosses you lose 30L and you're net 6 lots. The breakout swing needs to be 5L in size to breakeven.

You need to choose your MAs such that the number of crosses is kept small during long ranges and such that the difference in price between the crosses is as close to zero as possible. The MAs have to adapt to the trend and the volatility of the market. During ranges you need a long period for the long MA and a small period for the fast MA so that the cross happens near of the long MA which position won't change fast. Understand the cross happens roughly at the same price (small L). During the trending market the two periods shall be close together so that you catch as much of the swing as possible. If the trend is volatile the MAs shall be longer than when the trend is straight to avoid getting stopped out a every retracement.

Please also consider Spyderman's post: you need a better exit plan than just "being positive". The profit shall make sense with respect to the drawdowns.

Thanks everyone for the contributions

I think two main things.

1. It's a totally unnecessary, essentially silly distraction from how you're trading, with absolutely no logic behind it at all, and all it will do is randomize your results.

2. It displays that you don't understand how position-sizing should be calculated, by knowing your edge, expectancy, risk management parameters, and understanding the basic statistics that all successful traders need at some point to learn.

Either your system has a genuine, objective, statistically proven, valid edge and a positive expectancy, in which you case all you need to do, to make consistent profits with it, is to calculate appropriate position sizing and risk management parameters (once you've read a book or two and learned how to do that) - or it doesn't, in which case you shouldn't be trading it at all, let alone with variable staking parameters!

(A moving average crossover, used as a trade entry parameter, is not - realistically - going to have a genuine edge or be inherently profitable in itself, but that's another matter, and not what you asked.)