In the second half of 2011, and in particular the last quarter, I was experiencing a frustrating time. The markets were volatile, and I suffered a number of trades where price initially broke out in my favour, only to sharply reverse and end up generating losses around a full -1R. This was because those trades had not been open a sufficient length of time for my trailing stop to start moving from its initial price level.
The trade that really stung was a long position in a small UK stock in November 2011. After initiating the trade, my open profits reached about +5R after only three days, yet my initial stop had not moved, based on the rules I used at the time. Then, price did a sharp about-turn, and three sessions later, I was stopped out for a full -1R loss.
After this happened, I retreated to cash and took a break to think.
The subsequent change to my own method resulted from reading David Ryan's interview in Market Wizards. Let's look at what he had to say:
"I want to buy it as soon as it goes to new highs."
"Stocks should be at a profit the first day you buy them. In fact, having a profit on the first day is one of the best indicators that you are going to make money on that trade."
So that tallied with my own beliefs as a trend follower.
Jack Schwager asked him the following question: If you buy a stock at new highs and then it pulls back into the range, at what point do you decide it was a false breakout?
Ryan's response was; "If it reenters its base, I have a rule to cut at least 50 percent of the position."
He went on:
"In some cases, it will break out and come back to the top of the base, but not reenter. That's fine, and I will stay with the stock. But if the top of the base was $20 and it breaks back to $19 3/4, I want to sell at least half of the position because the stock didn't keep on moving. Frequently, when a stock drops back into its base, it goes all the way back down to the lower end of the base. In the example, if it goes down from $21 down to $19 3/4, it will often go all the way back down to $16. Therefore, you want to cut your losses quickly."
As Ed Seykota would say, reading that was an "A-ha!" moment, and I implemented what I refer to as the 'breakout stop' into my own method. The only difference is that I am more aggressive than David Ryan in cutting the losses - I fully close the position if price falls back below the breakout level.
Using this type of stop does have its own frustrations - if price oscillates around the breakout level, you can get whipsawed around. But I can live with that.
Immediately, the average size of my losing trades began to drop. Since adopting the rule in the summer of 2012, the average loss suffered has reduced to -0.41R.
In addition, my losing positions generally are cut very quickly - normally within 24 hours of opening the trade. This means that I do not have equity tied up for long in a non-performing position.
When you are adopting a trend following method, where the typical win ratio is less than 50%, this aggressive cutting of losses will have a major positive impact on your equity when compounded over a period of time.
My latest losing trade demonstrates a breakout failure very well. Here is the chart:
You can see how price was starting to form a pattern of higher highs and higher lows from the early October price low. Throughout November, price started consolidating, and volatility started contracting. This gave me a potential setup. As soon as price broke out, I entered on an intra-day basis.
As we can clearly see, this attempted price move was immediately repelled, and price fell back below the breakout level, triggering an exit the next morning. In this case, the loss suffered was -0.25R.
We can also see what happened next with price, which again mirrored Ryan's own observations. As of yesterday's close, price had fallen well below where my initial stop would have been placed (which, without the breakout stop, would have created a bigger loss), and price is now threatening to breakout in the opposite direction!
I have seen far too many charts, and suffered far too many losses, where the use of my own version of the 'breakout' stop' has helped save me money, far outstripping the losses suffered due to any whipsawing.
This type of stop matches my own beliefs about trading and breakouts in particular, and was therefore easy to adopt from a psychological point of view.
I am a breakout trader, and if price breaks out, then at worst price will re-test the breakout level, but will not fall back within the previous price range. If it does, I bail out and move on - the breakout has failed, and the reason or rationale for entering the trade no longer exists.
As a result, the vast majority of my losing trades can be viewed simply as a cost of business.
Steve, thanks so much for sharing your thoughts on this. I really appreciated the Market Wizards example, and why it would have had a big impact on you.ReplyDelete
Personally, I had a similar realization when reading Nick Radge's book. He laid out a pretty compelling case for trend followers to trail stops more aggressively. As you said, I've found the smaller losses that result from this approach really help with the positive expectancy over time.
On the other hand, my system doesn't use ranges like you do. So this seems like an interesting area for potential improvement. I will have to start thinking about how to codify and test integrating even more tighter stops for these situations.
Hi Jay, which of Nick Radge's books are you referring to?Delete