FTSE system

[JTrader]

Let me see if I can clarify. A futures contract can be traded whereas a Cash index contract cannot be traded because there is no such thing. You need to understand (forgive me if I am being condescending) exactly what a futures contract is. In basic terms, it is a single instrument that gives exposure to the the entire underlying FTSE 100 cash index. Because it is used as a hedging vehicle it must perform in the same way as the Cash index, otherwise there is no point in using it as a hedge.

If you sign up to a broker that gives you direct access to CME Globex (Where E-mini's and many other contracts are traded electronically) then the futures quotes you get will be the same for anyone trading the E-mini through any broker. Which should be the same as the historcial tick data, which in fact it is. I purchased 2006 tickdata to check my trading and there is no difference.

I also use IG index and CMC markets SB and can say that the S&P500 Futures quotes IG Index give match the E-mini quotes I get from my Direct Access Broker, CMC are different as they quote 0.1 ticks like the big (PIT) S&P500 futures.

So what am I saying? If you are serious about trading an Index using a Mechanical strategy then you must be able to rely on both your quotes and historical data right down to the tick.

Hi
yes i understand that you cannot trade contracts in a cash index.

The point I was making is, can we rely on an SB daily index futures bet quotes, to follow the underlying futures price any more closely than we can rely on the SB daily index cash bet quotes to follow the underlying cash index value on an intraday basis.

 

[new_trader]

Hi
yes i understand that you cannot trade contracts in a cash index.

The point I was making is, can we rely on an SB daily index futures bet quotes, to follow the underlying futures price any more closely than we can rely on the SB daily index cash bet quotes to follow the underlying cash index value on an intraday basis.

Sorry, I didn't make my point clear enough. The only way to reliably trade a system is if the data you did your test on is identicle to the data you will trade. I trade the E-mini S&P via a direct access broker which is exactly the same as my historical tick data. If you want to avoid impairing the performance of your system then you cannot for example, test using yahoo data and trade with your SB company data. If possible, see if you can purchase/obtain historical data directly from your SB company.

 

[JTrader]

Sorry, I didn't make my point clear enough. The only way to reliably trade a system is if the data you did your test on is identicle to the data you will trade. I trade the E-mini S&P via a direct access broker which is exactly the same as my historical tick data. If you want to avoid impairing the performance of your system then you cannot for example, test using yahoo data and trade with your SB company data. If possible, see if you can purchase/obtain historical data directly from your SB company.

I agree, this is the only way to test a system really accurately - tradeable prices - and the smaller the timeframe, the more important this is likely to become.
However, if you can pretty sure that your retail dealers prices will be there or there abouts, give or take 1-2 pips then with this strategy, I think it is still worth giving trading it a go.................

City Index have a 2 point rolling cash FTSE spread, offer OCO and If Done orders, and endeavour to always be 1 point either side of the underlying cash price from 0800 onwards (according to their salesman). They also deal in .10 point increments. However, if you want to close the trade at 1630, you have to call them or close it online. Otherwise it is rolled over.
This is the only aspect they seem to be lacking in, in terms of functionality. If you have a job, you don't want the hassle of having to phone them every day telling them not to roll a bet over.

If they do ALWAYS endeavour to be 1 point either side of the underlying with their spread, they are probably the best choice to go with compared with WS CS CMC Cantor and IG, on this particular product, at present IMO.

 

[smccreedy]

I couldn't agree more that in an ideal world you really want to test the exact instrument you're going to trade.

However in my opinion stuff like yahoo is decent enough to test rough ideas, if when you test them they are only marginally profitable than I'd bin them anyway. If they look very profitable they might stand a chance when pushed on to the next research level.

If your strategy is down to a tick here or there to make it profitable or not it might not be a very good strategy unless you are seriously close to the market. ie just a few mess ups each year, power outages, bad fills etc could move it from profit to loss.

If you have a strategy like the 0.25% that promises quite a healthy return if you can get your orders on I don't think getting closer to the market should make too much difference.

Ideally I like systems that wouldn't be by the smallest detail but catch the larger moves and look at the bigger picture. Not slating people or provide liquidity / scalp but it's not for me.

Perhaps if you trade the futures but use the cash market to signal when you enter or exit might be a solution.

In my opinion though the future is end of day or even end of week systems.

We just need to raise the capital, find the patience and something to distract us from the fun of intraday trading in the mean time (far too cold for golf yet and Cheltenham isn't for another month!)

Stephen McCreedy

 

[JTrader]

Having thought about it, I've decided that trading this system, on the days when todays open is beyond the long or short trigger level, i would not want to enter immediately at the open price.

I would place my orders at around 0810 in the morning, once the open had begun to settle down.

I would do this because I would want to minimise losses and hopefully maximise profits on at least a few days.
This is based on the assumption that between 0810 and the close, price may cross back and forth over the calculated entry levels on a few occasions.
Lets say the FTSE closed at 6000. using 0.10% to calculate the long and short entries, the long entry for the next day is at 6006, the short entry is at 5994.
If the FTSE opens at 6030, and i entered at this open price, then starts falling back to the 5994 stop-loss, I have lost 36 points.
If I place an order to enter at 6006, and price comes back to enter me into the market, then hits the 5994 stop-loss, i have only lost 12 points.
I may not get entered at all, sticking to the calculated levels.
However, if I do enter at 6006, and the market then heads back up to 6030 at the close, I have made 22 to 24 points. If I had entered at the 6030 open, i have made -2 to 0 points.

As a result these few possible outcomes are -
i) The open was beyond one of the entry triggers, I do not get a chance to enter at the calculated price, and I either miss out on A) making a profit, or B) making a loss. I do not trade on that day.
ii) After opening beyond one of the calculated entry levels, the market comes back to that level, I am entered into the market. The outcome of this is that I either C) make a bigger profit or D) make a smaller loss, than I would have if I had entered immediately at the open.

I would use Use the 2nd-3rd most historically profitable calculated levels on the spreadsheet 0.10% or 0.15%, rather than 0.25% 0.125% etc.
However, in using 0.1% - 0.15% it is more likely that I will not get entered into trades than it would be if I used the 0.25% calculation.

Doing it this way, it may be possible to produce better results in forward testing than were on the spreadsheet.
However, there will be some days (I don’t know how frequent) when your orders do not get filled when the market opens beyond your calculated levels, and continues in that direction, without coming back to your entry level to fill your order/s – but at least you haven’t lost anything on such days!

The only way to find out how this plan would work out, would be to forward test it with real money. If things go well then fine.
The only problem I can see is not getting filled on some days if the market does not return to the trade entry level, having opened beyond it, or having moved beyond it between 0800-0810.
If the SB price and the live cash price were in line before 0810, I could place my orders before 0810..

Because the results are as good on the FTSE100 & Dax, I would plan to run this strategy on both, starting with the FTSE and introducing the Dax if things went well on the FTSE for a month or so. Adding more capital to get the Dax half of my trading plan off the ground. Thus hopefully, achieving my profit objectives in less time, trading 2 markets instead of 1.
Dax orders probably need to be entered an hour earlier than FTSE though, as DAX cash market opens at 07:00 UK time.

Hi Stephen

what do you think about the suggestions I made here, about how to manage this strategy?

What impact do you think this would have in terms of the overall performance?

Thanks
JT.

 

[new_trader]

I couldn't agree more that in an ideal world you really want to test the exact instrument you're going to trade.

However in my opinion stuff like yahoo is decent enough to test rough ideas, if when you test them they are only marginally profitable than I'd bin them anyway. If they look very profitable they might stand a chance when pushed on to the next research level.

If your strategy is down to a tick here or there to make it profitable or not it might not be a very good strategy unless you are seriously close to the market. ie just a few mess ups each year, power outages, bad fills etc could move it from profit to loss.

If you have a strategy like the 0.25% that promises quite a healthy return if you can get your orders on I don't think getting closer to the market should make too much difference.

Ideally I like systems that wouldn't be by the smallest detail but catch the larger moves and look at the bigger picture. Not slating people or provide liquidity / scalp but it's not for me.

Perhaps if you trade the futures but use the cash market to signal when you enter or exit might be a solution.

In my opinion though the future is end of day or even end of week systems.

We just need to raise the capital, find the patience and something to distract us from the fun of intraday trading in the mean time (far too cold for golf yet and Cheltenham isn't for another month!)

Stephen McCreedy

1 tick can make a huge difference to the bottom line irrespective of how you calculate your entry and exits, especially when your BUY stop is triggered and the market immediately retraces towards your SELL STOP. Trust me, this happens so often you think there are traders who have access to your trading account. Each time you "creep" your stops to avoid them being hit, the market chases them. It is uncanny but it happens. If you think 1 tick shouldn't make a difference to a "good" system then you are in for an expensive surprise.

 

[smccreedy]

new_trader

With respect what you say is only applicable to short term strategies.

You seem to only consider a system one that trades so often a tick here or there is make or break.

If you buy on six months highs and sell on six months low a tick doesn't have much of an impact.

You may consider your assumption that everyone who talks systems is talking scalping or day trading.

You seem pretty mad at the market and seem pretty sure I'm in for a surprise which I think is a little wrong when you have no idea how I trade, for how long I have traded or how well it has gone.

You seem to assume that all of use are complete novices and have no experience of operating any system at all. In relation to me this is simply un true.

I guess from your postings you had a nasty surprise at some point and seem to think we are all due one too, would this make you feel better?


JT

I'm sorry not to be more encouraging but my mantra is that if I can't test it I wouldn't like to say how I'd think it would pan out over the longer term.

The strategy you have built bares no resemblance to what the original was and as such I would not like to be drawn on if it helps or hinders the original ideas.

I think missing the day's you can't get your orders on with my system is to remove the big winners that eventually pay your profits.

What you are saying is 100% right, buying lower will give you larger profit or smaller loss but it doesn't address whether the likelyhood of buying a .25% breakout from 8.10 is going to work more often than not, or often enough to pay for the losers.

I don't wish to be negative I'm just saying untill you test it you won't know.

For what it's worth in my experience traders improve when the following happens;

1. They trade less
2. They trade on longer term trends and patterns
3. They adjust their expectations of what sort of return is possible with what they have.

These changes have helped me.

Stephen McCreedy.

 

[JTrader]

new_trader

With respect what you say is only applicable to short term strategies.

You seem to only consider a system one that trades so often a tick here or there is make or break.

If you buy on six months highs and sell on six months low a tick doesn't have much of an impact.

You may consider your assumption that everyone who talks systems is talking scalping or day trading.

You seem pretty mad at the market and seem pretty sure I'm in for a surprise which I think is a little wrong when you have no idea how I trade, for how long I have traded or how well it has gone.

You seem to assume that all of use are complete novices and have no experience of operating any system at all. In relation to me this is simply un true.

I guess from your postings you had a nasty surprise at some point and seem to think we are all due one too, would this make you feel better?


JT

I'm sorry not to be more encouraging but my mantra is that if I can't test it I wouldn't like to say how I'd think it would pan out over the longer term.

The strategy you have built bares no resemblance to what the original was and as such I would not like to be drawn on if it helps or hinders the original ideas.

I think missing the day's you can't get your orders on with my system is to remove the big winners that eventually pay your profits.

What you are saying is 100% right, buying lower will give you larger profit or smaller loss but it doesn't address whether the likelyhood of buying a .25% breakout from 8.10 is going to work more often than not, or often enough to pay for the losers.

I don't wish to be negative I'm just saying untill you test it you won't know.

For what it's worth in my experience traders improve when the following happens;

1. They trade less
2. They trade on longer term trends and patterns
3. They adjust their expectations of what sort of return is possible with what they have.

These changes have helped me.

Stephen McCreedy.

Hi Stephen

but also, these are the days that rack up the big losses when you enter at the opening price beyond the calculated entry levels, and when price reverses, you incur a bigger loss.

Perhaps in your framewrok you need all the big winners to make up for the losers. But if you didn't have the big losers resulting from entering at the open price, with price then reverses, you wouldn't need as many big winners.

 

[new_trader]

new_trader

With respect what you say is only applicable to short term strategies.

You seem to only consider a system one that trades so often a tick here or there is make or break.

If you buy on six months highs and sell on six months low a tick doesn't have much of an impact.

You may consider your assumption that everyone who talks systems is talking scalping or day trading.

You seem pretty mad at the market and seem pretty sure I'm in for a surprise which I think is a little wrong when you have no idea how I trade, for how long I have traded or how well it has gone.

You seem to assume that all of use are complete novices and have no experience of operating any system at all. In relation to me this is simply un true.

I guess from your postings you had a nasty surprise at some point and seem to think we are all due one too, would this make you feel better?


JT

I'm sorry not to be more encouraging but my mantra is that if I can't test it I wouldn't like to say how I'd think it would pan out over the longer term.

The strategy you have built bares no resemblance to what the original was and as such I would not like to be drawn on if it helps or hinders the original ideas.

I think missing the day's you can't get your orders on with my system is to remove the big winners that eventually pay your profits.

What you are saying is 100% right, buying lower will give you larger profit or smaller loss but it doesn't address whether the likelyhood of buying a .25% breakout from 8.10 is going to work more often than not, or often enough to pay for the losers.

I don't wish to be negative I'm just saying untill you test it you won't know.

For what it's worth in my experience traders improve when the following happens;

1. They trade less
2. They trade on longer term trends and patterns
3. They adjust their expectations of what sort of return is possible with what they have.

These changes have helped me.

Stephen McCreedy.

I'm sorry if I gave that impression. You must trade with much wider stops than I do.

(P.S: I don't know why you think a tick here or there makes no difference when it comes to hitting stops)

 

[Hoggums]

1 tick can make a huge difference to the bottom line irrespective of how you calculate your entry and exits, especially when your BUY stop is triggered and the market immediately retraces towards your SELL STOP. Trust me, this happens so often you think there are traders who have access to your trading account. Each time you "creep" your stops to avoid them being hit, the market chases them. It is uncanny but it happens. If you think 1 tick shouldn't make a difference to a "good" system then you are in for an expensive surprise.

New trader has a point - it only takes one tick to trigger a loss of say 50 points. When that happends your real system is 100 points short of what your spread sheet says you should be. Some SBs - particuarly IG - skew their quote a point or two in ahead of the current market direction.

 

[new_trader]

New trader has a point - it only takes one tick to trigger a loss of say 50 points. When that happends your real system is 100 points short of what your spread sheet says you should be. Some SBs - particuarly IG - skew their quote a point or two in ahead of the current market direction.

Exactly. The time frame is irrelevant. I am just trying to point out to anyone developing a system that 1 tick is as important as any. Like the old saying goes, watch the pennies and the pounds will take care of themself.

 

[smccreedy]

JT

I can only speak from my own experience and I only ever traded at the levels I worked out the night before, I managed to get these filled everyday by entering them after the close each day or entering them manually during the day.

I never took the approach that if I wanted to buy at 6330 but it opened at 6350 what to do, it didn't crop up as to have set following day limit at 6330 the market would have closed at 6314. Following that close I would have entered the order to buy at 6330 and would have got a fill at that during the night and been 20 up in the morning.

I only traded this properly following every signal for a month so can't say how often it'd be a problem it's just for me during that small time it wasn't.

Hoggums

I take your points; they are 100% correct but the point I make over and over again that no one seems interested in is that when you have to start thinking of other particpants skewing against you, you are trading a strategy that is too short termed to be trading with a SB company or even with our level of internet access etc.

To be fighting for every tick, as new_trader says you need to be direct through a gateway and entering on to the exchange order book itself.

I'm not saying anyone is wrong but for me anything where you have to worry about stopsbeing sought out, skews against you, ticks and .1 of a point isn't viable through SB and with the data we have and the time delay.

Just my thoughts, if you are trading from a distance like I am (I guess most of us are?) longer term systems have to be the answer.

Stephen McCreedy

 

[smccreedy]

new_trader.

If you buy Gold with a two year view you have your stop maybe at a point that if a whole month is spent below a certain level you exit at the market at the end of the month.

It's the kind of system where you might make hundreds and hundreds of pips and if you say I'm out of this six month trade if $635 1/16 is hit you're going to be out too soon, you're not fighting for that 1/16th you're weighing up longer term factors.

Go tell Warren Buffet every tick counts and he'll tell you his stops are probably 50% below the current price, he was also the richest man in the word at one point.

I'm sure for his bigger wins George Soros would say the same thing.

Don't worry about what the market is going to do in the next 5 mins, work out the next 5 months and bet enough on it to matter. I think that is a rough paraphrase of the great man.

It is simply wrong to say that your data needs to be accurate to 1 tick to weigh up the this sort of trading system.

All this frequent trading, if you're not close enough to the market, just makes commision for the dealers and puts you at a disadvantage.

We'll just agree to disagree on this one.

We're both right in our own way I suspect.

Stephen McCreedy

 

[new_trader]

new_trader.

If you buy Gold with a two year view you have your stop maybe at a point that if a whole month is spent below a certain level you exit at the market at the end of the month.

It's the kind of system where you might make hundreds and hundreds of pips and if you say I'm out of this six month trade if $635 1/16 is hit you're going to be out too soon, you're not fighting for that 1/16th you're weighing up longer term factors.

Go tell Warren Buffet every tick counts and he'll tell you his stops are probably 50% below the current price, he was also the richest man in the word at one point.

I'm sure for his bigger wins George Soros would say the same thing.

Don't worry about what the market is going to do in the next 5 mins, work out the next 5 months and bet enough on it to matter. I think that is a rough paraphrase of the great man.

It is simply wrong to say that your data needs to be accurate to 1 tick to weigh up the this sort of trading system.

All this frequent trading, if you're not close enough to the market, just makes commision for the dealers and puts you at a disadvantage.

We'll just agree to disagree on this one.

We're both right in our own way I suspect.

Stephen McCreedy

Stephen,

Are you designing an investment strategy or a trading strategy? Ask Warren What he thinks about trading currencies...

 

[smccreedy]

I feel he'd say the odds are stacked against you and he hadn't met anyone rich from trading fx.(Of course his idea of rich and mine are worlds apart!)

I think, and this is far from dictionary perfect, if you are looking at fundamentals you are investing if you are using TA only you are trading.

If I trade the 200ema or buy the 55 day high regardless of market sentiment, paper talk, what might happen in iraq etc etc then for me I'm trading not investing. The time scale doesn't come in to it.

If I am guessing the next tick on fx for me I'm gambling, not trading or investing.

That is not to say people can't make it work, that's to say I need to take a step back from the speed for it to work for me (I'm quite compulsive by nature!).

I like to think I'm develping a trading style, if you want to think it's investing it doesn't matter to me.

Even the best short term traders (Perhaps Buzzy Schwartz) view two or three days as short term, I guess you'd consider them an eternity?

I've never traded Fx so don't purport to know anything about it, this is just my view formed from my experience of trading stocks and indexes.

Stephen McCreedy

 

[new_trader]

I feel he'd say the odds are stacked against you and he hadn't met anyone rich from trading fx.(Of course his idea of rich and mine are worlds apart!)

I think, and this is far from dictionary perfect, if you are looking at fundamentals you are investing if you are using TA only you are trading.

If I trade the 200ema or buy the 55 day high regardless of market sentiment, paper talk, what might happen in iraq etc etc then for me I'm trading not investing. The time scale doesn't come in to it.

If I am guessing the next tick on fx for me I'm gambling, not trading or investing.

That is not to say people can't make it work, that's to say I need to take a step back from the speed for it to work for me (I'm quite compulsive by nature!).

I like to think I'm develping a trading style, if you want to think it's investing it doesn't matter to me.

Even the best short term traders (Perhaps Buzzy Schwartz) view two or three days as short term, I guess you'd consider them an eternity?

I've never traded Fx so don't purport to know anything about it, this is just my view formed from my experience of trading stocks and indexes.

Stephen McCreedy

Stephen,

I only day trade the E-mini S&P 500 futures contract. My investment philosophy is very different.

Let me explain it like this:

In your 3 day system, the FTSE100

Date............ High............ Low
12-Feb-07.....6,383.10... 6,344.70
9-Feb-07.....6,395.40.... 6,346.40
8-Feb-07.... 6,375.00..... 6,330.10

The next day you place a BUY order at 6395.40 + 0.25% and a Sell Order at 6330.10 - 0.25%
Giving BUY @ 6411.40 and a SELL @ 6314.30

Now, you test it with yahoo data and find that it says the high for the day was 6411.30 and the low for the day was 6340 and the close is 6345 making you think that no trade was triggered. Now, do you agree with me that 1 tick can make a difference, even if trading a 3 day system?

 

[smccreedy]

I'm afraid you've misunderstood the s.sheet on the three day system, it buys at highs or lows not a fraction above or below as the old system did.

From your data today's order is simply a stop at 6330 (the low of the last three days rounded to nearest whole number) as the system is already long.

If the yahoo data tells me 6330 has been hit the system would assumed I had been filled and am now flat.

Of course on occasion the market might hit not hit this level but I get a fill as the CFD or spread bet quote or whatever might quote a price that triggers it.

With a system like this though when the average win is 78 points, loss is 42 points and I win 55% of the time the fact I might get stopped out of a trade that carries on to new highs doesn't bother me too much, in this instance I'd lose 66 points untill I was entered again.

If a market has fallen 60 points from it's high to exit the trade the chances are it will fall further and not turn the second my broker exits my trade.

I checked each time a buy or sell signal was made and how far from the high or low of the day it was, the average was 27.8 points. ie if I sold at 6327.8 on average the low was 6300.

This makes me think the effect of getting stopped in or out just at the wrong time might be minimal.

I also looked at the times I was stopped or entered at a level within 5 points of the high or low and all the ones I checked before getting bored showed that if I hadn't got stopped or entered at that price, the following day I would have at a worse price, ie the trend was in the direction of my orders and although the market might have fleetingly hit that level and got me filled, in a day or two I'd be glad I had.

Stephen McCreedy

 

[new_trader]

Fair enough. Here is a simple formula to round values to tick increments if you are interested

=ROUND((A1/0.25),0)*0.25

Where A1 is the cell containing the Value you want to round and 0.25 is the tick value. Change it to suit. ie/ 0.1

You can also use ROUNDUP or ROUNDDOWN

 

[JTrader]

Lets say we are concerned about the 0.25% system being problematic due to the difference between official index prices, and the spreadbet quotes.
The only obvious way round this is to use FTSE100 futures data, and look to trade it with futuresbetting spreadbet. This way, we are doing our testing and trading on more or less the same prices.

Is there an inherent reason why this system would not test as well on LIFFE FTSE100 futures data?
If so, what?

If not, has anyone got about 9 years of OHLC FTSE100 futures data, that we could edit RunTheNumbers spreadsheet with?

Many thanks.

 

[smccreedy]

I agree JT

Test test test is my motto!

I'm afraid I can't help out with the data though, why don't you look to open an account with someone like www.interactivebrokers.co.uk , that way you'd be trading the exact instrument you are testing.

I don't have an account but am looking at opening one. It seems quite daunting to be honest so if anyone on this thread does or has experience of them please dive in and tell us if it's as complicated as it looks!

Otherwise if anyone has a good source for buying the data please let us know, I'm not against buying it it's just that I've never found a supply I'm happy with.

Thanks

Stephen McCreedy