[daedalus]

So I started digging into a little testing... and before we start lets get it out of the way... these results come from an obviously small sample of data. To be statistically significant these numbers would have to play out over 100's of trades at the least. But for the sake of discussion lets assume these numbers were to hold. What can we tell from them?



These were the results from the last week (all the historical data OEC provides) in 1 market (the 6AU0) on two different timeframes, the 125 tick, and the 250 tick. The triggers for the trades are systematic so there is no discretion involved in whether a signal fired and the entries use a buy/sell stop so the exact price of the fill is known for this data series (assuming no slippage which i'm not going to try and model - yes it can happen, yes it will happen, but a tick or two here or there either positive or negative should negate itself in the long run).

I tracked the total length of the initial impulse move until it retraced to entry or reversed and monitored the amount of backfill (stop level) required to get the most out of the move. The "Max Real Profit Fill" tracks the furthest the move went minus 1 tick to guarantee a fill at a given level. (ie. if I went long at 90.66 and it went to a high of 90.70, 90.69 was the highest point I could have gotten a fill to exit the position, thus a number of 3 would be put in this column).

The rest of the results are fairly straightforward. What I found interesting was the profit factors... even though the 2 pip exits had a nearly 100.00 profit deficit to the 3 pip results, the profit factor was substantially higher due in no small part to the significantly higher win rate.

But keep in mind that those numbers aren't exactly reflective of the results... 3 pips don't really have a 80% win rate, they still have a 93% win rate (only 1 trade flat out failed), but the 2 pip gained trades would turn into par's and not losses.

So what do you guys take away from this data (mathematically insignificant as it is)? I personally feel that the 3 pip profits might actually be a bit better even though the profit factors are smaller. I think in a larger data series the win rates between those two levels would even out a bit more especially when you consider the actual win/loss compared to the win+par/loss outcomes.

Anywho... interested to hear what you all think.

Cheers!

[MidKnight]

Ya know....if commissions are such a high percentage, you may be better off with a non commission FOREX broker than the futures. Typically AUD/USD spreads are in the 1-1.5 pip area.

Regarding your spreadsheet. As you can see here, PF is maybe not the best way to assess a method juts because we can see that the 2 pip target has the best PF but doesn't make the best money. Trade frequency, average trade, and win % give a decent well rounded picture of how the method does. If we examine those based on the data you have shown; the slightly lower win % with the significantly higher average trade on the 3 pip target versus the 2 pip target looks like the best choice. Targets bigger than 3 pip sacrifice too much win % for a smaller gain in average trade.

I don't know how you are choosing to enter, but if you are entering with momentum then slippage could be significantly more than is assumed and with such a small target it could really skew your test data. In part, than is why I like the idea of what that girl you mentioned was doing. Taking the slightly bigger bites reduce the effects of a couple pips slippage.

I've started my own investigation with 'swing scalping' and if I find anything worth posting then I will do so.

With kind regards,
MK

[BlowFish]

Just wondered whether you are considering slippage? On 50 or 100 pip moves a pip or two is no big deal however with more scalp like approaches it can knock what initially appears quite promising into marginal or worse territory. It really makes testing on any non live data tricky as it can have such a drastic impact.

[daedalus]

I am to a certain extent but i've traded these type of setups enough live that I know it will happen but honestly its a fairly rare occurrence for the areas i'm getting in at. But yea, it has to factor into these results and its something i'm very aware of.

The other thing though is that I get positive slippage roughly about the same amount I get negative slippage... getting filled at my price when the market doesn't move beyond it, etc... So I think overall the effects will negate themselves somewhat.

The other idea I was kind of playing around with is rather than take the breakout fill, since most of these moves retrace a couple pips waiting for a better fill maybe a tick or two below the breakout level for the fill which could reduce risk (keep same stop level just change fill price) and increase reward.

However, this would negate the moves that simply blast through those levels and turn them into pars rather than gains. But its something worth considering all the same.

[clmacdougall]

Think long and hard about it daedalus. Already tried this concept out, it ruins you psychologically. Had 30 winning trades straight than got nailed on the 31st. It's a pain that's hard to forget. It takes a strange kind of dude to risk like crazy for small profits and then get nailed for a huge loss and keep on stickin to that same plan. Good luck man.

[daedalus]

Can I ask what your typical win was and what your typical loss was? Obviously R:R comes into play but I think with a bit of compromise the logic could work.

The strategy i'm thinking of using would have a loss negate roughly 2 wins... certainly not 30.

[clmacdougall]

I did some extensive work on trendlines and used Demarks TD lines as my setup. I applied it to trading Corn futures and the ES. It, like many other trading methods, slants towards predictability instead of probability.
Eventually, as you might find with whatever system you should choose to follow, your stop has to grow in order for your method to "succeed" . This, to me anyways, is a massive warning sign that your method is out of step with market truth.
I can't remember what my r-r ratio was, but I know now that anything that slants more in the direction of your stop than your target is a waste of time. Just stop, take a breath like you've said you did in your initial post on this thread and try to look at the markets in a new and fresh way. Don't go for that high risk stuff man, it'll defeat you mentally.

[thalestrader]

This may be the best new thread at TL in many many moons - Good work, daedalus.

If I may interject a few of my own thoughts in response to the discussion thus far (and I have not yet read every line of every post, so if I make a redundant point, I apologize in advance to he who beat me to it). At any rate, here it goes:

I too went through a period of "if I can only get a couple two three ticks yada yada yada ..."
I cannot say it can't be done, but I can say that for me, it was beyond difficult - I cannot trade well for those objectives, and it was a losing proposition for me. I have found that whether I am day trading or swing trading, I am at my best when I am trading for a significant movement that is sustained in terms of extent (but not always in terms of time). As I get older, whatever the time frame, I find myself always trying to trade less but to take more when I do trade. Again, that's me, and you should continue your own exploration, and I will be interested to read what you and Midk find.

I conceive of RR in terms of where S/R is in relation to my entry. Yes, MidK, S/R is history, but markets, like humans, have memory, and quite often (more often than not) that history has meaning and relevance to the present. Does the market always swing neatly from S/R and back again? No, and that my friend, is what a stop loss is for. While I am no scalper, I have often said that for me, every trade starts as a scalp. That simply means that if price does not soon confirm my trade, I'm out. But I am always trading for something larger than a few ticks, though often that is all I get.

As far as expectancy goes, it seems to me that it would be easier to develop a system, method, approach etc. that has positive expectancy if you do so in connection with considerations of R/R. Ignore RR considerations at your peril. I hardly think they are as immaterial as Midk believes them to be, and in fact, I have found them to be quite useful when day trading.

A good start to a good thread, daedalus. I hope you keep it going.

Best Wishes,

Thales