Tuesday, February 16, 2010

Stop-Loss Points in Trading: Do We Need Stops When We Trade?

Henry Carstens has recently written about ways to manage risk other than through stop loss points. I've noticed with traders recently that stops not only tend to be price-based, but tend to be placed at price points that are relatively obvious. For example, someone who is short stocks will place a stop just above a recent high or vice versa. Stops just above or below recent price ranges are also common.

The problem with such stops is that they are natural targets for algorithmic programs that exploit asymmetries in buying and selling orders and tendencies. The trader with obvious stops falls victim to false breakout moves, exiting trades just before they reverse and go the intended way.

An important gauge of the value of your stops is to track markets *after* you have been stopped out. Do you stop loss points save you money on balance? Do they protect you from risk, or do they shake you out of opportunity? Most traders have never closely looked at the true value of their stops; they just take for granted that stops are needed and place them intuitively (and obviously).

Over the years, I've found that placing price-based stops further away from my entry at points where my underlying trade idea is clearly wrong helps keep me from getting shaken out of good trades. Those price-based stops act as "catastrophic" stops for me. More effective, as a rule, for my own trading have been indicator-based stops and time-based stops.

An example of an indicator-based stop would be a sudden surge in NYSE TICK to new highs or lows against my position or a surge in buying or selling volume against my position. As soon as I see the surge, I exit and reassess. More often than not, such a surge indicates that the short-term tide has moved against me and that it is not helpful to fight that.

Time-based stops are based on my observation that my best trades are executed well and tend to go my way relatively quickly, with little heat taken. If I have to sit and sit and sit in the trade, the odds that it will go my way eventually are diminished: the dynamics that placed me in the trade simply are not operative. Similarly, if the market moves against me very shortly after my entry, I generally find it's best to try to scratch the trade and reassess.

It *is* important to have stop-loss points for trades. As I've noted previously, everyone does have a stop-loss point: it is either explicit, based on market action, or it is unstated and based on pain. The latter can wipe out days' worth of good trading.

But while it's important to have stops, it's not necessarily the case that all stops need to be price-based. You should stop a trade when you see that the underlying idea is wrong, not simply when you've lost a certain amount of money.



RDV said...

I always use very short stops and I think it's a thing of getting a good moment of entry. If the market doesn't go my way soon, then it might have been the wrong moment of getting into the market and I'll step out anyway, with about one point loss max.

After long, long times of practising entries on the sim it's working out quite well. I really would hate to lose four or five points or even more, being stopped out.

madi said...

Helpful. Thanks.

RoachApproach said...

Great read!
I use for years a time stop on my positions and it also helps me to be patient and to stay in the trade.


vertical9r said...

If your stops are based on market action then how do you share size without knowing how much you are risking?

P.S. I love your work!

TWC Futures said...

I would have to say the biggest change to my trading and results in the past 2 years was changing from price based stops to event-based stops. My journal metrics support this as well as event based target prices Great job Brett!

Happy Camper said...

I have been a trading professionally for about a year and done > 7000 trades during that time. I have never ever used a mechanical stop-loss order.

However, I always have an exit strategy for every trade I make.

For me, a stop loss order is a kind of false security. It is like playing roulette and decide that you will stop gambling if you lose 2000 USD. Ok, you will not lose more than 2000 USD on that occasion, but it would not help you in the long run. Next time you will enter the Casino, your odds will not be any better than last time… the only thing you know is that you will not lose more than 2000 USD again.

In my humble opinion, trading is always holistic…entry, exit, position size, timeframe etc. If you know why you put on a trade and what you are looking for and what kind of profit you might expect, you will also realize when you are wrong or when something unexpected happens and you should get out of your position – which will be partly intuitive after some time. But I don’t like the idea of any kind of mechanical stop loss orders, for example 5 % below your entry, which is just naïve.

etoke said...

Excellent article as always.

I have until recently been challenged with the fact that while stops are a convenient way of quantifying risk per trade, they definately weigh down on overall performance.

Lately I have changed my risk-per-trade model from using hard stops (and measuring risk from entry to stop) to taking a statistical measure of volatility over a particular timeframe and measure my risk as the likely amount a position would move against me (or indeed in my favour for that matter) in that timeframe. Then measure this periodically throughout my trading to control the current risk.

I have found that this, in combination with smaller positions in multiple instruments (to further reduce the effect of statistical anomalies) has improved my results considerably.

Thank you for all your good work.