Honing a strategy

For the first time since I started dabbling with Forex, I've got a strategy that seems to be working, based on Udine's 00 Level Trading.

I've refined it over the last 6 months but for the last 3 months it's been bringing in consistent pips.


Udine, if everything goes to plan I owe you more coffee than you can ever drink.

Thanks also go to Nasap, Masterrmind, Luxing, Robinho, Yelena and many others who have contributed to the thread and also Hanover.

But, going back to my strategy, there are two variants. Over the last 3 months and they performing as follows: -

1. variant 1 is averaging about 98 pips/day with a win rate of about 75% averaging 20 trades/day.
2. variant 2 is averaging about 134 pips/day but only has a win rate of about 66% averaging about 32 trades/day.

The only difference between the two variants is that there is an additional filter on variant 1 which, as well as filtering out a lot of bad trades also filters out some good trades.

Both variants have a R:R of 1:1 (less spread - there is no commission as I'm UK based and using a spread betting account) and I use the SleepTight ea to lock in some pips.

My gut feeling says that the 130 pips/day is better as it's generating more pips...and more pips is good.

But I can't help feeling that, over the long term, the variant with a higher win rate might be the safer way to go (less draw down maybe).

Does anyone have any advice as to which way to go?

My advice is to go with the higher win rate. The difference in the pips will not matter to you once you are sitting on your yacht enjoying the life that successful Forex trading has brought your way!

If you really want to make up for the difference increase your trading size in variant 1. You will still have your 75% win rate but your increase in trade size will make up for 36 pip difference.

That is exactly what I was hoping someone would say. You can get to the same over $ gains by using the higher win rate but, because you're more confident that each trade will win, you can increase the lots.

Simple.

Sent you a PM.

If unable to spread the risk by trading both (and I can see from the trading frequencies that that might be the case, here), for myself I'd certainly go with the higher profit factor one, given that they each have plenty of trading opportunities. However, if they both have the same R:R ratios, then the higher PF one is by definition also the higher win-rate one, so in effect I agree with Pippington, above.

I think I could trade both.

The trades in variant 1 are a subset of the trades in variant 2 so I think it would mean doubling up on some trades.

I would never have thought of trading both. That's definitely thinking outside the box...I like it!

the one thing that worries me is the old saying "If it's too good to be true then it probably is".

Don't get me wrong, I've worked hard on developing Udine's strategy to suit my requirements. Over the last 6 months I can't remember the last time I went to bed before 12:30am. Not ideal when the alarm goes of at 6:00am.

But it turns out that the actual strategy is extremely simple. The part that has taken the time to get to the bottom of is what levels of specific criteria to use for specific currencies i.e. what's the minimum ATR for a specific pair and what hours to trade specific pairs.

Once you've worked this out it's extremely easy to make money from the Ash changing color.

A few people have asked me what the actual strategy is.

I don't want this to be a thread to be about specific details of the strategy.

Udine, Mastermind, Luxing, fxrabbi, yelena, robinho, nasap (and many more) are giving us gold.

What I did for a while were record details of every single trade for all 28 pairs (AUD, CAD, CHF, EUR, GBP, JPY, NZD, USD) taken when the Ash changed color.

Then, with a bit of analysis, it's quite easy to see the patterns so I started being a bit more selective with the trades: -

1. It just doesn't work for some pairs
   EURGBP and USDCAD are the biggest losers.
2. You get better results when market the currency is based in is open
   For example, European pairs perform better when London is open. How many times has this been stated?
3. There's a minimum ATR
   But once you go past a certain ATR for a specific pair there is no advantage in even higher ATRs.
4. There's a maximum spread
   Similar to the ATR, the is no advantages in really low spreads against spreads just under allowed maximum. Bizarrely, there's actually an anomaly where trades taken when the spread is exactly 1.0 pip drastically under perform against all other successful spreads.
5. You can ignore danger levels
   Big round numbers (BRN) and daily/weekly/monthly pivot points (PP) - there's just as much chance of the price pushing through those levels as there is in bouncing off them...more so in fact. If I stop taking the trades where there is a BRN or any kind of PP or the daily open price between the open price and the TP and the system become unprofitable.
6. Ignore the daily open
   There is no advantage in buys if the day is green or sells if the day is red.
7. Expire after 30 minutes
   If the pending order hasn't been triggered after 30 minutes then it's probably not going to be profitable.
8. Use trailing stops and/or break even
   I'm using SleepTight EA to move the SL to break even once the trade goes into 5 pips profit.
9. Ignore the strength indicators (Giraia etc)
   It sounds counter-intuitive but I've found that the base's strength relative to the quote's strength has absolutely no affect on the outcome of the trade.
10. R:R of 1:1
    For M30 (the only time frame I've used this on) a SL/TP of 20 pips works for me. I suspect there maybe more pips to be gained if the SL/TP is altered depending on the ATR of the pair but that becomes a bit difficult to test when you're using an EA to manage the trades once they're open.

I know some of these criteria go against what others are saying in the other 00 based threads but it just goes to show that there are many ways to skin a cat.

I don't know the system you're referring to (and that doesn't matter), but I'm rather impressed with your list of ten points above.

Eight-and-a-half to nine out of those ten points apply validly to all the routine intraday trading I do, and they seem to me, overall, to be an excellent set of principles.

(If it's of any interest, the only one that doesn't apply to my own trading is the trailing SL mentioned in point 8 above: for myself, I need to see really convincing statistical evidence that a trailing stop confers additional profit, before using one. My experience with them is that they're superficially attractive and appealing but the hard, statistical evidence very rarely bears that out because the "opportunity cost" with them is nearly always higher than one imagines without actually proving their value in any individual system. As it is, I use them for only one of the five systems I trade, in which I have proved that they're beneficial. However, I do sometimes move the SL of my "third lot" to breakeven, after closing the first two lots - but according to strict, objective, statistical parameters which take away the "guessing".)

I suspect you maybe right regarding the break even/trailing stop.

Going through yesterday's trades manually on the charts, it looks like the overall outcome would be about the same. i.e. some that were stopped out due to the break even would have hit TP if they'd been allowed to run but a few would have been 20 pip losses if the break even EA hadn't been in place.

What I'd need to do is start recording the trades without the EA running (as well as recording the trades with the EA running). I could then compare the results.

Sounds like I need to set up a virtual machine, in fact it sounds like I need a few virtual machines to record details of slightly different variations on the system.

1. Fixed SL/TP with break even EA running (as is currently)
2. Fixed SL/TP without breakeven running (as suggested by Xela)
3. Varying SL/TP based on the ATR with break even EA running
4. Varying SL/TP based on the ATR without break even EA running

Microsoft Virtual PC here we come!

I suspect that to answer this question definitively, you're probably going to need to compare the actual, traded, observed, proven outcomes of 300+ consecutive trades with the trailing stop with the projected alternative outcomes of those same trades without the trailing stop. If the difference ends up being very small, I would certainly err on the side of not using the TS, myself.

(I don't and won't use EA's, or trade from Metatrader, myself, so I can't offer any insights at all about that aspect of it, I'm afraid.)

Good luck!

Thanks for starting that thread. Fascinating to read.

I must say that I am also very skeptical about the Giria thingy. Never found it useful. Also skeptical about the "danger levels".

Generally speaking, I am skeptical when everybody seems to agree about something...

What currency strength meter do you use if you don't use giria?

I am using the Giraia but not the original indicator. I "stole" the code and included it in my bot. The problem I had was when I was working out the ratio between the base strength and quote strength. The calculation was correct for buys but wrong for sells, so overall is had a net impact of zero on the overall outcome.

In my sample set there is a very slight advantage if there isn't a danger level between your open price and the TP, but if you filter out them out you end up with very few trades. This could be slewed by the fact I'm running the EA to break even once the trade gets to +10 pips.

Big,
Thanks a lot for sharing your 6 months of effort with all of us. A few will do that. Hope you won't mind answering couple of Qs.
You start trading from London open to close? Do you trade all the pairs except UC and EG? Do you trade every ash color change if the session time is right and the pair has enough range to move? Thanks again!

Thx for sharing Big

Looking forward to it, always great to see others points of view.

#7

#3

Attached Image

The bot runs from 01:00 to 19:30 GMT.

Currently trading EURAUD, CADCHF, EURCAD, GBPUSD, GBPCHF, AUDUSD, AUDCHF, GBPJPY, GBPCAD, USDCAD, EURNZD, NZDUSD, EURCHF, CHFJPY & NZDCHF but this changes. It depends on how the pair is trending on the higher timeframes.

Not bothered about the daily range, I just care about how much it's moving. It the ATR(20) is low then there's not much chance of it moving quickly enough to hit the TP before it changes direction.

A lot of strategies involve a set list of criteria and then some discretion to avoid the traps.

This is strategy is different in that the criteria are constantly changing - which pairs to trade, the best ATR to use etc.

So what works this week might not work next week.

Hi Big,

Don't take this the wrong way but it looks to me like you are simply curve fitting a strategy to very limited data. I don't know much about the strategy but based on what I can gather here it already seems too 'delicate' for actual market conditions with its excessive degrees of freedom.