[ScarletPip]

These are my rules for a strategy I'm calling HALO. I came up with the idea while studying HOLO/HiLo although my idea has more to do with following a trend, than with scalping.

You can watch the progress as I test it manually (badly ) here.

http://www.forexfactory.com/clemmo17#acct.35

As I write this I am on day 3 of testing. It's a little early to post such an untested strategy, but I want to get in the habit of sharing my research.

The problem I have is this is a system meant to trade hourly, over the whole day and I don't have the stamina or the free time to do this entirely manually. As a result I made several errors that have reduced the performance. You can spot them as deeper dips in the profit curve. I'm also not able to trade all 3 sessions, which would be my preference, as the win rate is 50/50 so the length that trends are allowed to run is proportional to profit.

Halo
Configurable extern option inputs in green

[01] – General Requirements
=======================================
Chart: Any
Timeframe: Any, but default to H1
Indicators: Bollinger bands (2 or 3 deviations) (optional)

[02] – Entry
=======================================

1. Start at the beginning of the trading day (first hour Asian session). Allow 3 candles to elapse.
2. Wait for a candle to close higher (for a bull candle) or lower (for a bear candle) than the prior candle’s open +/- 10 pips*. Call this the ‘signal candle’. If that candle closes higher - enter long at market at the next candle open. If it closes lower, enter short. See Money Management section for position sizes.
3. If the candle height is less than 20 points (basically a doji) then ignore this candle for the purpose of signaling. It doesn’t exist. Wait for the next candle. Such candles also do not contribute to exit signals.
4. Stop loss (can be real or virtual to hide it from broker criminals) is placed at the signal candle’s high price for a short order, or its low price for a long order +/- 10 pips.
   1. Optional (true/false): Default true. Move to break-even after 20 pips and jump tighter by 12 pips every 10 pips of profit
   2. Optional (true/false): Default false. Use trailing stop of 10 pips. These cannot both be true.
5. Optional multi trade (true/false): Once price is 20 pips away from the entry point, place a new order of ¼ lot size in the direction of the trend. Move the stop loss 2 pips above entry to protect the profit of the first order. Do this every 10 pips.

[03] – Exit
=======================================

1. If price hits the virtual stop, then close the order. If there is time remaining in the candle immediately open a trade in the opposite direction and exit at the candle close.
2. If the order is short AND a candle closes higher than the previous candle’s opening price, exit.
3. If the order is long AND a candle closes lower than the previous candle’s opening price, exit. Note: see my future ideas section below where I suggest that you should give trades going with the dominant trend more leeway. I'm favoring this approach now.
4. 10 minutes before close of that day.
5. 180 minutes before close of the week.
6. After 3 consecutive OR total loss trades, stop trading that symbol for the day. 0 disables this option.
   [6]Optional: defaults false. Partial close ½ of position if price reaches the 2ndOR 3rd deviation band.

[04] – Re-Entry
=======================================

1. If there are no open trades, wait for a signal candle to close higher/lower than the previous candle’s open. (see Entry 2 note)
2. If that happens AND there are more than 5 candles remaining in the trading day then enter long or short according to the signal.

[05] – Money Management
=======================================
Here, my ideas are not original.

• Lot size - fixed
  
• Lot size - percent of balance
  
• RiskPercent
  
• LotsPerDollopofCash
  
• SizeofDollop
  
• UseBalance
  
• UseEquity

[06] – Trading Hours
=======================================
I see this EA trading all week, with no interruptions, but closing all trades with a basket closure at the mid-point of each session, when balance is positive.

• Check broker time, compare to local GMT offset and ensure that the hour count begins on the first hour of the Sydney session.
  
• TradeMonday -defaults true
  
• TradeFriday -defaults true

[07] – Risk Management
=======================================
Again, nothing new here.

• Stop all trades and remove EA if balance closes below SafetyAmount -default -500 units of currency
• Stop all trades and remove EA if balance closes below SafetyPercent - default -5 percent of balance
• Close all trades when profit exceeds ProfitAmount - default 500 units of currency
• Close all trades when profit exceeds ProfitPercent -default 3 percent of balance
  
• MaxSpread
  
• MaxSlippage
• Send email on any event in this group. If I can use something besides MT4 default email function that would be great since it doesn’t work for anyone.
• Send push on any event.
• Send alert on any event.
• MinMargin - stop all trades and remove EA if margin falls below percent

[07] – Miscellaneous
=======================================

• Magic Number: 7462
• Trade Comment: HALO buy/sell/sl/tp

[08] – Flow Diagram
=======================================
https://drive.google.com/file/d/0B7jwC7 ... sp=sharing (draw.io)
https://drive.google.com/open?id=0B7jwC ... i1jQnlyd2c (PDF)


[09] – Ideas for possible inclusion in future versions
=======================================

• If a signal candle moves more than Z% by the ¼, ½, ¾ mark of the current candle, allow EA to enter in that direction before it closes.
• Dashboard, trading multiple currencies/commodities at the same time
• Currency pairs - comma delimited list
• Avoid metals? They don’t seem to be helping. OR
• Trade higher timeframe (H4) OR
• Trade only in the direction of the D1 trend. (commodities) OR
• When trading WITH the D1 trend, allow 3 contrary candles to close before exiting. AND when trading AGAINST D1 require 2 confirming candles to close before entering, UNLESS a trade was just exited due to a trend change, then you can enter without confirmation. <- chart test this.
• Wait x minutes for retracement, then enter.
• Trade-management/basket-closure
• Prevent making the same symbol trade in the same direction after a failure

[ScarletPip]

So today this strategy had a rough drawdown. It is a sign of a major problem.

Of course it happened immediately after I posted all the rules for it.
Actually it's a good thing because it gives me a chance to learn about the market, and to talk about theory.

If you think about the best known strategies and the EAs behind them you will soon realize that even the best among them have a limited 'shelf life'. They work for a while and then they stop working as well. There are at least a couple of likely reasons for this.

• The brokers who are losing money due to the successful strategy devise methods for counteracting it.
• The fundamental inefficiency or pattern in the market that the strategy was exploiting, was discovered and countered. This is one of the basic functions of a market. To correct price inefficiencies.
• The market simply changed over time. This happens frequently.

Maybe there are some other reasons. Please feel free to submit your theories.

Some people like to devise a strategy based on a particular indicator, or based on a repeating pattern that they find in a particular symbol. This is a good method, but it does mean that the strategy may become ineffective once that pattern stops repeating. I like to start by trying to find a fundamental mechanic of the market that isn't going to change at all, or for a very long time.

In this case the 'edge' is that a candle going up (or down) is more likely (better than random chance) to be followed by another candle going up (or down). This is a very trivial insight - almost beneath notice, but the more basic the edge, the more solid it is as a foundation. A better than 50% chance is all we need to create a successful strategy as long as the other parts are working. Of course, if we can do a lot better, closer to 75%, that's even better.

The edge in HOLO (a different strategy that inspired this one) is that a new daily high is likely to be followed by a retracement that you can exploit, as long as price isn't breaking out. It's a scalping strategy so you get your pips and you get out. I like that.

HALO is not a scalper, it's a trend follower, and today, the trend (US dollar strength) reversed and everything went pear-shaped. Almost all my trades went negative and as I started to eliminate them one-by-one (by stop-loss or by the 2 contrary candle rule) the drawdown just got bigger and bigger. This has led me to a few initial conclusions.

• I am not putting enough importance on the dominant trend. It's hardly worth trading countertrend if I'm not going to scalp pips. Better to trade only with the dominant trend, as defined by a higher timeframe.
• If I'm with the dominant trend, I need to be more flexible with my reversal signal. Instead of 2 contrary candles, make it 3 or even 4, maybe, and relax the stop loss distance.
• I should stop trading after 3-5 consecutive losses, depending on how many simultaneous trades I'm making. If I'm losing this many trades in a row, it is a sign of a longer-term reversal. Wait for it to establish itself, and then get back with the trend.
• I'm ignoring candle height. Taller candles are stronger signals, and short candles are a sign of a ranging market.
• The most successful pairs have been EURUSD, USDJPY, and GBPUSD. Probably because they are highly liquid, smooth-trending pairs that don't whipsaw as much as crosses. I should focus on them.
• The least successful pairs were commodities. Probably because they tend to range for long periods and then violently trend for a short time. By the time I enter on the second trending candle, most of the move is finished, and I get whipsawed in the ranges.
• Exiting trades one a time a time is a losing strategy. A few times I saw my position was overall profitable, but left it and it didn't improve with time. It might be better to set a profit target and exit a basket all at once instead of having individual trade rules for exits.
• If we can have basket exit rules why not have basket entry rules? Use CSS to find the best weak/strong pairs (for a day or a 4H period) and enter on a signal.
• To avoid entering on an exhausted trend maybe our entry candle should be one timeframe higher, and be preceded by a contrary candle?

There's enough here to come up with an entirely different strategy and here is one of the hardest parts (for me) of testing out new strategies. Should I throw out the baby with the bath water, and try an entirely different strategy or just tweak this one? Certainly it works well on EURUSD, as I saw when I tested it on a chart, and it works well in live demo. I don't necessarily need this strategy to work on all pairs.

One thing to consider is how a strategy fits with your personality. It is exhausting (for me) mentally and physically to follow a chart, hour by hour, and check conditions manually. I could restrict this strategy to just one session. I could also make the 'check-in' requirements less stringent.
I will probably go ahead and make the new strategy, test it on a chart, and tweak this strategy and continue to test and report the results in this thread.

[Geko]

Hi Clem,

I didn't had time to study your strategy today as I was busy with latest things to be disclosed soon (a few tweak in CDM and a strength stuff in Blast DB), but I will for sure.

Keep on the good work

[ScarletPip]

Thank you Geko.

Well, I don't know if it is good work yet, but it will hopefully lead to something. I've already got some new ideas from the exercise. I wish I had some clones so I could test them all.

I noticed today that all my EAs went into drawdown, even the most conservative ones, so something unusual happened, but I'm not sure what exactly. It didn't seem like an extreme day. It could just be bad luck or something more. The lesson for live accounts is always be cautions around red news (like the Yellen speech). For this strategy I will try to change things slowly so we can see if today was an outlying event or a regular problem.

Looking forward to see what you have in store for CDM and Blast!

[ScarletPip]

I've had a chance to go over my charts and look at what happened this week.

First, let me show you a monthly chart with an early 'minimum viable proposal' - where I outlined, according to the rules of the system, winning trades in gold, and losing trades in crimson, using no trailing stop or tp, and always falling back to the worst-case scenario.



More gold than red? Ok, then we can proceed.

Now let's see what should have happened this week, if I traded this plan according to my system, with the same perfectly mechanical performance and outlook?


About right. 12 losses and 11 wins. The losses are a bit deeper than average. Note that even a basic rule to protect against retracement would improve a system that already looks profitable.

Based on this chart and the previous one I've noticed a couple of things.

1. Most of my loss trades are happening near the start of the day. In the Asian session.
2. Many of my losses are due to trading against the dominant trend based on a weak signal.

It's not hard to theorize the reason for the first failure. The Asian session has less influence on the trend for the day than the other two. Also, the more dominant sessions tend to reverse the direction of the trend that was started.

Second, the entry for this system is not much better than random. We aren't paying attention to candle size or prior patterns. So, let's fix this, while still remaining true to the original spirit of the idea, with a few rule additions/modifications.

1. Don't trade the Asian session
2. Enter only on a candle that is taller than 0.5 ATR.
3. I'm also going to restrict future trading to the top three performing pairs, EURUSD, GBPUSD, and USDJPY.

The reduced hours should (hopefully) allow me to trade this manually without collapsing, and the reduced scope (3 pairs only) should improve my focus. If results are good enough (profit factor over 1.0, with a positive expectancy) over the next couple of weeks we can look at further optimizations. These might include:

• Hold winning trades overnight
• Trade only with the dominant trend (D1 - looking back 7 candles)
• Exit as a basket at a defined profit level or at a defined time
• Relax the stop loss (set at highest/lowest point of the day so far) and the contrary candle rules (4) or eliminate this criteria? If price hasn’t broken the day’s high/low it doesn’t matter?
• Close only part of a trade on a sequence of contrary candles
• Close part of a trade if retrace is more than 75% of entry or signal candle
• Open new trades on a sequence of confirming candles
• Don’t trade the same symbol in the same direction after a failure.
• Move to break-even
  
  Some of these I think I will retain for a new system, but others might be worth trying out in December.
  
  Next - I'll take a look at my 'idealized' trading and actual trading performance and see how they (didn't) align.

[ScarletPip]

Looks like the spoilers aren't showing so here are the images.

Monthly chart pre-test


Weekly chart post-test

[ScarletPip]

Comparing theoretical performance with actual performance. Compare these trades with the second chart in post #6.
EURUSD

Summary: I scored 7 out of a possible 23 (including a pity bonus point!) for EURUSD. Pathetic human! No wonder robots are going to take over the world. The good news is that this means my system might be more viable than my FF pip total indicates, and the bad news is I suck at manual trading.

Feel free to skip the rest of this post unless you are curious about the small details of the system. Or the fallibility of Man.

Starting with Monday, Nov. 14 - there were 4 trades signalled for EURUSD.

1. I entered trade 1 exactly as I should have, and exited early on what was a clear retrace, outperforming my chart.
   http://www.forexfactory.com/clemmo17#ac ... t.26514349
   Notice how different FF's entry candle (4th candle) looks from the one on my chart!
2. I miss the rest of the trades for EURUSD this day, one of which would have included a sizeable win. Probably couldn't keep my eyes open. This puts me at 1/4 effectiveness for Monday.

Tuesday Nov. 15
A difficult day for the system. Look at how small is the entry candle. ATR is about 140 points and this candle is less than 40.

1. First trade of the day is positive pips but it's a mistake according to the first version of my system! I should have waited for this candle to close below the prior candle's opening. Discipline -1.
   http://www.forexfactory.com/clemmo17#ac ... t.26531055
2. Did I catch the reversal opportunity when price broke my stop-loss (the signal candle's high)?
   http://www.forexfactory.com/clemmo17#ac ... t.26534267
   Indeed I did, but I broke another rule when I didn't sell at the close of that candle. This breach of protocol was positive for my pips though. In fact the rule that requires closing with the close of a reversal is one that I'm waffling about. It seems to help in some cases, and in others it doesn't.
3. The next trade is a pyramiding on the previous one. This is permitted by my system but it is devilishly difficult to keep track of these trades when trading multiple pairs, at least without an EA to help. At least for me.
   http://www.forexfactory.com/clemmo17#ac ... t.26534614
4. The next trade is a loss and is perfectly in accordance with my system. It's hard to know when a trend has ended.
   http://www.forexfactory.com/clemmo17#ac ... t.26535270
5. Another win, and in accord with the system rules although it looks like I entered this a little early, to my benefit.
   http://www.forexfactory.com/clemmo17#ac ... t.26537063
   I'll give myself 2/4 with a bonus point for the pyramid trade for Tuesday.

Wednesday, Nov. 16

1. I miss the first trade opportunity. Was I napping after being up late all night? Saved me some pips though.
2. Second trade is no better. I miss the second trade and...
3. I miss the big reversal opportunity, entering very late in this candle for a few measly pips. Oh noes!
4. I miss the third trade signal. Am I even following the plan anymore?
5. I miss the fourth signal, miss the reversal, and end up placing a winning trade based on I don't know what.
   0//7 and 200 points from Gryffindor for this embarrasment!

Thursday Nov. 17
Will I get back to the trading plan today? Let's find out.

1. Trade 1 is a sell order, and a loss, as expected, but I enter a candle late and exit a candle late. Did I dismiss the first signal because it was too close to the opening of the prior candle? I do give myself a pip-buffer for close entries. I should probably remove that now that I enforce a minimum signal candle height. http://www.forexfactory.com/clemmo17#ac ... t.26565193
2. It looks like that is what I was thinking as I placed a reversal trade for some pips gained. Fair enough. http://www.forexfactory.com/clemmo17#ac ... t.26567906
3. Third trade is according to plan but I enter late and it costs me some pips.
4. Aaand where is the reversal trade?! Nowhere to be found. Passed out from exhaustion, poor baby. This missed opportunity and the next one cost me enough pips to undo all the illicit pips I gained from bucking the system. The lesson is - trade the system!

No trades placed on Friday, as I was already at my endurance limit. As you can see a frail organic lifeform like myself NEEDS strong robots to survive in this world of Forex.

[ScarletPip]

GBPUSD

Here is the same type of theoretical scenario weekly chart but for GBPUSD. Again, using only the worst-case scenario for each trade with every signalled opportunity; gold is a win and crimson is a loss.


Obviously chances for success are lower than with EURUSD. How did I do?

Summary: Overall very poor tracking of the original plan.

Monday

1. http://www.forexfactory.com/clemmo17#ac ... t.26514350
   First trade is a loss, as predicted. Note that I wasn't trading the right lot size and lost more than I should have on this trade.
2. Reversal is a win. http://www.forexfactory.com/clemmo17#ac ... t.26514957
3. Next trade is a nice big win, outperforming the chart because I anticipated the retrace. This isn't technically in accordance with my system which currently has a crude method of dealing with retraces.
   http://www.forexfactory.com/clemmo17#ac ... t.26515688
4. I miss the last trade opportunity. So, 2/4 for trade discipline.

Tuesday

1. http://www.forexfactory.com/clemmo17#ac ... t.26531013
   First a winning trade, and by the book.
2. However I miss the remaining trades for this day. Dodging a bullet in 2/3 cases.

Wednesday

1. First trade is a loss. http://www.forexfactory.com/clemmo17#ac ... t.26547246 Which I hold for too long but probably end up losing about what I'd have lost if I'd followed the plan.
2. Second trade (third on the chart) is a win that I enter and exit early to my benefit. http://www.forexfactory.com/clemmo17#ac ... t.26551988
3. The next trade is a mystery. Its a loss like the one on the chart, but I don't understand my entry or exit choice. http://www.forexfactory.com/clemmo17#ac ... t.26556013
4. Another mystery trade. http://www.forexfactory.com/clemmo17#ac ... t.26558849
5. A third trade that doesn't comply with system rules. http://www.forexfactory.com/clemmo17#ac ... t.26556013 No idea what I was thinking here. Addled by fatigue?

Thursday

1. http://www.forexfactory.com/clemmo17#ac ... t.26564343 As per chart.
2. The reverse trade. http://www.forexfactory.com/clemmo17#ac ... t.26567854
3. Nothing else for the rest of the week.

[ScarletPip]

Last comparison - using USDJPY.
USDJPY


Summary: Closer, but still lots of room for improvement.

Monday
Epic fail right off the bat, as the best trade of the week is exited early (WHY??) I think this must have been a fat fingers mistake.
http://www.forexfactory.com/clemmo17#ac ... t.26531038

Tuesday
First trade of the day is not allowed, although it is profitable. Did I dismiss the fourth candle as being too short?
http://www.forexfactory.com/clemmo17#ac ... t.26531038
I don't see the second trade of the day, but this might be the third.
http://www.forexfactory.com/clemmo17#ac ... t.26535319
No other trades are taken which means I missed another sweet end of day opportunity.

Wednesday
First trade of the day is bypassed, maybe because the candles are too short. Fair enough
Second trade is hit! http://www.forexfactory.com/clemmo17#ac ... t.26535319
No other trades today.

Thursday
It looks like I missed the first entry so try to make it up with a late entry.
Failed. http://www.forexfactory.com/clemmo17#ac ... t.26535319
Third entry is timed right but I don't exit, hanging on to turn it into the profitable 4th trade. Things like this make me want to rethink the 2 bar retrace rule.
http://www.forexfactory.com/clemmo17#ac ... t.26570255

No trades Friday.

[ScarletPip]

Updated system rules are here, and I will post new versions as they are modified.
https://docs.google.com/document/d/12AF ... Uo1CY/edit