Readers Questions Answered ............................
Today we'll make a slight diversion and answer a couple of reader questions. Hopefully we can all learn from others questions that maybe we didn't even think to ask. And as always, if you have a question worth answering here, please send it in. Maybe your question will benefit all of us.
Question 1: What is the recommended loss percentage in a trailing stop?
Answer: Good question. What the reader is referring to is a specific type of exit methodology. A trailing stop is a stop-out level that changes every day, along with the price of a stock or index. The idea is to let a good trade 'ride' as long as things are going the right way, and scoot the stop level upward along with the stock.....just slighty below it, so as not to interfere with the natural ebb and flow of a chart. When things reverse, your trailing stop level should be close enough to the current price to get you out quickly if things turn sour...........but not so close that it might get you out of a trade too soon just due to a little volatility. The nice part about these kinds of stops is that they are often automated, if your broker will allow them.
The question specifically is asking how much loss should I tolerate from a high point (or peak) in a trade to a low point in the trade? Or to say it another way, how much of a percentage drawdown can I accept before I pull the plug on a trade?
To answer the question.....it depends. Some stocks are volatile and need a lot of wiggle room, where other stocks are consistent and should be played tight. Indexes can behave both ways. So the question really is how much room do you need to allow for your style of trading? We back-tested a variety of stocks and indexes, and the optimal trailing-stop percentage for us was anywhere from 4% to 12%. But is seems like the best overall results occurred when we used 6%. For an index, it was slightly less.
Did we just give away a proprietary secret? Nah - anybody with back-testing software could have done the same. Plus, getting the right trailing-stop is only one of numerous things that has to be 'just so' for system-based trading. However, we hope that little tip helps.
Question 2: Could you explain when it's appropriate to use the "simple moving average" as opposed to the "exponential moving average"? Is it important to make a distinction between the SMA and the EMA or can they be used interchangeably? Thanks! I enjoy your educational columns.
Answer: It's actually a pretty important question. We use both simple moving averages (where all the bars or prices used in calculating the average are weighted equally) as well as exponential moving averages (where the most recent prices or bars are overweighted when calculating the average). As you may guess, the exponential moving average, or EMA, is quick to respond, while the simple moving average, or SMA, is relatively steady. So to answer the question, yes, it's important to make a distinction, and no, they really shouldn't be used interchangeably. Otherwise, you might miss the best feature(s) of one type or the other.
To figure out which style you want to apply in your case, think about your particular method of trading. Do you want to respond quickly to recent changes in an effort to catch a piece of new momentum? If so, an EMA will give you a lot of responsiveness. The downside is, you have a very volatile moving average line that could lead you to a lot of fakeouts. On the other hand, if you apply an SMA to swing trade signals (lasting only a few days), by the time you got a signal from your moving average line, the "swing" may be over. You have a low-volatility line, but it's slow to generate meaningful crossovers....or whatever your signal is. That's the trade-off - responsiveness versus stability. If you need to act fast, use an EMA. If you specifically don't want to react too quickly, use an SMA.
We like to apply EMAs to our short-term charts and trades, and SMAs to our longer-term (multi-week) charts and trades. There's no rule of thumb behind the reason we do so - just years of observation.
BigTrends.com
(This message is linked to previous articles.) |
