Hey,
Maybe somebody will be interested in this. Today made experiment with different Portfolio Management method, it uses MA(moving average) on your single systems profit loss balance. If your systems PL drops below that MA(20), you stop trading it until it trades profitably up again. Read somewhere that some traders usis this approach. Tried to see what I would get using this method on my systems portfolio (27 systems) witch runs demo since January.

Results is not very promising, probably better to stay away of it. Hoped to see better performance of this, but still will share with you and maybe some interesting discussion will develop.

Sometimes this method cuts losing systems very quickly, but there is trade off like always, it reduces profits too when systems are ranging or pullbacking to MA It wins some profit factor, less draw dawn, less trades, but losses some returns if winning. The end results are in the attached pics.

blue - real time mt4 trading PL
red - 20 optimized  trades from backtest + real time unseen trades
grey - 20 period MA of red PL line
yellow - PL with applied MA filter

As you can see if red line drops below grey line, system stops trading, and starts again if red line rises above grey line. Sometimes it wins sometimes it losses.

For 20 MA period I used back tested data, which was taken from optimized fsbpro journal, I added it to front of real trades. After 20 trades it started to use unseen real trades. Here can be little performance killer, because most of the time systems real time performance drops from optimized, and you get some MA lag and more losses than if used only unseen data's trades, but I think it should perform better anyway...

P.S -  used fsbpro journal PL data, not real mt4 statement, there is some  unfavorable big differences to clean it out later.