try
- #23
- Jul 11, 2007 9:55pm Jul 11, 2007 9:55pm
Dr. Van Tharp demonstrated in one of his books that sound money management, exit strategy and postion sizing can make random entry a profitable system. Weird.
Hi Stevegee
Have you tested this, and if so, over what time frames?
Sorry to be so negative, but it sounds too good to be true. If trading was this easy, we'd all be millionaires.
If it's possible to trade profitably on the basis of a coin toss entry, then all of the entry systems that have ever been devised are a complete waste of time, as is the maxim "the trend is your friend", since you'd be counter-trend trading, randomly, 50% of the time.
Cutting losses short and letting profits run is profitable only to the extent that markets trend, and far enough to overcome costs.
In all honesty, I'd be delighted to be proved wrong, and I'd be the first to congratulate you.
I admit that I'm skeptical, but I apologize if this any of this sounds facetious or mocking, I don't intend it to be.
Best wishes
David
[My apologies to Inflection - I didn't mean to hijack your thread. This is my last post on this subject.]
[Edit #2: perhaps I've been too hasty. It's very possible that the best systems are simple. Combine this idea with (1) trading in a long enough timeframe to make costs less significant (and perhaps allow positive carry to offset this further), and (2) trading only in the direction of the trend in an even longer timeframe, and I suspect the probability improves significantly. But of course that's no longer a truly random entry.....]
"Though this be madness, yet there is method in it." -Hamlet Act II
My apologies, Inflection. I inadvertently hijacked your thread.
I may just move this whole coin tossing thing to my own thread. Or maybe I should just attach a lightening rod to my head.
Not at all!
My goal in starting this thread was not to say, "Hey, I'm so smart. Look at my system! Use this and you'll get rich like me!"
The half-baked system I've posted may in fact be completely bone-headed ... hence, the title, "Stupid System"
I've already had to make edits and corrections and even the simplest of questions from other participants has made me realize that I've overlooked this or that.But that's the whole point.
Discussion.
And really what I want to discuss is an approach that is not common ... statistical, probability based systems that are as mechanical as possible. I'm not opposed to discretionary trading. Nor am I opposed to chart studies based on indicators and patterns. I'm just trying to look at an approach that depends only on price action and probabilities. I may have the wrong idea altogether. I may not even be able to do the math correctly (as I have proven before).
But somebody else may have the right idea, or the genesis of an idea. And that's what the conversation is about.
It's not my thread, it's ours. If I was really bothered, then I would tell you to take it elsewhere. But I'm not. Not at all.
Now let me add this to take the discussion further.
The only reason I'm even attempting to develop something like this, rather than just paying for a signal service or using a discretionary approach, is because I've been doing this with stocks for a long time. I currently use a purely statistical system for trading equities based on an approach developed by a Russian physicist. The theoretical return for the past 10 or so years is more than 80% a year. Beats the market hands down. My real-world return is less ... about 30%, 40%, and 50% for the past several years ... but still pretty good considering that I take vacations, make mistakes, miss the market occasionally, am unable to short the stocks that the theoretical system would pick, or just try various unsuccessful ways to tweak the system.
It's completely mechanical and I only have to work about an hour a day. It's not a bad gig, but it just takes a fair amount of money if you really want to make a living. If I make closer to 50%, then no problem at all. But if I make closer to 25%, then I can't really be living where I'm living, doing what I'm doing.
I don't think the approach is directly applicable to forex, but it gives me the confidence to hunt for something that might work.
Anyway, my system probably won't work, but that's the idea.
My apologies, Inflection. I inadvertently hijacked your thread.
I may just move this whole coin tossing thing to my own thread. Or maybe I should just attach a lightening rod to my head.
Hi stevegee88, hanover, and DonQ ...
I found stevegee's idea interesting. So, I plugged the formula into my spreadsheets containing all the historical OHLC prices for each major pair. Without looking back at stevegee's ruleset, I'll put the formulas here:
Cells A-E
Date, Open, High, Low, Close
Cells F-I
Entry, Status, Gain, Risk
Entry: if previous day is holdover, do nothing, else enter (long) at the open
=IF(G3="HOLD";F3;B4)
Status: if today closes lower than open or lower than prev day, exit, else hold
=IF(E4<B4;"EXIT";IF(E4<E3;"EXIT";"HOLD"))
Gain: if today is exit day, subtract entry from close price
=IF(G4="EXIT";E4-F4;"")
Risk: subtract entry price from low of today (max risk for this trade)
=D4-F4
You can see the results in the uploaded image...
Gain = total gain for the period (2003-2007)
Maxloss = maximum single day loss
Maxrisk = maximum single day risk (low - entry)
Trades = total # of trades
R/R% = Maxrisk / Gain (used to sort the list)
Portfolio = cumulative gains from top to bottom (list sorted by r/r%)
Pips/Mo. = Portfolio / 54 months
If I typed in what I think I typed in (and I've been known to make mistakes), then I consider this stupid system to be way better than my stupid system.
What do you all think?
- #32
- Edited Jul 13, 2007 3:04pm Jul 12, 2007 8:07pm | Edited Jul 13, 2007 3:04pm
- Joined Oct 2005 | Status: Pip Slappa Extrordinaire | 1,013 Posts
You are in a maze of twisty little passages, all alike.
