No Stop-Loss System

[trendie]

I am fascinated by systems developed by people that have no stop-losses.
I have been googling grid-trading style systems.

I am curious to deconstruct such ideas. The basic idea seems to be to take small gains, and let losing trades accumulate until price returns and they can be closed out for BE or a gain. (eventually)

Are they simply a reflection of peoples inability to accept the fact they are wrong??
Are grid-trading systems fundamentally flawed?

I know that overnight charges can affect the end-result. Aren't there conditions where the overnight charge is positive, ie, you are paid? Could you construct a grid-trading system where losing trades result in the trader being paid positive interest charges?

I am a curious orange.

EDIT: being a trend-follower, I am inclined to add to winning positions. (although none of my rules use this facility)

What If:
We used Point and Figure as a starting point. A 30-pip box size. 3-box reversals. What if we added to each new box, ie, every 30 pips. but didn't close and reverse direction unless there was a 90 pip (30pip box times 3 box reversal) change of direction?
That is, the market would need to move 90 pips to change direction, but once changed, you add a new trade every 30 pips it moves in your favour???
This way, you are adding to positions in the direction of the trend, adding to winning direciton, rather than adding to losers??

I am thinking too much aren't I??

 

[shadowninja]

Are they simply a reflection of peoples inability to accept the fact they are wrong??

The premise for working no-stop-loss systems is that you have a general direction to trade in and a reason for a specific entry ie price action suggests that you should open a position. Then you use PA to determine when to take profit. In the event that it goes against you, you have an area through which if the price moves you'd get out. This means you do in fact have a stop but by the time it's hit you've taken so many pips that it won't affect things. Snag is, knowing my luck, I'd take a hit the first trade I go in to. The second snag is, will you really bail after taking 100 profitable trades and accept your first loss or will discipline fail and you'll give back a year's profit? You could put a 500 pip hard stop, I guess.

One of the biggest dangers, of course, is over-leveraging and getting a knock on the door by IG's bailiffs.

-SN

PS What kind of stuck up **** uses the word "premise"? Oh...

PPS having registered in 2004, surely you knew this already?

PPPS If any new traders are reading, I highly recommend not doing this as I've seen so many new traders wipe out their accounts very quickly by not using a stop.

 

[SvenFoster]

trendie.
Grids /gridding was very popular for a while and can work if you like a simple life and have a basic understanding of PA and use small enough positions.

If your trading something like AUDJPY you'll get something like 4% P/annum on a position.

I've seen variious attempts at it involving
1) ahead of the price only
2) above+below the price
3) Bi directional grids, on the same pair / on multiple pairs to attempt to negate the Drawdown on moves.


They all can work but the biggest issue I think is that the position sizes are to small to make it interesting.
That said, I know a trader who was quite happy doing it and while I was telling him it wasnt worth it he was adding to his account, every day.

I can see a modified version involving SL's working for me, but the whole stoploss free thing doesnt appeal.

 

[MrGecko]

:smart::rolleyes:whistlingwhistling:innocent::innocent::innocent::innocent::clap::clap::clap::clap::clap::-0:-0:-0:

 

[temporarilylight]

They all can work but the biggest issue I think is that the position sizes are to small to make it interesting.
That said, I know a trader who was quite happy doing it and while I was telling him it wasnt worth it he was adding to his account, every day.

pretty much my experience.

I done it a good few years ago now whilst i was working a regular job.
I decided that stops were for girls and decided to trade small stakes alongside my regular job using limit orders that didn't require more than a few minutes per day screen time, and made roughly £300 per week which was a nice addition to my wages (although i know that most of you would get out of bed for that amount)

I done this for the best part of a year and made a good few thousand pounds, although trading without stops means that at SOME POINT you are gonna get wiped out. It pretty much guarenteed, even when trading UNDERleveraged. You still need to understand the basics and be able to read the price action, but in my mind, ive got about a 90% chance of increasing my intial pot using such a system.

It never sat comfotably with me though for the reason stated previously. At some point a crazy, relentless move will wipe you out, and because of that i dedicated lots of time to working out a way to trade intraday with 'propper' stops. Not easy in my experience. Ive had a few years of profitable consistency now, but that's not enough for me to consider myself a consistently profitable trader in my eyes.

My dad (whose made more money trading that i could ever dream) believes that it is the way one SHOULD trade, but that you need LOTS of money. Enough that when you are very underleveraged, you can still make a 'decent' weekly wage and be able to stand serious moves against you.

 

[trendie]

SN: yes, I have been here since 2004, but no, rejected grids because of all the reasons stated. The idea of adding to losers and holding on never appealed. But surely the flip-side, adding only to winners, would reverse the notion, resulting in near-term losses, and longer-term account-growing?

MrG: I can only apologise! I am going through one of my what-if phases, and trying to look at existing systems/ideas in a fresh way.

I dont like the 20 or 30 pips idea. How about percentage of current price? Such as, for GU, at 1.5000 say, you buy/sell a grid every 0.5%, in this instance, 75 pips. This would constrain the amount of capital used, and thus exposure, to 20 units at most. actually price would need to drop to 0.7500 for even half th units to be used.

 

[SvenFoster]

I think that accumulating in the direction of the trend esp if interest diff is +ve is a much better premise.
If using grids as your entry criteria(aka no TA required) then how do you manage the positions once on, for example on the losing side.
assuming long = +ve
a) it goes up and down, but generally up
b) it goes up and down but generallly down
c) it oscillates up and down.
d) it rockets
e) It plumets.

traditional trendfollowing grids do well in all except e) and then get wiped
traditional bi directional grids on range bound pairs do well in all until d) or e) as they always have more positions on the losing side.

Until you can address that, it is doomed for the fatal event when it happens.

Like you I keep thinking that flipping that equation so that you have more position while it moves is the way forward. but how to do it?

I think the idea is enticing but unless it can benefit from trends and reduce DD on adverse moves its doomed...doomed i tells ye

 

[barjon]

trendie

I think it's pretty safe to work without a stop-loss and to add to losing positions using a well constructed grid if you're playing the difference between co-related instruments (ftse/dow for example).

Otherwise it's an 'orribly dangerous game.

'Course you could argue that stock markets have had an upward bias for two or three hundred years, so you should be pretty safe hanging on and adding to longs. Always assuming you've got deep enough pockets and the occasional ten years or so to wait of course Mind you, you might find that the profit that was going to buy you a house would only get a small car when you cashed in.

good trading

jon

 

[trendie]

I agree, SF.
I think grids work as long as you only buy, or only sell.
since, if you are buying, and you get a sell signal, opening a sell merely "locks in" a position rather than allowing it to come to median or not.
There could some rule about closing out averaged positions to reduce the total loss. eg, if youre long 1.50, 1.51, 1.52 and 1.53, and price has fallen to 1,49, but hooks back upto 1.506, you could close out 1.50 and 1.51 for marginal BE (1.5 + 1.51 has average of 1.505), and reduce your losing overhead.

You could theoretically have a secondary mirror-account with shorts only, and manage accordingly.

However, the underlying essence is the unwillingness to take a hit. That is, the need to be right.
On balance, you're better off taking the hit and walking away until a prime opportunity arrives.

Yes, I thought the positive swap would help to mitigate a losing position if held long enough.
I have too much on my hands. Ad not enough spammers to chase.

 

[kimo'sabby]

I think probably most traders have toyed with methods that use no stops. Once a trade goes against a person it proves that the trade idea was not right, so the trader must realise that they can get it wrong again and again. Plus, why have losing trades hanging around like bad smells, they can be really off-putting and without a stop reduce profits even more. Get rid is what i say, the faster the better, why rely on hope when you can trade?

 

[trendie]

I think probably most traders have toyed with methods that use no stops. Once a trade goes against a person it proves that the trade idea was not right, so the trader must realise that they can get it wrong again and again. Plus, why have losing trades hanging around like bad smells, they can be really off-putting and without a stop reduce profits even more. Get rid is what i say, the faster the better, why rely on hope when you can trade?

a perfect conclusion to this thread.

 

[barjon]

trendie

I think it's pretty safe to work without a stop-loss and to add to losing positions using a well constructed grid if you're playing the difference between co-related instruments (ftse/dow for example).

Otherwise it's an 'orribly dangerous game.

'Course you could argue that stock markets have had an upward bias for two or three hundred years, so you should be pretty safe hanging on and adding to longs. Always assuming you've got deep enough pockets and the occasional ten years or so to wait of course Mind you, you might find that the profit that was going to buy you a house would only get a small car when you cashed in.

good trading

jon

Oh, and as far as adding to winning positions is concerned - I'm all for that.

General rule, though, is not to allow any loss on addition to wipe out what profit you've already accrued - so not without a stop-loss please.

jon

 

[counter_violent]

Barjon has it correct when he talks about a properly constructed grid.
Also people need to think about rule sets...you can have one rule set that will suffice 95% of the time....it's the 5% that perhaps needs additional rule sets...supporting and possibly quite different.

 

[trendie]

Barjon has it correct when he talks about a properly constructed grid.
Also people need to think about rule sets...you can have one rule set that will suffice 95% of the time....it's the 5% that perhaps needs additional rule sets...supporting and possibly quite different.

true enough. thats why I thought a percentage of current value was better than an arbitrary fixed amount.

 

[counter_violent]

true enough. thats why I thought a percentage of current value was better than an arbitrary fixed amount.

Very difficult one this as current value is only a snapshot in time and would have little relevance to a price that was only temporarily halted.

 

[shadowninja]

SN: yes, I have been here since 2004, but no, rejected grids because of all the reasons stated. The idea of adding to losers and holding on never appealed. But surely the flip-side, adding only to winners, would reverse the notion, resulting in near-term losses, and longer-term account-growing?

I was purely talking about not using stops.

 

[meanreversion]

It's usually worth testing all commonly held beliefs with historical data. Some truisms are borne out by stats, some are revealed to be bogus. E.g. always move your stop to breakeven when you can simply doesn't work over time (according to my analysis). The trend is your friend - this one does seem to help.

The problem with trying to backtest with no stops is that it's impossible to know what percentage of your account you're risking. I've tried removing stops with a couple of my systems (just betting a fixed amount each time) and the results are not particularly encouraging. You will make money sometimes, but in the long run it's not overly profitable, i.e. expectancy is close to zero.

Another saying is "cut your losses, run your winners".. this rule does hold up well in testing, so my conclusion for the moment is that trading without stops is statistically ineffective.

 

[Rossini]

I don't use stop losses as such when swing trading or trading over a fairly long TF. I find them a pain tbh.

I think its widely accepted that not using stops increases profitability, but eventually that one trade will cause you to blow up. So I use a strategy that enables me to exit with hopefully and usually (small) losses without having to use a SL. Having said that regardless of whether its day trading or longer term stuff I always enter a disaster stop to protect me in the event of a market crash, or if the first system should for whatever reason fail.

 

[Rossini]

It's usually worth testing all commonly held beliefs with historical data. Some truisms are borne out by stats, some are revealed to be bogus. E.g. always move your stop to breakeven when you can simply doesn't work over time (according to my analysis). The trend is your friend - this one does seem to help.

The problem with trying to backtest with no stops is that it's impossible to know what percentage of your account you're risking. I've tried removing stops with a couple of my systems (just betting a fixed amount each time) and the results are not particularly encouraging. You will make money sometimes, but in the long run it's not overly profitable, i.e. expectancy is close to zero.

Another saying is "cut your losses, run your winners".. this rule does hold up well in testing, so my conclusion for the moment is that trading without stops is statistically ineffective.

Good post.

Removing all kinds of SL's will obviously make it very tricky to calculate the total risk on any one trade, but if one is using a crash stop then you can at least work out in a disaster situation what % of your account you will lose.

Also you will have the knowledge of your average loss without using a SL and some kind of conservative position sizing formula could be used with this information.

 

[SvenFoster]

Good post.

Removing all kinds of SL's will obviously make it very tricky to calculate the total risk on any one trade, but if one is using a crash stop then you can at least work out in a disaster situation what % of your account you will lose.

Also you will have the knowledge of your average loss without using a SL and some kind of conservative position sizing formula could be used with this information.

I think thats even more so with stop and reverse type systems that are always in the market or use the inverse signal to liquidate and reverse. Often you'll have to base your position size on somekind of historical stats.

When it comes to manual/discretionary trading I think as long as you understand your risk you should do whatever sits comfortably with you. No point in putting SL on if it messes with your head, esp if your happier with less leverage and managing exits yourself