What is decent return?

Hmmm that's a pretty good response, entirely logical. Of course you are right, trading live is on "out of sample" data.

I'm not completely sold on the merits of using walk-forward testing (on known out-of-sample data). It looks pretty sexy, and I can see how it probably shows up very flaky systems, but it adds a whole new layer of complexity to the testing process.
 
The merits should really be independent of how easy it is to accomplish with your tools? It's very easy with my software.

It can tell you so much about your optimisation that you really want to know before trading live. I don't see how forward testing can be left out at all - certainly for the way I develop systems. I forward test in historical data, and then let systems forward test paper trading (all my systems are automatic, which I guess makes things a lot less effort) as a last check before going live.
 
I suppose you're right for a system like the opening posters, meanreversion. One reason I think forward testing (or whatever the correct term for it is) is more useful than testing on another set of data, is that it can show you how hard the system is to actually trade.

Example being, If you try to trade a system that enters on a moving average, because you've seen price appears to bounce off it. When you go to the paper or demo trading test, you will notice that some of those entries that looked perfect on past data, weren't even possible to enter, because the moving average was in a different place until the entry bar actually closed.
 
What I will say about mechanical systems is that you can test them until you are blue in the face (and you probably should), but the hardest part is the implementation.

By this I mean adhering to the rules, despite everything. This is the tough bit. The most likely way a mechanical system will fail to perform as the trader might want is due to human error, i.e. skipping signals.

I'll give you a couple of examples. I once had a buy signal in AUD/USD, but the previous 5 trades had been losers, so I skipped it. Ended up being a good trade, cost £15k or so.

I was away at the weekend, and missed a signal to sell soybeans. I "should" be short but am not. It would cost me £5k to get onto the trade now (relative to the system). Should I do it? It might retrace and not be an issue, or it could become the trade of the year, who knows.

In practice it's easy to miss signals, and for the most part I simply enter at a worse rate and put it down to slippage. But with the soybeans, I'm hesitating slightly as it's a meaningful amount of cash, and I'm fully loaded on several other positions as well. It's not easy..
 
wantochangemylife said:
I started with £500 and placed bets at £100 per trade.
Click to expand...

Does that mean you actually risked losing one fifth of your funds per stop loss hit ?

If that's true I guess what you might want to have a pretty close look at with that kinda position sizing is this:

http://traderfeed.blogspot.com/2010/01/how-to-avoid-risk-of-ruin-in-trading.html

You can have a super duper system with I dunno a 90% win rate or whatever, but even that will go through losing streaks when it hits the inevitable outlier events that are part and parcel of a traders career.

Risking 20% per trade as you are if I got that right your long term odds of survival are realistically zero.

5 losing trades in a row - which can always happen - and you've pretty much blown your account.

http://ftalphaville.ft.com/blog/2009/07/28/64176/dragon-king-of-the-outlier-events/

Just another old cliché but true nonetheless, there are old traders, bold traders, but no old and bold traders.

I think people always overestimate to their own detriment what they can do in the short term, while severely underestimating the long term power of compounding.

Good luck tho as so far the Gods have smiled on you.
 
BSD said:
I think people always overestimate to their own detriment what they can do in the short term, while severely underestimating the long term power of compounding.
Click to expand...

Wow, that hits the nail firmly on the head. People overestimate their tolerance for drawdown, yet fail to realise that +20 pct a year compounds quickly... i.e. they want to get rich overnight.
 
If you are trading once per day and the longest run of losers is 6 days - you MAY be doing something right, although generally speaking there are a number of data points that are needed to evaluate the system. For example, it'd be good to know average size of winners, average size of losers, percentage winners, percentage losers, average time in a winning trade, average time in a losing trade.

The whole $100/trade is something people shouldn't focus on - as a novice, you can't be expected to come up with the optimal position sizing approach for that strategy.

To the OP - the answer to your question 'is it worth it' is - MAYBE. There's not enough info to tell.

If you want - PM me and I'll send you my MSN/Skype ID and we can have a chat about it - there's no rocket science here - a lot of the information is in the public domain & it'll probably take 20 minutes to point you in the right direction.
 
Hello All,

Ok I have been tuning my system and it makes money- I modelled it on starting with £700 and betting £20 a trade. The only time (in 8 years of backtesting) I got totally wiped out was once in 2009when the ftse rose 11 consecutive days!

I just want to thank everyone for the help/links and offers of help. DIONYSUS TOAST ..thank you for your offer of a chat I will take you up on that at some point soon. As said I am a novice and not to proud to learn from other people!.
 
Please read "Trade your way to financial freedom" by Van Tharp (this is a must read for any trader). Then, if you're into coding, read the book about mechanical trading systems by Richard Weissman. The two will set you back £80-100 but are well worth it.
 
  • Like
Reactions: BSD
Just stick to 2% risk,..and you needn't fret !
Why folk get greedy is beyond me,.it only creates fear
Don't pass to the dark side young Jedi
:cheesy:
 
2% risk per trade I would consider a lot, but it depends on the trading style. For a breakout system, 2% is dangerous --- with a low win rate, a very uncomfortable drawdown is likely even if the system is ultimately profitable.

Beware though --- too LITTLE risk means you won't make enough money either.
 
Weeelllllllllllll, I don't mean to be a pedant, but a percent is quite a big difference. For example, on my system, the max drawdown over the last 10 years is 52 pct, with an initial stake of 0.5 pct. If I increase that to 1 pct, the drawdown goes up to 84 pct.

I don't think anyone can take 84 pct drawdown... 50 pct is bad enough! (Although the annual return for the 1 pct is much higher)
 
wantochangemylife said:
I have been testing a sytem for predicting ftse closing up or down using snapshot data.

I have back tested this theory on every day for 3 years worth of trading.

I started with £500 and placed bets at £100 per trade.

Year 1 - proft £3600
Year 2 - profit £1100
Year 3 profit £1700

It works BUT it involves me making a trade everyday and whilst it clearly works is this a viable return for the effort? C

THanks
Click to expand...

are u using ftse cash data for your backtesting?
 
Hairy palms,.mad and blind,. haven't got a lot going for me, huh !
Anyway, get back to work, slave ! : )
 
You must log in or register to reply here.

Similar threads

Q
trading spreads without using stops
2
Replies
24
Views
8K
inktomii
I
F
Forex Managed Accounts - Stable Returns
2
Replies
24
Views
10K
anley
A
L
SB Return Rate on IG Index
Replies
0
Views
401
lending900
L
T
Is 2% return per week realistic?
4 5 6
Replies
112
Views
37K
ffsear
ffsear
R
  • Poll
What to do in the situation of undercapitalization?
Replies
5
Views
2K
Akuma99
Akuma99
Top