Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

How Do You Compound Your Account While Paying Yourself? 22 replies

I dare you show your statement 13 replies

Twoblink tweaks and mods 61 replies

Vegas 4 hour Tunnel mods 4 replies

Attachments: First-Strike & One-Night-Stand users, show yourself and your mods, if you dare!!!
Exit Attachments

First-Strike & One-Night-Stand users, show yourself and your mods, if you dare!!!

Quoting birdt
I have followed Joel's blog for several months and although I have not followed or traded the strategies as he has done, there has never been any suspicion in my mind that he might be fluffing the results.......
Many thanks for the excellent insights.

One concept that I learned from Joel's blog is that of aggregating edges. If, for example, Monday's initial breakout (as captured by First Strike) does give a >50% clue as to likely direction for the rest of the week, then it would be possible to use this info as a kind of long term trend indicator, and seek only trades in that direction, using entries provided by shorter term strategies that in themselves have their own edge.

Of course this is much the same principle behind multi-timeframe analysis, confluences across different S/R types, and across correlated pairs, etc. The more (preferably independent) reasons one has that price will move in one direction, the greater the probability of a 'good' entry.

Hope that makes sense.
  • Post #62
  • Quote
  • Edited 10:23am Jan 17, 2009 2:11am | Edited 10:23am
The first half of this post confirms my own suspicion of joel's equity curves on his site. He is quick to write articles for magazines and post what seems a seemingly indestructable strategy to a website, but is not answering emails with legitimate concerns about the strategies.

deathbait, I'll take you one step further. Go to Tkimble's Weekly Scalper thread, and look at Joel's posts. He has more pretty graphs with high timeframe data (which confuses the backtester.) I got the same thing he did, until I tested with m1.

This further concludes my point that anybody can post a fake grail to a website and start a business.

Quoting deathbait
I've been lurking here for over a year, and finally got off my ass and registered an account just to put in my 2 cents worth regarding joel's strategies.

The gains he claims to have been made from first strike and ONS( he plots them via tradestation, and so do I) have been accurate. He has indeed been making alot of money if he used these system the last 1-2 years.

Unfortunately, tradestation only keeps minute data since mid 2003, and his graphs run well before that. Notice that the performance graph from 2003-present breaks even only till recently,...
I think Joel's concept is very solid. The main idea of it is - we cannot control the markets. People who try to exit before Friday/Monday think they can and in my opinion thats the problem of "the 95% traders" in the long-term. Cut your loses, let your profits run - thats the only edge I can depend on in the markets with swing/position trading. 51% + probability trades/predictions happen in daytrading only.
just sent a PM to TheRealThing pointing this thread. He was online 49hr ago.
If anyone has been trading this consistently for the past 6 months, they could now easily take off their original stake plus 10% and have a guaranteed winner no matter what happens from here.

The OP was looking for other people who are actually trading the method to compare notes with. As always, the thread is turning into a bash fest and has taken the normal turn away from the original intent. If only trading was equally predictable.

The people who are trading this live seem to be doing well with it. If it breaks down, they'll probably either stop, pause or adjust. That's trading.

The core of the system, or its edge, hasn't been mentioned by the backtesters. If the edge is valid, at its simplest level then the method can't be far off. I'd like to say to say, it can be made to work. But for me it already does.

How about it all you backtesters. Do you think the edge is definable and valid? You must know what it is.
Quoting billflet
If anyone has been trading this consistently for the past 6 months, they could now easily take off their original stake plus 10% and have a guaranteed winner no matter what happens from here.

The OP was looking for other people who are actually trading the method to compare notes with. As always, the thread is turning into a bash fest and has taken the normal turn away from the original intent....
Here are my FirstStrike Plus backtest results using Alpari 1M data from January 2001 until November 8th, 2008:

EUR/USD: 47.14% Win Prob, 1.31 W/L Ratio, Kelly: 0.068, Math Exp: 0.089
GPU/USD: 41.95% Win Prob, 1.57 W/L Ratio, Kelly: 0.050, Math Exp: 0.078
USD/CHF: 45.07% Win Prob, 1.28 W/L Ratio, Kelly: 0.022, Math Exp: 0.028
USD/JPY: 45.45% Win Prob, 1.45 W/L Ratio, Kelly: 0.078, Math Exp: 0.114

Averages:

44.90% Win Prob, 1.40 W/L Ratio, Kelly: 0.05542, Math Exp: 0.0776

Though not the most profitable system out there (Joel describes it as a basic system which can be improved apon), it is most definitely profitable.

Glen
I have read Joel's blog from the beginning. Several months I began testing it with a small real account. So far, so good.

The system's core is an Opening Range Breakout (ORB) of Donchian Channels. Think Toby Crabel and Larry Williams. First Strike Plus is Joel's version. The core of the system has been used successfully across all tradeable markets for many, many years.

Will FS plus stand the test of time? I don't know. What I do know is that the core can be adapted to work in any market.
Quoting daytrading
. (Jan 09 is at +1.9% as of y'day's close)
These are the results on demo for the last 2 weeks using FSP. Maybe, these are just 2 lucky weeks, time will tell. But these results are a fact which come from a forwardtesting. You decide if the MM used and the profit are good or bad. My opinion is already known to me.
Attached File(s)
File Type: doc FSP_results-Mid Jan.doc   153 KB | 630 downloads
A good trade is when I Break Even. A great trade is when I Take Profit.
Quoting billflet
If anyone has been trading this consistently for the past 6 months, they could now easily take off their original stake plus 10% and have a guaranteed winner no matter what happens from here.

The OP was looking for other people who are actually trading the method to compare notes with. As always, the thread is turning into a bash fest and has taken the normal turn away from the original intent. If only trading was equally predictable.

The people who are trading this live seem to be doing well with it. If it breaks down, they'll probably either...
Bill,
thanks for saying what i am sure lots of us are thinking,
whilst i am glad i opened this thread because it has shown quite a few people using the methods in different ways, i am also dissapointed and a bit imbarrassed that it has brought joel's name into what you called a bashing fest. I dont have a problem with contructive criticism but some people, i think, like to go out of there way to be destructive.
hopefully we might be able to continue with some positive discussion.

Also, where are the Aussie's trading this!!!!! cmon aussies cmon cmon,

As i have said before i am just trading the systems straight up as they are presented by joel, but from descussions i am wondering how to maximize my trades, is it simply to give more weight to the trades that are with the trend like dave does in his variation?,
these are the type of ideas, simple but maybe powerful in the long run, that i was hoping to unearth with this thread.
I look forward to hopefully a lot more positive contructive interaction with like minded people.
cheers
geoff
What is the granularity of your testing data? Is it 1 minute? I'd love to see the logic of your code and compare it to my own.

Quoting gnllr
Here are my FirstStrike Plus backtest results using Alpari 1M data from January 2001 until November 8th, 2008:

EUR/USD: 47.14% Win Prob, 1.31 W/L Ratio, Kelly: 0.068, Math Exp: 0.089
GPU/USD: 41.95% Win Prob, 1.57 W/L Ratio, Kelly: 0.050, Math Exp: 0.078
USD/CHF: 45.07% Win Prob, 1.28 W/L Ratio, Kelly: 0.022, Math Exp: 0.028
USD/JPY: 45.45% Win Prob, 1.45 W/L Ratio, Kelly: 0.078, Math Exp: 0.114

Averages:

44.90% Win Prob, 1.40 W/L Ratio, Kelly: 0.05542, Math Exp: 0.0776

Though not the most profitable system out...
Why use the 1m "zoom in" for testing a weekly strategy? I think if you wanna test a weekly strategy a maximum "zoom in" to 1H is more than enough.
Also, what kind of strategy for the management of the open trades, your EA has? Did you use exactly the rules Joel mentioned in his blog? 2 weeks can be easily manually tested and compared with the EA's results.
A good trade is when I Break Even. A great trade is when I Take Profit.
Quoting Cornelius
Why use the 1m "zoom in" for testing a weekly strategy? I think if you wanna test a weekly strategy a maximum "zoom in" to 1H is more than enough.
Also, what kind of strategy for the management of the open trades, your EA has? Did you use exactly the rules Joel mentioned in his blog? 2 weeks can be easily manually tested and compared with the EA's results.
I can only imagine that it would matter due to news events, and you would have to take into account market liquidity during these times. Is there a reliable way to do that, tdion? Can anyone attest to liquidity of Oanda accounts during stuff like NFP?

Also, did anyone think that the not-so-great period of backtesting could be a smaller dip in a bigger time frame test? We don't see a few months of drawdown as a big deal across a 10 year test...should we see a few years of drawdown as a big deal across a 100 year test? Does anyone have the huevos to stick it out across 5 years of flatness?
Hindsight is 20/20
Quoting tdion
What is the granularity of your testing data? Is it 1 minute? I'd love to see the logic of your code and compare it to my own.
Hi tdion,

Yes, it's one minute data.

Regards,

Glen
Quoting hanover
Excellent point. Without wishing to sound too condescending, I would guess that the vast majority of folk here have little idea as to the relationship that exists between sample size and statistical confidence.
I do. And I have you to thank. Your posts piqued my interest and prompted me to buy and read a few books about basic probability theory. Very illuminating for people who are willing to explore beyond trading specific books.
Quoting tdion
The first half of this post confirms my own suspicion of joel's equity curves on his site. He is quick to write articles for magazines and post what seems a seemingly indestructable strategy to a website, but is not answering emails with legitimate concerns about the strategies.

deathbait, I'll take you one step further. Go to Tkimble's Weekly Scalper thread, and look at Joel's posts. He has more pretty graphs with high timeframe data (which confuses the backtester.) I got the same thing he did, until I tested with m1.

This further concludes...

Yes, I've already done that. And I've raised that point with Joel too. Like I said before, I've concluded that Joel's data isn't accurate/complete enough for backtesting. The only thing I'm unsure of is whether he already knows it, and keeps using it as justification, or it's an oversight on his part.

Thing is, Joel hasn't ever replied my queries...and that's worrying, really. I do analysis of trading strategies as part of my job, and if someone told me my data set had problems, i would thank him and instantly correct that, not ignore the person.

Quoting Cornelius
Why use the 1m "zoom in" for testing a weekly strategy? I think if you wanna test a weekly strategy a maximum "zoom in" to 1H is more than enough.
Also, what kind of strategy for the management of the open trades, your EA has? Did you use exactly the rules Joel mentioned in his blog? 2 weeks can be easily manually tested and compared with the EA's results.
It's ok to use the 1 hour charts IF the results are the same as the 1 min charts. The problem is, they aren't. In fact, that holds for the 1 min chart too. If the results for 1 min charts aren't similar to the per tick charts, then the 1 min chart results are worthless too.

Yes, his rules are simple and as such, easy to code to the exact word. The problem with Joel's results is, the results for the last 2 years or so are accurate to the minute. Results before that aren't. So if all you're testing just the last 2 weeks, you'll get it right.

I only managed to spot his error because someone coincidentally spotted a similar error in my data series recently
  • Post #76
  • Quote
  • Edited 11:14am Jan 18, 2009 3:25am | Edited 11:14am
Yes, we are on the same page. Apparently he uses Tradestation, which I am not familiar with.... but

If it is anything like MT4, there are multiple timeframe tables in the database for each pair. And if one timeframe has 8 years of data, while another has 2 years, it will use the 2 years first (if it is a lower timeframe) and then supplement the test with the remaining 6.

You've really got to know what you are doing before you start publishing your results.

-T

Quoting deathbait
Yes, his rules are simple and as such, easy to code to the exact word. The problem with Joel's results is, the results for the last 2 years or so are accurate to the minute. Results before that aren't. So if all you're testing just the last 2 weeks, you'll get it right.

I only managed to spot his error because someone coincidentally spotted a similar error in my data series recently
I just went to infiniteyield.com again and almost wretched when I saw the title of the site....
"Guaranteed profits" Who does this guy think he is? What's my guarantee, Joel? If I lose money, will you reimburse me. I get sick of websites that feed that crap to the general public.

A proper title would be "High probability of losing your ass."
  • Post #77
  • Quote
  • Edited 6:41am Jan 18, 2009 6:38am | Edited 6:41am
I would like to express my gratitude for those who take the trouble to address the questionable issues surrounding ONS and FirstStrike, for I have been wanting to do this, but haven't do so because I do not want to openly accuse Joel of anything and incite a flame war.

When I first found out about his strategies, I did a quick backtest and found that it losses nearly every trade. Later I knew why. That's because the SL in his manual is not the same as the one posted in his trade alert. I also found out that he uses a proprietary algorithm to manage the pos. sizing. The problem with this missing ingredient is: he can always put winning trades with 2x size and losing trades with 1x size (or whatever you can think of), which of course artificially turns the strategy to become profitable. I never knew whether he did this or not. But one thing is certain, Joel is not a billionaire yet. If he can achieve such a performance year-in-year-out, there's no reason why he isn't a billionaire now, given that he have already traded for xx years. Don't tell me he can't use his strategies in fx futures or other futures before retail fx spot arrived.

Another thing to note: he actually sold the proprietary pos. sizing algorithm to a South African firm for $10k. I do not make this up, it was in one of his posts long time ago. (PS. post can be deleted, so don't whine if you cannot find it). I sincerely urge everyone interested in his strategies be sensible and logical. Put this in your own shoes. If I have so good a strategy, and I have already traded for so many years (means I already made a lot of money), why do I want to sell one great strategy for a mere $10k?

One more thing to note: Joel actually organizes some sort of traders meeting (though not frequently) and charges excessive fees to anyone wanting to attend. You should have got an invitation if you subscribe to his trade alert.

By doing all the mentioned, Joel has nothing to lose but everything to gain. These are the questionable elements that need to be answered, and I doubt anyone can come up with a reasonable answer aside from nonsense like sharing is caring or whatever Samaritan sayings you can come up with.

For those who have been profiting these past few months. I would say caveat emptor. You are just lucky because the market has been volatile. Big crisis like that only happens every 40 yrs while small crisis happens every 10 yrs. One thing is for sure, we'll never know what Joel has really done unless he posts his pos. sizing algorithm.
  • Post #78
  • Quote
  • Edited 7:30am Jan 18, 2009 7:28am | Edited 7:30am
Quoting tdion
I just went to infiniteyield.com again and almost wretched when I saw the title of the site....
"Guaranteed profits" Who does this guy think he is? What's my guarantee, Joel? If I lose money, will you reimburse me. I get sick of websites that feed that crap to the general public.

A proper title would be "High probability of losing your ass."
tdion - you go on about back testing accuracy and yet you have difficulty reading a website... there is no guarantee in Joel's title.. there is a Google related link to a system seller's website, who offers some form of guarantee - nothing to do with Joel. If you actually bothered to read Joel's blog, you would see he makes no guarantee. So that' the second inaccurate statement you've posted in just a few posts.

So let's review what's happening on this thread : several people share their real profits, derived from someone giving something for free. Others who are not trading the method bash it completely saying that it either doesn't work at all or not in all years or not long term... my goodness, there are a lot of very angry failed traders out there !

There is a trading fundamental : to trade any method well you must make it your own. If it is a mechanical method, you must test it, backwards, forwards and preferably both and satisfy yourself that it is worth trying. Then trade it if you choose to. If you do this, you can ignore all the nonsense posted by others (including myself ) and actually become a real, non-whinging trader.
Anyone who trades blindly just because of a strategy posted is doomed to get into trouble. I had been trading his strategy since I found it last year from some posting in FF from someone who modified it to his liking from becoming a bad trader to a better trader than before.

Anyhow, back to the point. I have never believed in backtesting and draw conclusions based upon the past. The only way is to try to demo forward, and as they all say, how one trades in demo is different from trading live [totally agree]. I have attached a demo chart I sent to someone showing what is the potential of this past week trade [ignore that buy alert - was testing Firebird EA]....

Oh yes, I do not wait until Friday to close a position when I was demoing it, I would close it before hand. If I had crystal ball, then I would be rich by now....
Attached Image (click to enlarge)
Click to Enlarge Name: GBPUSD_1H.jpg Size: 159 KB
http://www.infiniteyield.com/

What is the first word you see highlighted in gigantic red letters in the title of the site?

Care to retract your statement about my inaccuracies?

Quoting jtrade
tdion - you go on about back testing accuracy and yet you have difficulty reading a website... there is no guarantee in Joel's title.. there is a Google related link to a system seller's website, who offers some form of guarantee - nothing to do with Joel. If you actually bothered to read Joel's blog, you would see he makes no guarantee. So that' the second inaccurate statement you've posted in just a few posts.
Jtrade, please do not preach to me. I have been here a year longer than you have. I have a Masters Degree in Software Engineering, and am extremely miticulous with my approach to backtesting and concluding the validity of systems. I have been known to make a mistake sometimes, but I am no "angry trader" ... simply a realist.
0 traders viewing now