Flippy method

Indicators for above post.

january 2005 backtest - dec 2005 for flippy with ROC.

Using generic rules (should work to keep minimum drawdown in any market types) :

1. opposite direction trades never exist.
2. on 3 lines formed, enter on support/resistance breakout, only if price is coming from opposite support/resistance direction.
3. on 2 lines formed, enter on pivot(yellow lines) breakout, only if if price come from existing support/resistance direction.
3. fixed initial SL 50 pips. open 2 order at a time, each use half of risk size.
(if risk is 1 %, each position use 0.5 % risk)
4. Every new day, move 1 order sl to trigger line (ie. if entry is from yellow breakout, move to yellow each new day, if trigger is red/blue move to red/blue each new day)
5. the other open order use fixed 1/1 ratio.
6. when opening buy/sell stop order look at rules 1. if there's existing trade that hasn't been closed, do not enter stop order.

Backtest is using alpari history data. using 1% risk each trade. max drawdown recorded in 2005 is around 4 - 5%. Backtest is finished, please see next post for result.

finished backtest finally.

result of manual backtest from jan2005

trades with 1/1 risk reward :
total trades = 226
win = 134
loss = 92
pips gain = 6700
pips loss = 4600
average profit ratio = 1.46

trades with open TP :
win = 86
loss = 140
pips gain = 10982
pips loss = 4600
average profit ratio = 1.25

equity risked each trade = 1%
balance over 2 years = 210%
max drawdown = around 5%

this backtest using 'generic' rules, no adjustment for trending/ranging market. It does miss some of big moves & it caught some whipsaw, but all in all, it manage to keep low drawdown.

I will post some examples of this generic rules entry & exit in chart screenshot later.

sample #1

A. 3 lines. dont enter buy stop here, price doesn't hit blue lines prior red lines breakout.
B. 3 lines. enter 2 sell stop order here, price hit red lines before this blue breakout. Put initial SL 50 pips for both order. put TP 50 for 1 order, the other keep open TP.
C. Both short entry stopped out. we dont move open TP order SL to blue lines, because it has higher value than initial SL.
D. 2 lines. As we have no open order. and price hit blue before breakout. we enter 2 buy stop order @ yellow lines.
E. 1st buy order hit 50 TP.
F. yellow lines value still below initial TP by 1 point, we wont move SL to this value yet
G. yellow lines value is above initial TP. we move open TP to yellow lines.
H. Exit manually at beginning of day, because price has went under yellow lines already.

sample #2 (whipsawed, no need to worry, my backtest include all types of this whipsaw, it filtered some and got hit by some whipsaw. still manage to keep low drawdown)

A. 3 lines. Price has not hit blue lines yet. don't enter buy stop order.
B. 3 lines. Price has hit red lines before. Enter 2 sell stop order.
C. 3 lines. Price has hit blue lines before, enter 2 buy stop order here. Previous short order closed out at buy stop point.
D. 3 lines. Price has hit red lines before. Enter 2 sell stop order. prev long order closed out.

1 short order hit TP 50 pips. next day move SL to blue lines (we enter at blue lines breakout trigger on this short)

E. the other short order stopped out at blue lines.
F. no open orders. 3 lines. 2 Buy stop order hitted here.
G. 3 lines. both long order stopped out. 2 sell stop order hitted here. 1 hit TP, and the other move it SL to blue lines each day (blue line breakout trigger)
H. manually exit 2nd short order. price has opened above blue lines already.

sample #3

Some opened trades are occasionally ignored with some special condition.

A. 2 lines. We have 2 long orders here with yellow lines trigger
B. long order with open TP stopped out at yellow lines. the other with fixed 50 TP was still open.

but watch, this particular 50 TP order doesn't hit SL or TP within 1 day after it was opened. And by obvious H&S pattern broken at shoulder we ignore this 1/1 RR opened order.

C. 3 lines. previous long order is stopped out. enter 2 sell stop order at blue. 1 hit 50 tp.
D. 2nd short order closed out at blue lines.

sample #4

sometime this generic rules dodging whipsaws

A. We have 2 long orders here.
B. 1st long order hit TP @ 50 pips.
C. 2nd long order with open TP stopped out @ -45 pips. survived from loss.
D. no short order opened, price has not hit red lines before.
E. 2 long order opened.
F. 1 long order with open TP stopped out, as price opened below red lines.
G. the other long order hit 50 pips TP. survived from loss too.
H. 3 lines, no open order, entered 2 short here. 1 hit 50 tp, the other chasing the profits.

The reason I use 2 type of order here is to lock breakeven & reducing the 'profit runner' order chance of being stopped out by spikes.

As I mention it before, for experienced traders, you might want to use daily trendline breaks, chart patterns, and candlestick pattern to filter which box breakout you want to take. I'm sure it'll filter a lot of fake breaks. This boxes its just a tool to simplified TP/SL and entries.

This indicator yellow lines is actualy 2 previous day pivot. I use it as initial guide line for entry trigger. Filtered with ROC indicator. thats why 2 types of entry : 2 lines, and 3 lines. 2 lines entry will be triggered if there is strong moves that breaks that pivot line. 3 lines entry exist when ROC value is around ROC equilibrium point, thus why the trigger has to come from opposite support/trendlines, to validate breakout strength in a price consolidation area. (roc hanging around @ equilibrium ~ price consolidating)

haha, just've read the other thread who ask for MM.

Well, this method is implementing all that you guys have taught to me about trading.

1. it cut it losses @ % risk equity value for losing trades.
2. it locked breakeven value.
3. it minimize chance of being spiked out for the winning trade.
4. it don't open 2 opposite trades mostly. because bull & bear wont share a room.
5. it try to detect 'change'. because the only consistent things that happens in market is 'change'. Not a trend nor anything else. Even trend begins @ these 'change' phase.
6. it try to take same direction of current market sentiment in the beginning phase, it doesnt look back historical high low etc etc. When it went wrong, it'll closed out.
7. it run the profit until there's another 'change' happening

Thanks for you all for these priceless lessons.

Hello "dutdot",

thanks again for sharing your work. Look really fine.

Can you explain me how you was using the "ROC" in your backtest.

Thanks


Josef

ROC -> Rate of Change percentage = (Y-Yx)/Yx

"Y" represents the most recent closing price, and Yx represents the closing price a specific number of days ago. So, if the price of a stock closes higher today than it did 10 days ago, the ROC value point will be above the equilibrium.

I use ROC indicator as internal custom indicator inside flippy. Using x = 24 (24 hour ago in H1 charts). so I could compare rate of change on yesterday close & 2 days ago close.

As we see in chart. when market is about to retrace, having a 'ross hook' or enter ranging condition usually we see ROC[24] is hanging around 0 value as prices slowing down. To filter out this condition I choose value between -0.3 and 0.3 as momentum slowdown.

When ROC is between this value, flippy indicator will draw 3 lines to identify 'ranging box' for that day. To validate this box breakout, I want a good momentum price movement. Thats why my rules are only open stop order, if prices has touch the other side of the box when 3 lines exist. I dont want to risk trade on fake breakouts from ranging/consolidation area with low momentum.

On the other hand, when ROC is above or below -0.3 to 0.3 value range, flippy will draw 2 lines only. Any breakout to yellow lines from indicated direction (red line is a possible short breakout, blue is long), is considered as potentially strong movement that could indicate a trend reversal. (imagine, when ROC[24] is flipping from above to below it possibly making an engulfing candle when seen in daily chart right?

Hope that explain things. Cheers.

Thanks, that make it clear

Inserted RoC function to flippy, so you only need 1 indicator now.

warning: you can't use this indicator for IBFX mt4 client yet. IBFX has a "holiday candle" in the beginning of new week, and it ruined neutral lines/roc value. I'm still learning on how to put 'day begin & day end' function. so no matter what mt4 client timezone use, will have same value for neutral lines & roc.

We need EA on this method...

I've been doing some other backtest on different charts. 2 notable results is on GBP/JPY and the other is japan NIKKEI index. It seems this method works better with high volatility (as other trend following method is).

Still doing forward testing by now on GBP/USD, USD/JPY, USD/CAD, EUR/JPY. I dont know if its valid enough to forward test on what I call low correlation pairs, because when I learn about impeccable hedge I've noticed that every spotfx pairs is always correlated each other in some ways. But we'll see how it goes. I'm hoping to hit 200+ trades on forward tests for statistical edge.

I wish I'd discovered this years ago!