Trend followers: how good has the last quarter been?

[greg10]

Hello,

this is my first post so please be gentle if I breach any etiquette.

I've been a spread-betting trend follower on and off for a few years and began again in earnest at the start of September 2008. My target is an annual return of 20%+. The logic behind the target is that compounding 20% for 20 years = a pleasant pension.

Three months in and it's going well - I'm betting 1% of capital per trade and I'm on course to meet my target. I plan to increase to 2 or maybe 3% per bet in the coming months as my confidence grows.

With the economy in such turmoil, I find it really difficult to judge whether my success is just down to the last three months being exceptionally kind to trend followers, or whether I really can be optimistic of long term success.

If you've been trend following for a number of years, I'd really like to hear your assessment of the last quarter.

Thanks,

Greg

 

[montmorencyt2w]

Greg,

What markets are you trading as a matter of interest?


I started just under a year ago, trend-following (or at least I thought that's what I was doing), and after blowing my (very small) account a couple of times, corrected the more obvious of my mistakes and started to make a bit of dosh.

However, I started getting greedy, and began to lose (although I was still in profit), so I just got out, for a while, in order to take a break, and get my head in order. The break was a bit longer than originally intended, but I restarted just over a month ago, or so.

Whether it was just that I'd got rusty, and forgotten the lessons I'd learned, or just that the markets seemed to have changed so much, or what, but I found it much harder coming back. Perhaps I now knew enough to know how little I really knew!

Consequently, I'm only currently trading in a small way, in order to re-learn the craft. Still essentially trend-following, but keeping myself open to new ideas and new techniques.

I'd be interested to know which markets you have found successful.

When I was starting, oil and gold were very predictable, as was GBP/EUR and USD/EUR, although I made most money on AUD/various (for a while). That all changed after the credit crunch...


I was initially inspired into trend-following by Mark Shipman, and then by Michael Covel.

It's probably the simplest approach there is, and probably the most difficult to actually succeed with.

Not meaning to dampen any enthusiasm here, but I'd be really cautious about increasing your bet size if I were you (knowing what I know now). It can really mess with your head if you are not very careful. Try to make sure that your technique works in a range of markets in a range of conditions, or at any rate, try to become as fully aware as you can of its limitations, and its bound to have some.


Regards,
M.

 

[greg10]

M.,

I trade commodities and currencies but I have no preferences other than that (i) it is offered by CMC for spread betting and (ii) it is covered by the Maddock Data daily data set. I'll probably include indices and treasuries in due course, too, but I don't have the stomach for shares. I use a very (probably over-) simple Donchian algorithm. I too am inspired by Covel and also Curtis Faith. I've read a few others but not found them as inspiring. I think the key to my 'success' so far has been a distant initial stop loss that has avoided my being stopped out too often.

To give an idea of numbers, I've made about 4.5k but it's been rather volatile: I've had several £900+ days and one -£2000 day. Most days I've had half a dozen or so bets running but that's crept up in the last couple of weeks to fourteen as more and more new highs or lows are cropping up. Fourteen sounds too many to me, too, but I haven't put the work in yet to decide a sensible maximum.

Greg

 

[montmorencyt2w]

Well Greg, I think you are a bigger/more successful player than I got to be, although I think I did manage to lose about £1000 one day ... probably about when I decided to step back for a bit!

That does sound like a lot of concurrent positions, but on the other hand, if you are holding them for quite some time, as I imagine you are, it's not so bad.

These days I'm effectively day-trading (something I swore I'd never do ), and some days practically scalping over very short time scales, and under those conditions, I can't manage more than one at a time

I'd like to get back to a slower, more relaxed, less stressful way of doing it really. But as I've said elsewhere, I'm currently trying out various approaches and "systems", with and without indicators, etc. I did start out using Mark Shipman's version of Donchian's channnel breakout method, based on trading on the highest high (or lowest low) of last 12 weeks, on weekly charts. And/or trade on above/below 40 week SMA, with uptrend/downtrend MA direction. It seemed to work well enough at the time, but at the time I got back into the game, I couldn't see any markets that I was familiar with that seemed to lend themselves to that approach. Maybe I wasn't looking properly


Regards,
M.

 

[montmorencyt2w]

Greg,

You may be vaguely interested in taking a look at this:

EOD trading system

It's a "mechanical" system based on EOD, and is I believe, effectively a form of trend-trading. I haved not traded it myself (I might borrow some of their indicators though ), but it is not dissimilar to one approach I am currently using. They use stochastics and CCI, with some special indicators. I use stochastics and MACD to do something similar, but also confirm with CCI and OBV and maybe volume. Mine is a bit more discretionary though, and I've been doing it on a shorter timeframe.


Regards,
M.

 

[greg10]

Thanks, M, that was a new site for me: I do need to spend some time reading around.

I'm not up to speed with all the acronyms yet but I think there are a few key differences with my approach:

1. I don't wait for a retracement - if yesterday was a 25 day high or low then I buy or sell today.
2. I have no target price in mind, instead I have a 10-day low or high trailing stop.

I think it is probably the fairly distant initial stop that has given me the luxury of simple signals.

Like that strategy, I do just have a once-per-day intervention, complicated only by the trading hours of some of the US commodities. Like a lot of trend following, the median bet loses money but for now, anyway, the mean is positive.

As I write this I seem to be having my best day ever despite being stopped out of some bets. The drawback with my algorithm is that the stops get left far behind when there are big movements so it can all disappear next week.

I suspect that I've just been lucky on hitting upon simple parameters (25 day extreme for entry, 10 days for exit) that have worked for the last few months. I will back test to see how often that has been the case just as soon as I finish writing some back test software...

Greg

 

[davidnmfarrell]

Hi Greg,

This is an interesting question. I have been trading my Trend-based system since November 2008, and so far I have returned 23% from a win percentage of approx 30%.

I have a proprietary indicator that I use, which is based upon several MAs and I too use an MA as my stop loss. I only trade major currency pairs, and adjust my trades once per day.

Regarding whether or not the system would have made money historically, I have back-tested the system over several years worth of data and markets, and it was shown to be very successful (but due to back testing flaws, the results are over-stated). It also showed profit in Commodities, but not nearly as much.

Cheers,
David

 

[megamuel]

Care to share your system then David?!

Sam.

 

[Dacamic]

With the economy in such turmoil, I find it really difficult to judge whether my success is just down to the last three months being exceptionally kind to trend followers, or whether I really can be optimistic of long term success.

Being unfamiliar with your system and markets, I do not know whether you are currently sailing in very favorable weather. As a suggestion, you should consider evaluating your system over several years (as David did for his) to help answer that question yourself.

 

[meanreversion]

I've found using MACD as a trend filter on a longer time frame is useful, e.g. if MACD line crosses over signal on daily chart then take this as a change in trend for a shorter time frame and use oscillators there to get on this "trend". This thread is months old, how have things been going for you in the meantime greg10

 

[rogue8]

Alright, so I'm late replying to this thread (just saw it today!), but here's some food for thought...

Don't know about the last quarter up to the date of the initial post, but a 35 day period from Wednesday 05/11/2008 to Tuesday 23/12/08 would have yielded a 3448.8% return with a simple CCI-indicator trend-following system (and some people turn their noses up at indicators!). This is for the EurUsd on a 5-minute chart (no optimisation/curve fitting of CCI value: just an arbitrary but reasonable value) with strict entry/exit rules, and strict trade and money management rules (most important ingredient). If rules are followed precisely, allowing for some margin of error + minor slippage, there is very little room for ambiguity in the results. So £10,000 yields around £345,000 net profit (4056pips after 0.9pip spread per trade) with a fixed 2% Risk per trade (162 trades in all).

Take a look at the Daily FX charts like EurUsd when it tanked from late July to late October 2008. Some people (trend-followers?) must have made a fortune from that. Are any of them here treading these T2W boards??... Unfortunately it's a rare event and not a good benchmark.

Now compare the results of the same CCI system for the same period in 2009 (Thursday 05/11/2009 to Wednesday 23/11/09)... A more down-to-earth 284.5% over 212 trades (1052pips) so £10,000 would yield around £28,500. I'm not sure what percentage the prop-desk pros normally return for 35 working days but I think this is a good amount. (N.B. Entry and exit signals, as I've read and found out myself in practice, are not actually that important).

So basically, trend-following appears to still be profitable but only with discipline, persistence through losses and adaptation to changing/worsening markets.

Happy hunting!
:shuriken:

(Results are from empirical data)

 

[meanreversion]

Any trend following system trading FX should have made money in February. Aside from EUR/USD, the move in GBP/USD was fairly forceful as well. Michael Covel's books on trend following are worth a read.