Backtesting leading vs. trailing

[mmillar]

What would be a sensible way of backtesting whether one index/share leads another?

I just can't get my head around this at the moment. Some of the ways I'm thinking about it are...

1. %change charts. I've looked at %change charts, but it is quite difficult resetting the start date all the time.

2. Maybe a crossover - if the leading crosses the trailing then buy/sell etc. Would probably only work on a % change.

3. Buy on open if the leading is above the trailing etc. Probably not subtle enough.

4. Some kind of Rate of Change looking at the leading to then see if the trailing follows.

Any ideas (or TS code ) just to help get my brain into gear would be most appreciated.

Cheers

mmillar

 

[stevet]

mmillar

are you already trading successfully?

 

[mmillar]

yes. on the s&P e-mini.

maybe I'm thinking about this the wrong way. maybe I'm just looking for a pairs trading strategy in reverse. too late and too tired now. will think about it some more tomorrow.

 

[fowkesp]

mmillar

Have you ever looked at NeuroShell Trader? This can "discover" rules like the ones you describe. I have the EOD version, but there is a realtime version as well. It even uses the eSignal datafeed nowadays.

I've had some fun with it in the past but have never got around to looking at the realtime version. More info at http://www.neuroshell.com/

Paul

 

[Trader333]

mmiller,

Can you give a bit more information on what you are trying to achieve ?


Paul

 

[stevet]

Trader333

he's trying to make bucket loads of money by finding a leading indicator for a tradeable contract

 

[Trader333]

stevet,

Aha, now you have explained it is obvious


Paul

 

[mmillar]

fowkesp - thanks. I've seen NeuroShell Trader but for now it would just be a distraction. Boys toys and all that.

stevet & trader333 - aren't we all trying to make bucket loads? actually, someone has told me that they think one indicator leads another. I don't think it is true but I wanted to look at it in an objective way by backtesting. however, the more I think about this the more complex it seems. It's only a side issue and isn't important to the main plot. i think I'll just put it on the to-do list for now.

Thanks anyway.

mmillar

 

[iq200]

Trader333

he's trying to make bucket loads of money by finding a leading indicator for a tradeable contract

stevet, haha. Isnt everyone mate

Imran

 

[stevet]

ImranQ

no, everyone is keen to make bucket loads of money - but its the way you go about it and the time you commit to it that makes the difference - and seeking a leading indicator is one of the keys to trading succcess - but especially if you think of different ways to use the leading indicator scenario

mmillar

to create a good back testing system could take a lifetime - and there are so many more parameters than people realise and even then the results may not be as valid as hoped for

but that does not stop realtime evaluation of methodolgies - so you have realised that back testing is a nightmare - but just check out ideas in real time and see what y9ou come up with

you are on the right track - but if you do find a usable leading indicator - you may have to think out of the box to utilise it in a profitable way

 

[Trader333]

mmiller,

The biggest problem in doing this is determining the timescale that one will lead the other by if it happens. I know that Naz has said previously that the Nasdaq leads the other indices.


Paul

 

[stevet]

the nasdaq for sure does not lead other indicies - it just happens that its price action is different and it may appear to lead - but at the same time it may also appear an equal amount of time to lag

the simplistic concept is that since the nasdaq has stocks with more volatility, therefore its better to pile in to make the most money for any potential moves - but if only it was that simple!

at different times durng the trading day - each index has a stab at being the leading index and bonds sometimes lead stocks, but it is key to realise that their is an inter-relationship between all stock markets and bonds etc

and that the inter- reaction is not fixed - but does exist tick by tick day in day out

 

[mmillar]

PRICE DIVERGENCE!!

Easy. Thought of it in the shower. Surely if one stock is making higher highs and the other lower highs etc then the one that wins is the 'leader'? May be simplistic but it will get me what I want.

Does anyone have any good divergence spotting code for TS?

 

[stevet]

mmillar

that woud be great if they were exactly the same stock -but of course each company has a different business model, different debt structure, better shareholder pr , different dividends etc etc

so perchance not as easy as it might seem

but you are definetly approaching trading with the right attitude

and there are a lot of opportunities out there

heres one - how about price divergence of the ADR of a european stock quoted in the US against the exact same stock quoted in Europe

 

[bonsai]

when you sort something out, let us now who led the sell off this afternoon ?
classic I think, just as some indices made new highs.

 

[dazzam]

mmillar what I want to know is how is the weather in Tenerife??

 

[Lambchops]

Me too cos I`m going next week.

 

[dazzam]

Lambchops,

I am sure it will be a damn sight warmer than where we r.

 

[mmillar]

stevet,

Thanks for that. It's an interesting idea. I'll look at it in the future, but for now I have to concentrate back on my current plan (otherwise nothing will get done).

Cheers

 

[mmillar]

dazzam & lambchops,

It depends where you go in Tenerife. Most tourists go to the South...

http://www.bbc.co.uk/weather/5day.shtml?world=0045&links

The North (where I live) is currently...

http://www.bbc.co.uk/weather/5day.shtml?world=1563&links

If you go to the South then it is virtually always dry. In the North right now it is 'changeable' i.e. we had four days of glorious sunshine then yesterday it rained. However, wherever you go shorts and t-shirt will suffice.

Cheers