Stop/Losses - the myth(?)

[in2uxs]

The myths about setting a stop/loss

Now then, time to cause an argument I think.

Common conjecture is that we must set a stop loss in order to make money on the stock market - for example if one goes long on the FTSE and it drops 200 points then if we had set a stop loss at say 20 points then all we would have lost is £20 and not £200.

Again, in the same example, if the FTSE rises 200 points then we would make £200 profit. If we hit the trend 50% of the time then it would seem that a nice guaranteed profit is to be had.

However, unless I'm missing something this in practise doesn't seem to work out. A market never goes up in a straight line and the chances of it going up 200 points without falling back several times under a sensible moving stop/loss is very rare indeed. On average I'd be surprised if you made more than 40 points or so before your moving/stop loss is fired.

Again if we buy the FTSE at 6000 and set a target at 6060 and a stop loss at 5980. Again, in theory, every 2 trades we would on average win one lose one (even at random) netting us a £40 profit.

However, in reality would this be the case? There is the mathematical question of probability that cannot be ignored. For example if you buy the FTSE at 6000 points probability would suggest that there's a 3 times greater chance of it hitting +-20 points (therefore selling us up) before hitting a +- 60 points. Therefore over the couse of time we would draw level at best. I've ran tests on excel and this does hold true.

So, do stop/losses actually gurantee a profit or just allow us to stay in ther game that bit longer?

 

[Crap Buddist]

The myths about setting a stop/loss

Now then, time to cause an argument I think.

Common conjecture is that we must set a stop loss in order to make money on the stock market - for example if one goes long on the FTSE and it drops 200 points then if we had set a stop loss at say 20 points then all we would have lost is £20 and not £200.

Again, in the same example, if the FTSE rises 200 points then we would make £200 profit. If we hit the trend 50% of the time then it would seem that a nice guaranteed profit is to be had.

However, unless I'm missing something this in practise doesn't seem to work out. A market never goes up in a straight line and the chances of it going up 200 points without falling back several times under a sensible moving stop/loss is very rare indeed. On average I'd be surprised if you made more than 40 points or so before your moving/stop loss is fired.

Again if we buy the FTSE at 6000 and set a target at 6060 and a stop loss at 5980. Again, in theory, every 2 trades we would on average win one lose one (even at random) netting us a £40 profit.

However, in reality would this be the case? There is the mathematical question of probability that cannot be ignored. For example if you buy the FTSE at 6000 points probability would suggest that there's a 3 times greater chance of it hitting +-20 points (therefore selling us up) before hitting a +- 60 points. Therefore over the couse of time we would draw level at best. I've ran tests on excel and this does hold true.

So, do stop/losses actually gurantee a profit or just allow us to stay in ther game that bit longer?

hmm, I think or look at them a stop loss as a way to stop a loss, not a trading loss, but a loss bigger than anticipated. (a disaster stop if the market dont give you an out) i think you/we maybe need to have a look at an individuals method approach to the market. Do they take a random approach or trade on the basis that markets are just random ? or on the basis there is order or logic (although it appears and can be random at times) in price movements.

Some might use a stop order to take them out of a trade in profit, ie trail up the stop, and some may use a stop order to open a trade first . This was exampled on another thread with someone trading ES, although I do not like that approach, of using a stop order to open counter to original plan.

Many different views on stops, including, dont use them. I think whilst people are learning it maybe I good idea to have use of them. Then maybe as we learn more we dont have to wait for the stop to be hit, we can act first, but that assumes people are at the screen being able to watch tic by tic prices I guess.

 

[nine]

Stop losses will reduce the (simple) profit of most systems (Stridsmann) when compared with an alternative exit strategy like reverse on opposite signal. This is not always true but can be shown in a large number of cases.

This effect is most obviously true if you set them too tight for the volatility of the market being traded but even if larger stops are used there are quite a number of situations where a stopped trade returns to profit in a well designed system as long as the stop is not exercised.

What a stop loss does that is extremely important to profitability is reduce the variation in the size of losses. Because the std deviation is lower you can make much larger bets for the same risk of ruin. Thus you can buy 3 contracts with a 10% risk of ruin with 100k of capital (say) instead of having a 10% ror with only 1 contract. As a consequence your (complex) profit is higher with well chosen stop losses.

 

[TheBramble]

The myths about setting a stop/loss

Now then, time to cause an argument I think.

Common conjecture is that we must set a stop loss in order to make money on the stock market - for example if one goes long on the FTSE and it drops 200 points then if we had set a stop loss at say 20 points then all we would have lost is £20 and not £200.

Again, in the same example, if the FTSE rises 200 points then we would make £200 profit. If we hit the trend 50% of the time then it would seem that a nice guaranteed profit is to be had.

However, unless I'm missing something this in practise doesn't seem to work out. A market never goes up in a straight line and the chances of it going up 200 points without falling back several times under a sensible moving stop/loss is very rare indeed. On average I'd be surprised if you made more than 40 points or so before your moving/stop loss is fired.

Again if we buy the FTSE at 6000 and set a target at 6060 and a stop loss at 5980. Again, in theory, every 2 trades we would on average win one lose one (even at random) netting us a £40 profit.

However, in reality would this be the case? There is the mathematical question of probability that cannot be ignored. For example if you buy the FTSE at 6000 points probability would suggest that there's a 3 times greater chance of it hitting +-20 points (therefore selling us up) before hitting a +- 60 points. Therefore over the couse of time we would draw level at best. I've ran tests on excel and this does hold true.

So, do stop/losses actually gurantee a profit or just allow us to stay in ther game that bit longer?

A stop loss doesn’t guarantee you a profit directly. It guarantees you small losses.

Let’s assume you’re trading with the main trend. This is a given for my trades. That will generate you more profitable trades over the long run than any other single trading rule.

Depending on your trading timeframe, your stoploss will keep you from getting out too soon, providing you tailor your stop size to your trading timeframe and factor in your risk.

You obviously need to allow significantly more wiggle on a trade you expect to hold for months than you would on one you don’t plan on holding past the end of the day. You’d look back over the previous retracements in uptrends and note what the average and maximum retrace was in terms of points. You obviously want to tread very carefully that thin line between getting you out early enough into a trend reversal to allow you to keep the largest part of your accrued profits and not slipping out of the trade on ‘normal’ secondary or intermediate retracements. You need to be quite clear what to you, for your given instrument and based on that instruments historical retracement profile, what size you need to calculate for your stop.

To use your example, buying the FTSE in an uptrend at 6000 with a target of say 6400, and previous retracements on uptrends in recent history being no more than 66 points, you’d set a stop of say 70pts. If you’re timeframe was intraday and you were looking for it to move from 6000 to 6050 that day, you’d check your intraday charts for recent uptrend performance and see the most likely secondary retrace would be in the order of 33 points. So you’d set your stoploss at 35. Or would you?

Some will suggest ATR as a stop size factor, others a fixed point or percentage trailing stop. Each have their merits. But for my part, there is simply nothing to beat profiling your stop size on the recent historical performance of the instrument itself.

The additional benefit in my view of taking this trouble is that you will get a reasonably clear idea of what your risk is going to be, from that it’s simple arithmetic to calculate your position size. But before all of that, and to answer the question I raised above, it gives you the opportunity to compare your stop and therefore your risk, against your potential reward. It is at this juncture that you decide whether it’s worth the effort and the exposure. A stop/risk of 70pts for a potential reward of 400 looks rather good. But a risk of 35 for a gain of 50 – is that acceptable? For some, it would be. But you need to do the calculations in order to be able to make that informed decision. Just going in Long on an uptrend with a stop loss does not a sensible trade make. Risk and Reward must always go hand in hand.

 

[fibonelli]

My learning so far is that successful traders have

1. A nice proportion of big winners
2. Lots of small winners
3. Lots of small losses

The stop loss is there to cut out the big losses but not the big winners.

Setting the stop loss seems to be an art and a science! Could use Charts, ATR, fixed %, fixed £ or a mix of these.

 

[alexander]

Common conjecture is that we must set a stop loss in order to make money on the stock market

Stop losses are used to increase the average win to average loss ratio of a suitable strategy that has already a positive expected gain.

Stop losses are not applicable to all trading strategies.

Stop-losses on the buy side offer liquidity in the order book and thus very dangerous to have in place upfront. Better keep them as mental although this requires discipline.

Stop-losses on the sell side take out liquidity from the order book and thus very hard to hit. Better keep them as mental also and never place them upfront.

Tight stops is a way of transferring your wealth to the market.

I'm sure others have some things to add.

Alex

 

[tightstops]

tight stops arent necessarily a losing proposition neither are wide stops or no stops at all. im sure its possible to be profitable overall trading any of these three approaches. i got an email the other day saying we could watch john carter trade for an hour on a paticular platform he must have been advertising for the company. he was the author of a book i quite liked so i took the oppertunity to watch. he had a few stop and limits preset a certain way that he could apply to a setup he liked. one of them was to enter a position with a 9 point stop loss and close 1/3 of position after 2 pips tighten the stop a bit then another 1/3 after 4 and the last 1/3 after 6 pips. seemed a bit strange to me but it obviously works for him. he said the much touted 2 or 3:1 target is one of the reasons why most people fail.
i think people just tend to fall into theyre own stop and limit ideas based mainly on how their setups work. obviously say tight stops would be a losing proposition for some approaches but may be the optimum approach for other trade ideas. to generally say tight stops are a bad idea i think is a mistake.

 

[One Eyed Shark]

Intersting debate...

Surely a stop has to be placed at the point where you can clearly state, if price reaches that level my entry is wrong.

One certain way I can see to bleed to death in this business is to trade the markets without a stop.

Hope is not is not something that will be in the armoury of a successful trader.

For the record I trade the ES/6E with a stop that is between 3 - 6 ticks from my entry price

 

[TWI]

When using a stop it is important to remember that if a stop goes and the setup is repeated it is a requirement to re-enter. Stops will not work if you are not prepared to continue to follow your rules after a number of stops have been taken.

 

[fibonelli]

This video adds an interesting idea to setting stop losses and of managing a trade.

http://www.tradethemarkets.com/public/341.cfm?sd=61

 

[Hoggums]

Intersting debate...

Surely a stop has to be placed at the point where you can clearly state, if price reaches that level my entry is wrong.

One certain way I can see to bleed to death in this business is to trade the markets without a stop.

Hope is not is not something that will be in the armoury of a successful trader.

For the record I trade the ES/6E with a stop that is between 3 - 6 ticks from my entry price

I totally agree with this comment - placing an abritrary stop of say 20pnts makes no sense to me - it could easily get taken out and then reverse direction - you should set it at the point where you can say - if the price hits this point then the market is not going to go in my intended direction so I'll get out.

 

[TheBramble]

Before you get too bogged down in the purely mechanical and arithmetic aspects of stop size calculation, and there is a lot to get enjoyably bogged down with in that area, there are also the non-mechanical aspects to consider. The most important of which in my book is local Support and Resistance.

If you’re going to take a Long/Short position, with the main trend of course, your initial stop is usually most sensibly placed just below/above the most recent area of Support/Resistance.

This of course, like any other element of your risk assessment calculation, needs to be factored into your Risk:Reward and your Position Size calculations as well. But this is just your initial stop. Once your trade is in play, you need to be constantly monitoring current action from a number of angles. The first, which I mentioned yesterday, is the profile of the price action compared with its recent historical performance (recent being relative to your trading timeframe). Another factor to consider is tightening your stop dependent upon volatility.

The natural tendency is to tighten stops as volatility increases. Wrong. While you never move your stop against your risk, increased Volatility is a time to sit tight on your stop. When Volatility slackens or even flattens, that’s the time to aggressively move your stop in close.

The above is only counter-intuitive and non-rational until you intuitively rationalize it and you can’t do that if I explain it in detail now.

However, if anyone is having trouble I’d be happy to expand on that comment privately in order not to detract from others’ efforts to enter the void.

 

[firewalker99]

The natural tendency is to tighten stops as volatility increases. Wrong. While you never move your stop against your risk, increased Volatility is a time to sit tight on your stop. When Volatility slackens or even flattens, that’s the time to aggressively move your stop in close.

The above is only counter-intuitive and non-rational until you intuitively rationalize it and you can’t do that if I explain it in detail now.

Very true.
Excellent stuff The Degrees. Interesting thread developing here too.

In my opinion how wide your stops are should depend on the volatility of the market at that time and where your entry price is. If you have a long entry very near to support, you can set a tight stop, but if you can't catch it you either need to skip the trade or use a wider stop. Setting a tight stop in terms of the amount of money you are willing to risk, will only guarantee you will in effect lose that money.

 

[TWI]

I agree. Stops should be worked out basis volatility and as such, from that, trade size can be determined based on capital you wish to risk. IMO if you wish to create something more generic it is useful to use volatility in many parameters of a system.

 

[One Eyed Shark]

Two different discussions

So implicity by the fact that we trying to place the correct stop does that mean we are answering the initial question in the thread that stops are essential whether mental or hard are in place ?

The question of the correct placement is too subjective to be discussed as one subject and has to be relevant to the timeframe and trading style as already pointed out earlier in the thread.

Only advice from experience I can give to a new trader is DO NOT TRADE WITHOUT STOPS.

So much emphasis on learning to trade is where to find the perfect entry, equally important in my humble opinion is identifying at what level does price have to reach before I am prepared to admit my entry is wrong.

That is your stop level and you should never enter a trade without that level indelliby printed in your mind as clear as your entry .

 

[in2uxs]

Thanks to contributing to the thread.

I tend to view the stock market as a giant maths game rather than anything very serious. I have tried to treat the stock market seriously, but alas due to the folly of it I gave up on that approach years ago. To me the stock market is as random as a string of numbers generated by the excel function =randbetween(-1,+1). I agree trends occur, but trends also occur in random numbers too.

Personally, I have a problem with stops. I agree that they are all well and good for preventing total annialation due to unforseeable events - there is only one greater enemy than that, and that's going into the market using emotion and "it can't go any lower than this" mentality. Have that attitude and you'll end up burning your hands under hot water due to the shame of it.

The problem I found with setting a stop is that in breed's greed. If I think the FTSE has bottomed out, and I set a stop say 6 points below then I am very tempted indeed to risk a greater chunk of my capital - instead of buying say 1 FTSE index points I'll buy probably 20. My theory been well the most I can lose if I'm wrong is £120. This in itself becomes a gamble and the mood has changed from playing the market as if it’s a game of chess to one of flipping a coin. Get it wrong 8 times (this can be done easily within the day when panic sets in) and you've lost nearly a grand. However, if you'd have brought just 1 index points in the first place the stock market would have to drop 1000 points in order to amass such a loss.

A six point stop is nothing - to judge a significant rise so accurately without getting bumped out by the noise of the stock market is actually very mathematically remote. I ran tests over the weekend using past graphs to test out such stops, and trailing stops, and found myself breaking even at best. More often than not noise would hit a stop - even if I hit the trend right - and even if I hit the timing spot on then before the market had moved too far then it would zigzag back down and hit the trailing stop which in affect eliminated any kind of great gain. I also created a simple random number generator on excel to simulate a market, (it actually worked very well and to be honest the graphs it created could match any FTSE daily movement even using an unbiased formula such as =randbetween(-1,+1). Using this I used stops and failed to make any decent profit..

Personally, I do set stops but they are set very low - its just a safe guard in case someone nukes New York in the early hours rather than if it various 20 points due to Dubbya swallowing a chicken.

 

[fibonelli]

The reason I posted the video in post no. 10 is because stops shouldn't be thought as being set in stone.

Where they're set is a different subject altogether!

 

[One Eyed Shark]

in2uxs

Big difference between a 6 point trailing stop and a 6 point stop loss from entry.

To trail with a 6 pt stop I can agree that it maybe untradeable.

The difference on how you approach the markets, dictates your stop strategy.

I am at the other extreme to your point of view and do not accept that the markets are random.

Day in day out the markets repeat specific patterns that can be traded, not random but precise, the skill of the trader is to identify the patterns as they develop .

If you are in tune with the market, you know where you are right and where you are wrong.

 

[Crap Buddist]

"Ive paid my deposit sir, how biggeth my pot be please?"

"ohhh, fraid I cant be telling ye that, but blow for ya I will sir. You's just gonna have to watch til me puff starts to ease."

"Right ye old codger, blow my size and make it a good un"

 

[Crap Buddist]

What about this?

If we know why we start a trade, then we will know when to stop it.

Its beautifuls.... CB.