EOD FTSE Trading Strategy. Tweaking Ideas Appreciated!

[FetteredChinos]

Ok Guys, Long Post Alert (LPA!)

I have been lurking on this BB for what seems like ages,but this is going to be my first post. Lets hope this thread becomes a constructive one.

Following on from the interest generated by the Index King threads, and the work BigBusiness did in fathoming out the formula. I did some backtesting using the yahoo historical prices (unreliable i know, but we can discuss the merits of that later!). It seems that IK's method was not robust enough to cope with the slippage and movement in the FTSE prices after the close. There were ways of optimising his system further, but they required using the closing data from the previous 50-100 days, and that seemed like too much of a pain in the ****.

So I had to change tack, and have spent the last 4 weeks devising, what appears to be a more robust system. For the sake of provoking thought, and inviting comments and suggestions, i am going to divulge it here, as any system can be improved, but Im struggling on how to do it in this instance. I have tried programming in stops, using all sorts of closing criteria (ie number of days, profit targets, but i cannot markedly improve the system)

At the moment, I work full time, and therefore can only trade End of Day. Owing to firewalls here i cannot run MM5 or in fact access the online D4F trading platform, so i have to telephone in my orders to the dealers. Thus one of the basic requirements of the system was that it would be simple to check the price action of the day and then decide what action to take as the close approached.

Having played around with the OHLC "data" proferred by yahoo, i came up with the following Stop and Reverse Strategy. It tries to take advantage of the fact that markets only trend for about 30% of the time (an oft quoted figure, but i have seen it in many places, so it cant be far off being correct)

Simple the rules are:

Buy if Today's Low < Low of Previous 2 Days
Sell if Today's High > High of Previous 2 Days

All trades to be activated as close to the Close as possible.

Trade continues until counter-signal is generated.

I have tested it on the FTSE and it performs well, and on the IBEX where it performs slightly less well, but still generates a reasonable profit. I would like to test it on the DJIA and SP500 , but yahoo seem to be having problems in this regard (or are they planning to start charging for it???)

I have attached a spreadsheet of my backtest going back to Jan 1996 , and this returns a profit per trade of 20pts , about 5 trades a month. Obviously spread and slippage will eat into this, which is why im asking you more knowledgable people on how i could tweak this for a few extra points.

Im at work at the moment, and so the Excel file im attaching is just a top level version of my backtest. The file i have at home is more complete, and can break down the analysis further (eg Long Profit vs Short Profit - negligible difference btw) If anyone by some miracle wants to see it, then ill be glad to upload it in future.


all comments would be gratefully received. but lets try to keep this thread on topic please.

cheers guys, and good trading to u all!

FC

 

[JonnyT]

Ahem, this is a known previously published system....

I beleive Drawdowns are large, capital requirements are large.

Can you give details on the max drawdown during the period

JonnyT

 

[ChartMan]

Looks good, but what about those huge losses? any clues as to avoidance? 300 and 500 point loss is a monster whacking. Even so, your work seems profitable.....

 

[FetteredChinos]

I know about the dock-off great losses. thats why ive come to you guys for ideas on how to moderate the profits/losses. Id like the whole system to be more resilient and less open to massive swings. Any ideas??

In my backtest sheet (alas sadly at home) i have tried to implement a means of averaging down (i know this is against conventional wisdom) when for example, the end-of day loss is greater than a certain figure. I think i calculated that somewhere between 60 and 80 improved things somewhat:-

the overall profit/trade went up to about 28 (precise figure to be added later) and the win % went up to 78% or thereabouts, averaging 180 points a month (20 trading days)

HOWEVER, those big losses were only augmented further (twice the stake going the wrong way - bad move)

JonnyT - i can work out the maximum drawdown later tonight and post the results tonight/tmr . But from the last 45 months trading it had had only 14 losing months, and on a rolling 3-month period, i think it has only lost 5 times. So over a long period of time it is robust, but it is a bit too volatile to trade at the moment (at least without going prematurely grey!!) in the short term.

cheers

FC

 

[nautical]

Maybe you could include some kind of volume component which will obviously complicate the entry critieria but may reduce those big moves against you if they can be shown to provide some level of confirmation or not of any change in direction. Just top of head thoughts...haven't applied to your spreadsheet.

Alternatively, what about a simple stop loss when x points or y% down?

Chris

 

[FetteredChinos]

Hmm, thanks for the ideas. Had a bit more of a play with the basic sheet today implements an absolute stop (not reversing). Had a look at various levels of stops ( -50 to -200) and it doesnt really seem to help much. In fact it has the opposite effect!

The nature of this method is that the trade reverses when a new high/low is hit. therefore a fair chunk of the profit is actually generated on the day of the reversal. now if only i could harness this somehow.

i attach the revised sheet (level of STOPLOSS can be changed by altering the figure in the yellow box). not pretty reading when stops are used. May be time to go back to the drawing board (getting pretty large now) and come up with something more productive. Still 5pts per day average on the FTSE is not to be sniffed at.......


pass the tissues

 

[bonsai]

Buy if Today's Low < Low of Previous 2 Days

Sell if Today's High > High of Previous 2 Days

Are these the only rules?

They imply you can only trade on the long side.
Suspect you may not have meant that.

Under what circumstances would you open a short after closing a long ?

Or are you always in the market ?

 

[FetteredChinos]

Always in the market, good sir. It's a Stop and Reverse strategy.

I know from being a keen student of your monthly threads that you dont employ such a method. But i would generally like the reassurance of always being in the market in order to capitalise on sudden moves (odds are that they would be on my side)...

Fingers crossed....

With a bit of luck....

and 10,000 lucky charms.






Maybe i should find a more "dynamic" exit strategy. After all, that coupled with money management is 90% of this game.

FC

(edited for rubbish punctuation)

 

[bonsai]

hmm
I thought that might be the case

my suggestion is not to be always in the market. But to select
separate rules for entry after you have closed a trade.

and then you can have reasonable stop losses ?

 

[FetteredChinos]

Agreed.

Thanks for your constructive input chaps. I have noticed from toying with this, that the winners tend to come in batches.

Ie if the trade reverses 3 days running then they all tend to be profits. maybe that could be a sound basis, or at least worth looking into?

I'll do some more investigation later, but in the meantime i'm off to unwind with some popcorn at the cinema.


Good Evening one and all, and good trading!

FC