Trends

megamuel

Experienced member
Messages
1,098
Likes
164
Just a quick question... Suppose you are intending on following a trend for approximately 1 week, how long would you want the trend to have been established for prior to this? 1 month, 3 months, 6 months? I know there is no clear cut answer to this but just wanted you opinions. Thanks,

Sam.
 
Just a quick question... Suppose you are intending on following a trend for approximately 1 week, how long would you want the trend to have been established for prior to this? 1 month, 3 months, 6 months? I know there is no clear cut answer to this but just wanted you opinions.
Hi Sam,
There may in fact be a clear cut answer, it's just that mine will be different to yours which will be different to the next person's etc. It goes without saying - but I'll say it anyway - that the earlier you jump on a trend and the later you jump off of it - the bigger your profit will be. Conversely, the more confirmation you need that a trend is in full swing - the later you'll join it and the smaller your profit will be. Therefore, for me, your question isn't one I'd ever try and answer. IMO, the questions that one needs to address are these:
1) Is the instrument trending according to my definition of what a trend is?
2) Has my tried and tested set up appeared for jumping on board the trend at the point that offers the minimum possible risk for the greatest possible reward?
When you can answer 'yes' to both questions, place your order along with your stop loss and trade. The length of time the trend has been in place isn't an issue, so long as the trend is strong and there are no signs of it coming to an end. Same applies to the length of the trade. Why get out after a day, a week or month if the trend is continuing in your favour? Getting into a profitable trend trade is the hard bit; having achieved that, why not ride the baby for every cent she's worth? Elliot Wavers and peeps into to cycle and time analysis may take issue with my comments, and that's fine if you use these concepts to define what trend is, how strong it is and when it's coming to an end.
Tim.
 
Just a quick question... Suppose you are intending on following a trend for approximately 1 week, how long would you want the trend to have been established for prior to this? 1 month, 3 months, 6 months? I know there is no clear cut answer to this but just wanted you opinions. Thanks,

Sam.

Following a trend assumes there is already a trend in place. Then you need to calculate the 5-day ATR. Make sure your stops are set at least twice that value otherwise you run the immediate risk of getting stopped out before 5 days. Keep in mind that this does not assure you will not be stopped out due to a sudden reversal for example. But if things continue normally with the same volatility, you will be able to ride the trend for at least 5 days.

I hope my idea helps. I am sure others here may have better ideas because they are more experienced with trend following than me.

Alex
 
Just a quick question... Suppose you are intending on following a trend for approximately 1 week, how long would you want the trend to have been established for prior to this? 1 month, 3 months, 6 months? I know there is no clear cut answer to this but just wanted you opinions. Thanks,

Sam.

I got into long-term trend-following after reading Mark Shipman's book "The Next Big Investment Boom". His specific comments on commodities are looking a little outdated in the current market, but long-term, he's probably not far wrong.

By definition, if a market is not long-term trending, then you can't use this method, and your only choice is to remain out of the market until it is. This is perhaps no bad thing.

Anyway he offers two specific methods, which he does not claim are particularly original which I attempt to paraphrase:-

1) Using a 40-week MA:

Only invest in, or stay invested in a market where the current weekly closing price is above the current 40-week MA and the trend of the 40 week MA is upwards.
(Looked at another way: you should never invest or stay invested in a market where the current weekly closing price is below the current 40 week MA -
And you should never invest or stay invested in a market where the trend of the 40-week MA is down.

[The above is clearly for long positions - if going short, then the reverse directions obviously apply].

He cautions against closing profitable positions too soon, advice you will see on this forum also. He notes that really long trends of this nature don't occur that often, but when they do, you want to capture as much of them as possible.

2) - Based on a strategy developed by Richard Donchian - the price channel breakout.

Donchian's weekly rule was: buy whenever the market moves above the highest price of the last 4 weeks. Hold until the price moves below the lowest low of the last 4 weeks, when the position was reversed. (Thus you capture uptrends and downtrends).

To avoid "whipsaws", a longer period is used and he suggests 12 weeks, thus his 2nd method is summarised as:

Establish a position when the current week's closing price is the highest price of the last 12 weeks. (This is known as an "upside breakout"). It is guaranteed to get you into every long-term trend.

You can of course combine methods 1 and 2: Buy when the current weekly close is the highest price of the last 12 weeks and it is above the current 40 week MA, and the trend of the 40-week MA is upwards.

Hold as long as the current weekly close is above the current 40-week MA and the trend of the 40-week MA is upwards.

Almost by definition, there is no way you are going to capture all of the trend - the aim is simply to get as much as you can.

His exit strategy is slightly more complicated (but not that much), and he includes a lot more about establishing what condition the market is in to begin with ("Stage Analysis"). The smart money gets in on trends before everyone else has noticed, and by the time everyone and his postman has noticed it, it is time to get out. Property was a good case in point.
 
You may also wish to use the idea of trendline touches. Essentially, you need two lows (highs) on two separate cycles to draw an upward (downward) trendline. The trendline is then confirmed with a third low/high touching and bouncing off it. Whether you wish to anticipate the third touch or wait until it happens is up to you.
 
hmmm :rolleyes:

if you're trying to catch a trend at the fourth touch isn't that kind of a bit late ?

a bit risky perhaps ? like jumping on a train whilst it's moving ?

dd
 
Better than jumping on a train which immediately reverses away from your intended direction of travel :cheesy:
 
Top