Bytebodger's Pip-Burning Journal

Quote

Disliked

This single factor will have a definite effect on your results

I'm a bit thick (see : a lot) , could you clarify the above statement...

Cheers

Rye

I don't want to speak for BB, but I understand his statement to mean that my profitability will be greatly improved if I am more aggressive in moving my SL to BE. In other words, anytime you have a position move into reasonably-positive territory, lock in a BE SL.

Of course i knew that really.....not

Thanks Bb

ya might have to drag me along kicking and screaming thru these conversations ...but i am enjoying the ride

Thanks all !!!

Rye

In the past few months, I have had a rather tempestuous relationship with trailing stops. At times, I have convinced myself that they are massively effective. At other times, I have determined that they are utterly useless. They sound so alluring - progressively lock in profits as the position goes in your favor and then gracefully stop out when the position turns against you. But the reality is usually somewhat different.

The biggest problem with the trailing stop is the price spike. I don't necessarily mean one of those lightning-fast news spikes (although these certainly qualify). A price spike could take minutes or even hours. But when you have a trailing stop in place, the spike effectively acts to take you out of a position that otherwise may have been profitable. For this reason, I was recently back in my "hate them" mindset.

All day today I've been rehashing my performance over the last couple of weeks. My performance hasn't been bad, but it hasn't been good either. I'm just treading water. At some times, this can be acceptable. After all, "treading water" means that you are still in the game for tomorrow's trades. In exceptionally bad market conditions, "treading water" can be a victory unto itself. That being said, I have been mostly placing fozzy trades and the following link really got me to thinking:

http://forexfactory.com/showpost.php...&postcount=278

billser only started this adventure a few weeks before me, but his performance is far superior. Furthermore, as I delved deeper into his trades, I could not find any magic key that defined the difference between him and me. He is using a slightly different set of indicators, but it's not as though he's hitting all the winning trades and I'm hitting all the losers. What's more striking is the fact that he has a much higher winning percentage.

I had mentioned a few posts earlier that my winning percentage seems to be unnaturally low. It may be too early to make any hard conclusions about my method, but I'm more than a little surprised to see 10 losses balanced against only a handful of wins, especially when other fozzy traders are experiencing a much better win rate.

But there is one key difference that seems to be separating me from others - the fact that I don't currently employ any trailing stops.

Based on BB's comment above, I went back and decided to see what would have happened with my 14 trades if I had quickly moved my SL to BE whenever possible. The result is that I would have avoided 3 losing trades and I would have saved ~125 pips. Then I went back and tried to figure out what would have happened if I used a ATR(50)x0.7 TRAILING stop. The result is that I should have booked roughly ~300 more pips.

What really struck me is this: My winners generally tended to take off very fast and didn't reverse much until they were ready to totally turn against me. But my losers often languished in moderate territory, sometimes for days, before ultimately turning against me. My GBPUSD trade is a great example. It had been up by as much as +30 pips at various times, but those times were always brief. Today it spiked to its high before reversing hard and finally stopping me out. Since I relied on a manual move of the SL to BE, that move never happened and I am left with my full loss of 70 pips. But if I had employed a trailing stop, I would only have lost ~20 pips on the trade.

Another key factor with trailing stops has to do with the 24-hour nature of the market. The opening of the European session happens while I sleep, and I have already experienced several scenarios where my position peaks into positive territory while I sleep, but by the time I manually peruse my trades, they are already stopped out or strongly into negative territory. Often, by late morning, those positions have been completely stopped out. So I never even had any chance to manually move my SL to BE if I expected to get any sleep.

So for these reasons, I'm strongly considering the redeployment of the trailing stop. But the key difference between this and previous attempts is that my trailing stop has to be wide - I believe I will stick with the ATR(50)x0.7 metric. I have even considered using the full ATR(50) value, but I'm not going to go that far yet.

Of course, I realize that there is no such thing as a free lunch. If I employ a trailing stop - of any width - there WILL be times when the trailing stop takes out what otherwise would have been a more profitable position. But the more I look at it, I really believe that a trailing stop has a place in my arsenal.

Trade update (new trades in bold):

(F) Sell 1 USDCAD @ 1.0895, SL @ 1.0787
(F) Buy 1 GBPUSD @ 1.9866, SL @ 1.9792 = stopped out @ 1.9792 (-70)
(F) Sell 1 AUDNZD @ 1.1271, SL @ 1.1311
(F) Buy 1 EURAUD @ 1.6438, SL @ 1.6367
(F) Buy 1 GBPJPY @ 241.40, SL @ 240.37
(A) Sell 1 AUDJPY @ 99.60, SL @ 100.15
(A) Sell 1 EURCHF @ 1.6481, SL @ 1.6516
(A) Sell 1 EURJPY @ 163.60, SL @ 164.25
(F) Buy 1 EURUSD @ 1.3449, SL @ 1.3402
(F) Sell 1 GBPCHF @ 2.4277, SL @ 2.4363
(F) Sell 1 USDJPY @ 121.63, SL @ 122.14


Fozzy Trades
Running Total: -41
Winning Trades: 3
Losing Trades: 7

ADX Trades
Running Total: -4
Winning Trades: 1
Losing Trades: 3

All Trades
Running Total: -45
Winning Trades: 4
Losing Trades: 10

Exactly............thank you Bb!

Words of wisdom.............I really enjoy reading your posts!

Trade update:

(F) Sell 1 USDCAD @ 1.0895, SL @ 1.0787
(F) Sell 1 AUDNZD @ 1.1271, SL @ 1.1311
(F) Buy 1 EURAUD @ 1.6438, SL @ 1.6367
(F) Buy 1 GBPJPY @ 241.40, SL @ 240.37 = stopped out @ 240.44 (-96)
(A) Sell 1 AUDJPY @ 99.60, SL @ 100.15
(A) Sell 1 EURCHF @ 1.6481, SL @ 1.6516
(A) Sell 1 EURJPY @ 163.60, SL @ 164.25
(F) Buy 1 EURUSD @ 1.3449, SL @ 1.3402
(F) Sell 1 GBPCHF @ 2.4277, SL @ 2.4363
(F) Sell 1 USDJPY @ 121.63, SL @ 122.14


Fozzy Trades
Running Total: -137
Winning Trades: 3
Losing Trades: 8

ADX Trades
Running Total: -4
Winning Trades: 1
Losing Trades: 3

All Trades
Running Total: -141
Winning Trades: 4
Losing Trades: 11


Tomorrow, and subsequent days, should be interesting. We will begin to see if my revised SL strategy bears any fruit. This GBPJPY trade really doesn't apply, since it was already down by 60+ pips when I decided to change the SLs.

Trade update:

(F) Sell 1 USDCAD @ 1.0895, SL @ 1.0787
(F) Sell 1 AUDNZD @ 1.1271, SL @ 1.1311
(F) Buy 1 EURAUD @ 1.6438, SL @ 1.6367
(A) Sell 1 AUDJPY @ 99.60, SL @ 100.15
(A) Sell 1 EURCHF @ 1.6481, SL @ 1.6516
(A) Sell 1 EURJPY @ 163.60, SL @ 164.25
(F) Buy 1 EURUSD @ 1.3449, SL @ 1.3402 = stopped out @ 1.3410 (-39)
(F) Sell 1 GBPCHF @ 2.4277, SL @ 2.4363
(F) Sell 1 USDJPY @ 121.63, SL @ 122.14


Fozzy Trades
Running Total: -176
Winning Trades: 3
Losing Trades: 9

ADX Trades
Running Total: -4
Winning Trades: 1
Losing Trades: 3

All Trades
Running Total: -180
Winning Trades: 4
Losing Trades: 12




That EURUSD trade was a turd almost from the time I opened it. Holding my breath for the ADP nonfarm payroll report...

Trade update:

(F) Sell 1 USDCAD @ 1.0895, SL @ 1.0787 = stopped out @ 1.0737 (+158)
(F) Sell 1 AUDNZD @ 1.1271, SL @ 1.1285
(F) Buy 1 EURAUD @ 1.6438, SL @ 1.6367 = stopped out @ 1.6383 (-55)
(A) Sell 1 AUDJPY @ 99.60, SL @ 100.15 = stopped out @ 99.75 (-15)
(A) Sell 1 EURCHF @ 1.6481, SL @ 1.6487
(A) Sell 1 EURJPY @ 163.60, SL @ 163.73
(F) Sell 1 GBPCHF @ 2.4277, SL @ 2.4293
(F) Sell 1 USDJPY @ 121.63, SL @ 121.80


Fozzy Trades
Running Total: -73
Winning Trades: 4
Losing Trades: 10

ADX Trades
Running Total: -19
Winning Trades: 1
Losing Trades: 4

All Trades
Running Total: -92
Winning Trades: 5
Losing Trades: 14

Last night was the first time that I put trailing stops in place on my orders. I'm still not very happy with my winning percentage, but at least some of my losses are getting cut short. EURUSD was originally opened with a 48-pip SL, but was stopped out at -39. AUDJPY had an initial SL of 60 pips, but was stopped out at -15. EURAUD had a 73-pip SL, but was stopped out at -55.

I have 5 remaining trades open. They are all, for the moment, in positive territory. Due to the rise in these positions, even if all 5 immediately turn against me and stop out, I will only suffer a combined loss of 65 pips.

Of course, what I really need now are some WINNERS, but in the interim I will settle for smaller losers.

Byte, I think I remember reading elsewhere in your journal that you are being very mechanical about your trades, and that you are taking all crosses all the time. I think this is in general a good idea, as it takes overanalysis out, emotion out, etc.

However, in your search for a good system that fits you (as I am in the same search), consider some price analysis, and that might help filter out some of your losers (you said you have had quite a few). Granted, an overanalysis can filter out winners, which is what we don't want. But on the EURUSD long that you took, if you look at the bar that confirmed the long, it's a long nosed pinbar (see the james16 chart thread, but you probably have already, not sure) that got in the vicinity of the top of the DOWN channel that's screaming "the sellers dominated this day and the buyers just didn't have the power". Now in a 100% mechanical system, that wouldn't matter. And it's TOTALLY possible that this could have just shot up the next day. But the odds were strong that it was not going to. Hindsight is 20/20, but even if this had gone up 200 pips today, I would still say that the setup was not good for a long. So while I know you want to be 100% mechanical, just the pinbars alone might save you from some big losing trades. Also, these bars have the double benefit of getting you in early on a trade. Say the RSI is crossing down towards the MA, but it's not quite there yet. The daily bar closes with a looong upper shadow (pinbar) and it's near a swing high. I'm going to jump in that trade regardless, but a fozzy cross would make me trade double lots there perhaps! You get the idea. In fact, look at the pinbars on yesterday's daily EURAUD, EURUSD, and GBPUSD, at their position in the trend, and the results today! Combine a little filtering like this with fozzy and you'll be unstoppable! And then you can show me what the hell I'm doing wrong and how I can start making money! :-) (yesterday I got in on the EURAUD and GBPUSD shorts and pulled in about $300 on my demo, but I have many losers that I'm trying to get rid of as well!!!)

This is a great commentary from JD!

Another suggestion: do some backtesting on your "mechanical" Fozzy system with an eye for identifying why some trades lose. Make a list, currency pair by pair. I think you will see a "pattern" that can be used to add a filter or two to your mechanical system to weed out some of the losers. Trust me: there are a few nuances to this system that can be learned by testing and experience!

I have just recently begun reading a book called Way of the Turtle. It's written by Curtis Faith - the most successful of the wildly successful group known as the Turtle Traders.

First, if you don't know who the Turtle Traders were or what they did, go out and Google it. It's an extremely informative lesson for what most of us are trying to achieve in forex.

Second, if you don't yet have this book - GO OUT AND BUY IT RIGHT NOW. I can't begin to describe to you here how applicable this book is to our trading efforts. I would be surprised if there is ANYONE out there who will not get something useful from this book. Most people will learn a ton from it.

It's amazing because on one hand, it tackles some of the nuts-and-bolts type of stuff that we attempt to cover in these forums - stuff regarding position sizing and drawdown and backtesting etc. But it also provides invaluable information regarding the psychology of trading.

Whether you are just now placing your first demo trades or you are an experienced trader, I really think you will be doing yourself a huge favor by reading this book.

In keeping with my previous post regarding Way of the Turtle, I wanted to share something incredible that I read this weekend that really hit home. Curtis Faith, a trader who has attained almost mythical status as the most successful member of the Turtle Traders, states very clearly in his book that trading is not easy. This may sound like a rather basic observation, but it struck me because he also makes it clear that trading is inherently simple. He makes it very clear that trading success does not depend upon any complex set of secret formulas. Rather, successful trading is typically based on a rather simplistic approach - but that does not mean that it is easy.

You see, when I first started to research forex, I was absolutely convinced that profitability could be attained through a simple approach - I wasn't sure exactly what that approach should be, but I knew that it should be simple. Curtis Faith's words have confirmed my belief. However, I was also confusing simple with easy. I felt that if I could just find the proper, simple approach, trading success would come easily. But that has not happened, because being simple does not necessarily equate to being easy.

The analogy I would draw comes from chess. Chess is, at its core, a tremendously simple game. There are only 6 unique pieces on the board (pawn, rook, knight, bishop, queen, and king) and each piece has only a few basic rules that govern its movement. You can take a complete chess newbie and have them playing the basic game in 30 minutes or less. In a few hours, you can illustrate all of the major tactical and strategic motifs (castle early, control the center of the board, develop your knights and bishops as fast as possible, etc.). But this does not mean that it is easy to play chess well. Far from it.

Anyone who's played chess enough can tell you that it can take a lifetime to master and even achieving proficiency amongst other regular players can take a ton of hard work and practice. I think that trading is very much like this.

What if you trade opposite to your signals ?

Maybe you will be opposite and successful

Um, you ARE joking, right??

I know that everyone has been worried sick about when I would resume posting in my trade journal, so here goes...

The past several weeks have been quite a learning experience. As witnessed in previous posts, I have been attempting to cultivate a mechanical trading style that will keep me actively trading and yet avoid the pitfalls of discretion. However, recent events left me in quite a quandary. My fozzy trades were not working. More perplexing, those fozzy trades had a horrible winning percentage while others in the fozzy forum were claiming to have much greater success. Posts by BeachBum and others recommended analyzing factors such as price action, but this threatened to bring me back to a discretionary road block.

Long story short, I have learned a LOT over the last couple of weeks. I went back and re-evaluated my fozzy approach. After completing a large load of backtesting, I have come to some startling conclusions about the fozzy method - and about most methods in general. I used to believe that a good method accompanied by a less-than-ideal SL/exit strategy would still yield decent results. However, backtesting has now shown me that most methods are only as good as your SL/TP strategy. In other words, you can take many sub-optimal systems and squeeze a few pips out of them with a superior SL/TP strategy. Conversely, you can actually take a very profitable system and LOSE A LOT OF MONEY if you are employing a poor SL/TP strategy.

This is very challenging because a system of SLs and TPs that works wonderfully on one system may perform quite poorly in another. In other words, there is no universal SL/TP strategy that can/should be applied across all strategies. Rather, the only way to know which SL/TP strategy is ideal for a given system is to do your own exhaustive backtesting.

This makes a lot of sense to me now (everything makes sense in hindsight!). I have seen systems posted by other members who claimed to be making X amount of money/pips. When I looked at these systems I often thought, "That system is crap. There's no way they could be making money with it." I now believe that some of those posters may be sincere in their claims. It's just that I didn't understand their process of SLs/TPs.

You see, so many of the systems/methods posted in this forum show you screenshots and indicators and say, "Open a trade when condition X occurs." Some of the more "advanced" ones actually go so far as to say, "Close a trade when condition Y occurs." But very few of them actually go so far as to say, "Set your SL at X and set your 1st TP at Y and close the trade at Z and trade Q number of lots based on your account size." And if you don't have ALL of that stuff figured out, you may very well be turning an otherwise-profitable system into a money pit.

In my recent adventures, I have really had my eyes opened with regard to manual backtesting. Like so many people here, I have spent many, MANY, MANY hours over the last 4-5 months manually backtesting countless systems. You know the drill: System X says to trade any time Indicator Y and Indicator Z display condition C. So you pull up the charts for a given currency pair, roll it back to some distant point in the past, and start writing down every time when you ostensibly would have made a trade under this system, along with the supposed result of that trade. But there can be a major flaw in this process.

As I have stated in my previous post, the relative success/failure of so many systems can be determined by the specific application of SLs, TPs, and position sizing. But attempting to figure out where/when a particular SL or TP level would have been hit can be extremely time consuming. In some cases, it can be nearly impossible to discern this merely by looking at the chart.

For example, the system says to open a trade at the beginning of the day and you are going to use a 30-pip SL and a 50-pip TP. The day's bar/candle has a low that is lower than 30 pips and a high that is higher than 50 pips. So what happened? Did the hypothetical trade get stopped out with a 30-pip loss? Or did it hit the 50-pip TP? The only way to know is to drop down to a lower time frame - in some cases, going down as far as the minute-level or tick-level data. But if you're looking at daily charts from 2, 3, 4 years ago or more, you may not HAVE the minute data to look at. What do you do then? And what if you're employing a trailing stop? How in the heck do you figure out where that would have been triggered? Like so many others, I have tried to look at these charts and make reasonable guesses/estimates. But I've recently found out that these guesses can be very misleading.

I have my own database of Alpari's minute data for most pairs going back several years and I'm a programmer, so I often write my own automatic backtests. But I have to admit that my programming skills often get challenged when I'm trying to program indicators. I can get the formulas to calculate indicator data, but often my programmed results don't match exactly what I see in MT4. And if my programmed indicator values don't match the MT4 values, then my backtest is useless. So, far too often I have resorted back to the manual "eyeball" test.

But last week I decided to take a different approach. I wrote a program that will give me the result of any particular trade in the past couple of years. I just have to identify the currency pair, the position (long or short), and then I also put in the SL (including trailing or static), and any applicable TP levels. By doing this, I am merging the eyeball test with raw, hard data. I don't worry anymore about trying to program all the indicators. Instead, I use MT4 to show me what the indicators looked like at any given point in time. Then, when I know that the indicators would have told me to Trade Pair X Long on Date Y, I put that exact trade into my simulator including my SL and TP levels.

The result is that I am now getting a FAR more accurate picture of how a system performs. Just as important, I am also learning loads about how a system performs under any particular configuration of lot sizing, SLs, and TPs. I'm also seeing how trailing stops effect a system as opposed to static stops. The results are really eye opening.

I have seen posts from highly respected senior members of this forum where they show their manual backtest results. In those results, they show that Trade X would have happened on Date Y with a final profit of Z. But I go back to my simulator and put in the exact same trade and I find that in real-time, the trade would have been a loser. Of course, I have found the same problem in my own manual backtest results.

This has led me to basically throw out all my previous notions about a particular system's performance based on my previous backtests. This has also made me realize that eyeballing a backtest is basically useful when you are trying to get a very quick feel for a system - "Do I even want to bother testing this further?" - but it is almost pointless for anything beyond that.

Of course, this harsh assessment varies by system. Some systems, based on their simplistic nature, can be more easily backtested by eyeballing charts. But generally speaking, you don't truly know how a trade would have performed unless you enter your exact parameters and let the computer go through the data - minute-by-minute (or even better, tick-by-tick) - and determine EXACTLY where a hypothetical trade would have been closed.

i agree, why not just apply the opposite. That way you have the risk and reward relationship that you are looking for which is approximately 2:1. I noticed you were averaging 15 to 20 pips profit and being stopped out at 30. The trade success rate would be greater then the break even rate of approximately 33%. I would keep a close eye on the trades by always taking a sample of the last 20 and seeing if the system is still maintaining similar results. Obviously there's going to be some variation from sample to sample but this way you can quickly identify if the system is losing consistency and a change needs to be made.

There's a lot of frustration at the beginning but as long as you know what you should do and keep trying to achieve that success will come