Thoughts or Tips on This Basic Strategy

[CarpeUK]

I will have to write it in its most basic form as I do not know any other way to write it.

Overbought/Oversold (80-20)
---------------------------------------
CHART SETTINGS
-------------------------
Time Frame: 2day & 15mins into frequency
Indicator: SMA 14 - lower indicator 1 - Slow stochastic
Chart Style price display - Candlestick chart size large

DOWN TRADE (short)
--------------------
Wait until Slow stochastic is above 80 then enter the market
Change the chart setting to 5mins and watch.

Will do the same as the above but in reverse for the up trade (long) but with (20)

Now as you can see this is very basic and at this time, I have no idea how this will work out BUT the question that is hovering over me is where my EXIT is

Ok anyone who has used this or has any tips on how to improve this please let me know, it will be most appreciated.

CarepUK

 

[dbphoenix]

When you say you "have no idea how this will work out", does that mean you haven't tested it?

 

[CarpeUK]

When you say you "have no idea how this will work out", does that mean you haven't tested it?

Thats right, not even paper trade yet. Just one basic strategy i have come across and thought i would ask the forum for there opinions and improvements

 

[Mr. Charts]

To be blunt, you need to do your own testing as the only person whose judgement and analysis you can trust is your own.
Lots of hard work in this business...........

 

[dbphoenix]

In order to offer opinions and improvements, one would have to have tested it himself. Suggest you go back six or eight months and test this through 20 or 30 trades to determine whether there's any profitability.

Then "forward test" it from whatever point you stopped in order to determine whether or not it still looks promising.

Then paper-trade it from this point forward in real time in order to find out if it's profitable in real time (its profitability won't mean much since you're paper-trading, but if it turns out not to be profitable, you can be fairly certain it won't suddenly turn profitable with real money).

 

[TheBramble]

Ok anyone who has used this or has any tips on how to improve this please let me know, it will be most appreciated.

Carpe, if OB/OS with any settings was the answer, don't you think someone (everyone?) would have been using it by now, thus flattening it out to uselessness?

Offered in the spirit of improving this system for you.

As for back-testing as suggested by DBP and Mr. C go right ahead, literally. Don't back test. Test it going forward using just the hard right edge. It's the only honourable way.

Then go read 'no inds', 'trading the ES', DBP's 'Price Volume...' and when you really have read all that and know it all, 'Journey from the Basement' and then begin to realise you know less than nothing.

That's not a negative remark BTW - it's genuinely inspiring.

 

[dbphoenix]

As for back-testing as suggested by DBP and Mr. C go right ahead, literally. Don't back test. Test it going forward using just the hard right edge. It's the only honourable way.
.

He'll find out a lot faster through back-testing that it doesn't work.

Either way . . .

 

[TheBramble]

He'll find out a lot faster through back-testing that it doesn't work.

Either way . . .

Maybe. But he'll lose a lot less be deferring his real trading by working forward...until he finds out it doesn't work AND he'll be far more aware of market action.

There's a REAL investment in WAITING for each (5 min bar, EOD bar) to confirm your suspicions that your system is crap than there is in the past kidding you it's a winner.

 

[dbphoenix]

Maybe. But he'll lose a lot less be deferring his real trading by working forward...until he finds out it doesn't work AND he'll be far more aware of market action.

There's a REAL investment in WAITING for each (5 min bar, EOD bar) to confirm your suspicions that your system is crap than there is in the past kidding you it's a winner.

True, but he's using the 15m as a trigger, which extends the waiting.

If one tests every hypothesis via forward-testing, he can become feeble by the time he finds something that seems to work. However, one can test and toss a dozen candidates in a weekend via simple backtesting.

Note, of course, that I'm not talking about computerized backtesting, but of going through the charts manually.

 

[TheBramble]

If one tests every hypothesis via forward-testing, he can become feeble by the time he finds something that seems to work. .

I'd rather be feeble than broke.



...if that's OK with you...?

 

[dbphoenix]

I'd rather be feeble than broke.

How does one go broke by backtesting?

 

[TheBramble]

How does one go broke by backtesting?

It's quite simple. You back-test with full optimisation to show your system works. Then you trade with no risk or money management until your trading capital runs out.

I'm not recommending this mind - just pointing out one way I managed to do it.

 

[dbphoenix]

It's quite simple. You back-test with full optimisation to show your system works. Then you trade with no risk or money management until your trading capital runs out.

I'm not recommending this mind - just pointing out one way I managed to do it.

However, that's not the procedure I suggested. See Post #5.

 

[TheBramble]

However, that's not the procedure I suggested. See Post #5.

Yes, I understood the procedure you suggest in post #5. I was just answering your question about how could one go broke back-testing (the way I did it!).

 

[TWI]

Overbought/Oversold (80-20)
---------------------------------------
CHART SETTINGS
-------------------------
Time Frame: 2day & 15mins into frequency
Indicator: SMA 14 - lower indicator 1 - Slow stochastic
Chart Style price display - Candlestick chart size large

DOWN TRADE (short)
--------------------
Wait until Slow stochastic is above 80 then enter the market
Change the chart setting to 5mins and watch.

Will do the same as the above but in reverse for the up trade (long) but with (20)

Now as you can see this is very basic and at this time, I have no idea how this will work out BUT the question that is hovering over me is where my EXIT is

Ok anyone who has used this or has any tips on how to improve this please let me know, it will be most appreciated.

CarepUK

Please first explain what you mean here. You have a 14 day SMA, is this to determine whether you take the long or short signals from the stochi or is it just there because it looks good?
I love the bit about when you put a trade on you change to "5min and watch"...Why is this?
Is the system actually just buying stoch below 20 and selling above 80 with nothing else?
If you clarify I will send you the horrific truth of your system back-tested on the market of your choice(if I have data).

 

[CarpeUK]

Please first explain what you mean here. You have a 14 day SMA, is this to determine whether you take the long or short signals from the stochi or is it just there because it looks good?
I love the bit about when you put a trade on you change to "5min and watch"...Why is this?
Is the system actually just buying stoch below 20 and selling above 80 with nothing else?
If you clarify I will send you the horrific truth of your system back-tested on the market of your choice(if I have data).

Twalker,

In regards to SMA 14 it is to determin the long and short signal. as for the changing of the 5min chart this gives you a clearer indication of what is going on.

The 20/80 is over bought and over sold. You start buing below 20 then sell before 80 and vis versa below 20 etc

As i said i am very new to this so no need to takle the P**S, this is why i have posted this to see how it works out but like the others said in to test it myself then post the results. the downfull of this is u need to be sitting by your PC and i do not have an exit worked out yet.

 

[TWI]

OK. I wasn't taking the ****. Just wanted you to clarify. You will find most things look horrific when you try them out. Just have to keep experimenting and keep it simple. Eventually you see what works.
This is a start so lets see if something can be built from it.

 

[FetteredChinos]

carpe,

i think TW was only trying to save you from yourself, and save you some time in the process.

by all means fwd test this system. it will work (like most systems) in a nicely trending market, but as soon as we become rangebound (like 70% or so of trading days ) then all those profits you will have accumulated will disappear as you get whipsawed to death (false signal after false signal)

to try to reduce these whipsaws, you could avoid trading during quiet periods eg the US lunchtime/last thing on a friday afternoon/evening.

also i dont recommend taking signals on one timeframe and then zooming in on another. causes great confusion about when to enter and exit, and you end up in one helluva mess.

been there, done that, eaten the popcorn.

try to keep your system as simple as possible (less indicators the better) and get to know it inside out in all sorts of markets. unless you plan on forward testing it for a year or so, to experience all sorts of market days, you may find this system largely futile.

you arent the first person to discover an indicator that "works" with some particular settings, and you certainly wont be the last.

exit is far more important than entry. flip a coin for entry (as good as using an indicator) and try to hone your exit skills. this will make you profitable. look back at the charts and see where you should have exited. see if you can spot any similarities. if you can, then you have the foundations of a system.




good luck though. perseverance is the key.. or was that love is the key?

hmm, liquid lunch...

FC

 

[dbphoenix]

Apparently, there needs to be a thread somewhere on how to define a strategy and the tactics to go along with that strategy. When someone like Carpe posts something like what he posted, it's difficult to avoid the impression of being discouraging and also point out even some of what needs to be addressed in the plan (and there's a lot).

 

[TWI]

Best to demonstrate by working through it:
Here are results for following.

15min bar on dow future(or did you have another market in mind?)
Use 14 bar MA with price above or below this to decide whether to go long or short
When above sell 14 period Stoch SlowD above 80
When below buy 14 period Stoch SlowD below 20

Had to have some exit criteria so just used a profit target of $2500 if wait for reverse signal....you don't want to know result.

Here is equity curve.

You can see that the system has net made money but it is very horrible with respect to consistency and unlikely to prove tradable with any degree of confidence.

What would you change?