FTSE system

[pssonice]

Dow 1984 ... 2007

with reference to last day (HIGH+1) and (LOW-1)

1 863 sum 2492 330 -1
2 574 No.loss 539 220 -2
5 209 No.gain 1260 58 -5
10 50 NO.Trade 1258 4 -10
15 16 No.Long 1058 0 -15
20 10 No.Short 729 0 -20
OTG NoOTG NTL OTL

OTG : One Trade Gain i.e. [863 times 1 point gain ]
OTL : One Trade Loss i.e. [1258 times 4 point loss]

 

[smccreedy]

Great well as I've said before the only way to test these things properly is to trade them and make sure you follow every signal.

The stop and reverse I found didn't work too well for my data, I ended up losing both ways although if your stops are further apart than mine it is less likely it'll works against you I guess.

Getting filled is the problem but with the market I am using and the way I'm doing it so far it has been alright.

Don't let the fact it is simple but effective put you off. I think often a lot of people only believe the most complicated systems can work.

The Donchian system of buying or selling at six month highs/lows with effective bet sizing strategy can still out perform many other ways of trading.

For me as I trader I find it helps to imagine that I earn my money from being more discplined than the next guy not for being smarter, having more expensive software, bigger computer, more screens etc.

Best of luck and let me know how you get on.

Stephen McCreedy

 

[juanchi]

But if your both orders are very closed, you will hit one first and many times and when the markets reverses you wil be stop out and miss a big move.

Let's say that the market opens ( or closes) at 6200 and you have 6225 and 6175 orders. if the market goes first above 6225 and then goes to 6100, your system will actually lose money that day. But with a stop and reversal i would make 25 points (6175-6100-(6225-6175))

However, i would got killed if the market for whatever reason goes again to the 6225 area and i have to do a new stop and reversal

 

[RunTheNumbers]

Stephen

Despite my reservations about this system I had a look. You do not appear to be using Yahoo data, which is definitely wrong for its recording of highs and lows, and so the methodology is worth looking more closely at.

I disagree with you in your assertion that the opening is not important. For instance if the market opens above your target long entry then you have a choice. Either you go long at the open, or you record a no-trade and do nothing.

I re-worked your spreadsheet, using my data which I get from Sharescope. This does reflect the fact that the open does not match the close except in unusual circumstance, but as I have never traded based upon the open I would need to test whether this was in fact the case. I make the assumption that if the opening is higher or lower than your long entry or short entry points then the system goes long / short respectively at the market opening price.

The fact that you don't know, in a volatile day where both entry triggers are hit, which comes first, I figure is irrelevant, as clearly either trade would have been stopped out when the opposing trigger was hit.

I have also used a spread of 2 points for th DAX and FTSE and 4 points for the S&P and DOW. I have not allowed for any slippage, which would occur if you had to open a trade at the market opening, but would likely not if the order was a limit order.

To my surprise the system does seem to work on both the FTSE and the DAX, although not at all on the DOW and would work if the entries were reversed on the S&P. I have a niggling doubt though, so I've attached the spreadsheet for review and would be interested in whether anyone can spot any flaws.

Regards

Ben

 

[Hoggums]

Stephen

Despite my reservations about this system I had a look. You do not appear to be using Yahoo data, which is definitely wrong for its recording of highs and lows, and so the methodology is worth looking more closely at.

I disagree with you in your assertion that the opening is not important. For instance if the market opens above your target long entry then you have a choice. Either you go long at the open, or you record a no-trade and do nothing.

I re-worked your spreadsheet, using my data which I get from Sharescope. This does reflect the fact that the open does not match the close except in unusual circumstance, but as I have never traded based upon the open I would need to test whether this was in fact the case. I make the assumption that if the opening is higher or lower than your long entry or short entry points then the system goes long / short respectively at the market opening price.

The fact that you don't know, in a volatile day where both entry triggers are hit, which comes first, I figure is irrelevant, as clearly either trade would have been stopped out when the opposing trigger was hit.

I have also used a spread of 2 points for th DAX and FTSE and 4 points for the S&P and DOW. I have not allowed for any slippage, which would occur if you had to open a trade at the market opening, but would likely not if the order was a limit order.

To my surprise the system does seem to work on both the FTSE and the DAX, although not at all on the DOW and would work if the entries were reversed on the S&P. I have a niggling doubt though, so I've attached the spreadsheet for review and would be interested in whether anyone can spot any flaws.

Regards

Ben

Very interesting - take a look at the DAX period Oct 04 to Apr 06 - the system makes no pips whatsoever.

I decided to take a closer look at this period. I've isolated it in your spreadsheet and I've put in some money management. The system will bet 1000th of the account size per pip on each trade.
Tke a look at the graph - I start the run in Oct 04 and it takes until 19th May 06 before the system gives any real return.

Most people would have given up long before they get to this point. ..

 

[rtsaur]

I'd prefer to leave the trade overnight if the stop-loss threshold is not reached. It should be quite safe if you use the guaranteed stop losses.

 

[RunTheNumbers]

I'd prefer to leave the trade overnight if the stop-loss threshold is not reached. It should be quite safe if you use the guaranteed stop losses.

Can I ask why? Remember the system exits

a) if the stop loss is hit; or

b) the close of the cash market on the day.

If you hold overnight what will be your target/non-stop system exit?

Ben

 

[TopIQ]

"If you want to know the secret of making money in the stock market, it is this: KILL YOUR LOSSES. Never let a stock run against you more than three-quarters of a point, but if it goes your way, let it run. Move your stops up behind it so that it will have room to fluctuate and move higher"
Old E. H. Harriman 1912

Need a mechanical system because your inner personality of fear and greed eat away at you inside, and makes you act at exactly THE wrong moment such as buying the very top or selling the very bottom.

Breakout systems are not just for when your in jail...hehehe!

 

[RunTheNumbers]

"If you want to know the secret of making money in the stock market, it is this: KILL YOUR LOSSES. Never let a stock run against you more than three-quarters of a point, but if it goes your way, let it run. Move your stops up behind it so that it will have room to fluctuate and move higher"
Old E. H. Harriman 1912

Need a mechanical system because your inner personality of fear and greed eat away at you inside, and makes you act at exactly THE wrong moment such as buying the very top or selling the very bottom.

Breakout systems are not just for when your in jail...hehehe!

I'm sorry. I don't wish to be rude, but I have no idea what you mean.

Are you commenting on this system, or all mechanical systems? If you are suggesting that all mechanical systems are a waste of time then why on earth would you think that the people who read boards devoted to discussing mechanical systems would be interested in your poorly worded ramblings? If you don't have anything constructive to contribute then please don't feel the need to write anything at all.

Regards

Ben

 

[rtsaur]

Ben

There is nothing to say if the stop loss is hit. If not, we can hold the position and adjust a new stop loss based on your own exit strategy, e.g. max( Close*(1-0.25%) , Close[1]*(1-0.25%) ) for long position or min( Close*(1+0.25%) , Close[1]*(1+0.25%) ) for short position in Stephen's FTSE system.

PS1. there is no target/non-stop system exit since we hold the position till the adjusted stop loss been hit.

PS2. we may add more positions later.

Can I ask why? Remember the system exits

a) if the stop loss is hit; or

b) the close of the cash market on the day.

If you hold overnight what will be your target/non-stop system exit?

Ben

 

[Zish]

Stephen

Despite my reservations about this system I had a look. You do not appear to be using Yahoo data, which is definitely wrong for its recording of highs and lows, and so the methodology is worth looking more closely at.

I disagree with you in your assertion that the opening is not important. For instance if the market opens above your target long entry then you have a choice. Either you go long at the open, or you record a no-trade and do nothing.

I re-worked your spreadsheet, using my data which I get from Sharescope. This does reflect the fact that the open does not match the close except in unusual circumstance, but as I have never traded based upon the open I would need to test whether this was in fact the case. I make the assumption that if the opening is higher or lower than your long entry or short entry points then the system goes long / short respectively at the market opening price.

The fact that you don't know, in a volatile day where both entry triggers are hit, which comes first, I figure is irrelevant, as clearly either trade would have been stopped out when the opposing trigger was hit.

I have also used a spread of 2 points for th DAX and FTSE and 4 points for the S&P and DOW. I have not allowed for any slippage, which would occur if you had to open a trade at the market opening, but would likely not if the order was a limit order.

To my surprise the system does seem to work on both the FTSE and the DAX, although not at all on the DOW and would work if the entries were reversed on the S&P. I have a niggling doubt though, so I've attached the spreadsheet for review and would be interested in whether anyone can spot any flaws.

Regards

Ben

Based on your numbers, the system is down 500 points on the FTSE since 20 June 06! Could this be a sign the system is no longer working? Or is this an acceptable level of risk. It seems if someone started trading this system last summer, they'd be wiped out by now.

 

[keavenen]

does this work for day trading and scalping? not sure I get it totally. so at the open if price moves up...buy or if it moves down sell. seems simple??? prob does work aswell but this might only get you 10 ticks or so at the open.

please state fully your trading day. thanks

 

[RunTheNumbers]

Based on your numbers, the system is down 500 points on the FTSE since 20 June 06! Could this be a sign the system is no longer working? Or is this an acceptable level of risk. It seems if someone started trading this system last summer, they'd be wiped out by now.

You might argue that the system has averaged over 3,000 points per annum on average for 8 years. So a drawdown of 500 points is only between 15-20%. Given this is a 10 minute per day system, then with sensible bet-sizing the system would not yet be hurting too much.

Plus I believe it would be crazy to trade this alone. I would look to assemble a basket of markets which it worked on so as to diversify out any significant risks.

I should state that I do not trade this and have only done the relatively cursory research into it that I have posted on this board.

Regards

Ben

 

[Zish]

Hi Ben, thanks for the reply. I was wondering if you could answer something for me. When testing this system on past data, how do you know whether a trade is triggered just by looking at the high/low for the day. Wouldn't you need to see how price moved throughout each day in order to establish whether a trade was in fact triggered, or whether the trade was stopped out instead.

i.e. say FTSE is at 6000 with corresponding entry/exit points at 6015/5975

if the high for the day is 6025 and the low is 5970, how do you know which trade is triggered first unless you look at the detailed price action for each and every day?

Apologies for my basic questions but i'm new to all this

 

[rtsaur]

Yes, you have to use the historical real-time data to back test this system.

Hi Ben, thanks for the reply. I was wondering if you could answer something for me. When testing this system on past data, how do you know whether a trade is triggered just by looking at the high/low for the day. Wouldn't you need to see how price moved throughout each day in order to establish whether a trade was in fact triggered, or whether the trade was stopped out instead.

i.e. say FTSE is at 6000 with corresponding entry/exit points at 6015/5975

if the high for the day is 6025 and the low is 5970, how do you know which trade is triggered first unless you look at the detailed price action for each and every day?

Apologies for my basic questions but i'm new to all this

 

[Zish]

Yes, you have to use the historical real-time data to back test this system.

Thanks. Where can I get this historical real time data from?

 

[RunTheNumbers]

Yes, you have to use the historical real-time data to back test this system.

I'm not sure I agree. One can work through the various combinations.

1. The high for the day is higher then the long trigger, and the low for the day is higher than the short trigger = a long trade is entered and closes out at the end of the day.

2. The low for the day is lower than the short trigger, and the high for the day is lower than the long trigger = a short trade is entered and closes out at the end of the day.

3. The high for the day is higher than the long trigger, and the low for the day is lower than the short trigger = two possibilities, the long trigger is hit first and a long trade is entered, but the system stops out when the short trigger is hit, or the short trigger is hit first and a short trade is entered, but the system stops out when the long trigger is hit. Either way the system is stopped out for a loss of the difference between the long and the short triggers.

4. The high for the day is lower than the long trigger and the low for the day is higher than the short trigger = No trade that day.

I therefore do not see why one needs intraday data.

Regards

Ben

 

[Zish]

Ahhh of course - thanks ben

I'm not sure I agree. One can work through the various combinations.

1. The high for the day is higher then the long trigger, and the low for the day is higher than the short trigger = a long trade is entered and closes out at the end of the day.

2. The low for the day is lower than the short trigger, and the high for the day is lower than the long trigger = a short trade is entered and closes out at the end of the day.

3. The high for the day is higher than the long trigger, and the low for the day is lower than the short trigger = two possibilities, the long trigger is hit first and a long trade is entered, but the system stops out when the short trigger is hit, or the short trigger is hit first and a short trade is entered, but the system stops out when the long trigger is hit. Either way the system is stopped out for a loss of the difference between the long and the short triggers.

4. The high for the day is lower than the long trigger and the low for the day is higher than the short trigger = No trade that day.

I therefore do not see why one needs intraday data.

Regards

Ben

 

[Hoggums]

I'm not sure I agree. One can work through the various combinations.

1. The high for the day is higher then the long trigger, and the low for the day is higher than the short trigger = a long trade is entered and closes out at the end of the day.

2. The low for the day is lower than the short trigger, and the high for the day is lower than the long trigger = a short trade is entered and closes out at the end of the day.

3. The high for the day is higher than the long trigger, and the low for the day is lower than the short trigger = two possibilities, the long trigger is hit first and a long trade is entered, but the system stops out when the short trigger is hit, or the short trigger is hit first and a short trade is entered, but the system stops out when the long trigger is hit. Either way the system is stopped out for a loss of the difference between the long and the short triggers.

4. The high for the day is lower than the long trigger and the low for the day is higher than the short trigger = No trade that day.

I therefore do not see why one needs intraday data.

Regards

Ben

You need intraday data because you do not know the entry point otherwise. Take today on the FTSE for example - there is no way you could have got filled for less than 6250 at open - that's 47 points above what YAHOO says the open is rather than the 15 or so points your system tries to target. Even at 9pm last night IG were quoting the FTSE at +25 pnts over the close when I opened my position. Let's say for arguments sake the market later took a tumble to stop out at 6187 - that's a loss of not 30 points as your spreadsheet would say - but 63 points (ignoring spreads).

I have also found that the SB overnight price than peak more than 15 points only to go the other way before market open because of a rise or tumble in the opposite direction on the asian markets. If that happens you will make a loss even though your spreadsheet may say you've profited.

 

[Zish]

So this system isn't workable in real life? It seems the difficulty in getting filled at the price you want renders it unprofitable.

smccreedy did say that he hasn't had any probs getting filled at the price he wants thus far. Maybe he'd like to comment?