22/2: DAX Testing with 2013 data – findings & insights – Part1

WARNING: Another lengthy and heavy-going blog post.  Remember I write this mainly for my own benefit of keeping all the thoughts and ideas in one central place.  Happy reading in any case!

I completed testing through all the 2013 data for the DAX.  This resulted in 74 trades and net win of 14.3R – which implies a very respectable edge of 19.2%.  A 19.2% edge means that for every £100 I risk on a trade, I expect to make, on average, £19.20.  This is referred to as expectancy.  In addition I have a further sample of 54 trades on DAX/Cable/Oil (a mix of paper and real trades) from early Dec to now.  They resulted in a gain of 6.9R.  If I add all this together, I get a return of 21.2R from 129 trades, implying an edge of 16.4%.  Naturally, adding these altogether might be over-simplifying things.

I believe that most retail traders would think that making a profit of 21R from 129 trades (well those retail traders who actually understand what R means!!) would consider the result laughable.  Not me – I think numbers like that would lead to a financially very-comfortable lifestyle.  Each to their own.

Findings from testings/review of charts:

1. Trading Session Preparations: Approach – Review the 1H charts for the last 1-2 weeks, identify key levels, and identify in advance at which levels I am willing to go long, short or either.  Here are some examples of what I mean

Cable Review Feb20

Oil Review Feb17

2. Checklist/Trade Setup template – I have refined and clarified my checklist for taking a trade.  I keep a template for it in Excel.

3. Cost-Benefit analysis of multiple targets/runners -I have done analysis on how much I am gaining by using runners on trades, as opposed to exiting the entire position at T1.  It seems that all in all I have actually lost a tiny 0.36R by using runners.  This implies that had I simply always closed the entire positions at T1, then my overall profit across the 74 trades would have been 0.36R higher – thus, essentially the same.  However, one big drawback from using multiple targets (or rather, “letting the remaining portion of the trade run”) is the increased exposure of having an open position in the market.

The increase in the average duration of trades with 2 targets was 80 minutes, and 165 minutes if using a 3rd target.  This extra time (which doesn’t render any benefit) does cost me in requiring more concentration & effort and more erosion of daily emotional capital.  This time and energy could be much better applied to looking for setups in other markets, in testing, and in considering re-entering prior positions.   To keep things simple I dare suggest that I just use a single T1 going forward and accept the fact that i will miss out on some big winners but equally that I will miss out on trades retracing and pulling back, and that all in all it doesn’t really make a difference.  (However, using multiple targets and runners does seem to work better when trading from higher timeframes such as 1H.)  But I still have some reservations about this.  I will look at it a bit further.  Is there some way that i can see in the trade where price goes further, and come out of the trades where price turns around? Why? Am I greedy? Am I scared that I am missing out?  Maybe there is a simple way of getting some more profits – surely it’s worth a look?

I guess what I have not considered is trades where I have exited the entire position at T1 – how much would I have made/lost, if I had only closed ½ the position – presumably the best estimate for that is to look at the 28 trades where I did close only ½.  And for those 28 trades, the net difference is minimal – some trades gained more, some pulled back.  Incidentally here is the chart showing this finding

DAX Testing impact of multiple targets

4. RR and Win Rate numbers: The average RR ratio is 1.3, which is quite acceptable from a day-trading perspective.  In terms of win rate – 47% were winners, 39% losers and 14% scratch trades.

5. Closing trades early on basis of EMA10/20 XO – during 2014 I spent a lot of time trading on the basis of crossovers (XO’s) of exponential moving averages (EMA’s) – all of my charts, regardless of timeframe, still contain the 10, 20 and 100 EMA’s (as well as the 200 simple moving average).  In discussing a live oil trade with my office colleague, he asked me why I wasn’t closing the trade after the EMA10 had crossed over the EMA20 to the downside (I was long) because based on his experience he found that once this occurs price would very likely continue going down for some time.  His reasoning made a lot of sense to me.  So I reviewed the 74 trades looking and assessing the likely difference in result if I had closed a trade anytime that there was a clear XO against me.

My gut feel was that this would in fact save me a lot of pips in places in lieu of me meeting for price to hit the stop level.  Well, it turns out that the result was quite different.  Out of the 74 trades, there were 21 trades were a XO would have caused an “early” exit.  The crux is that although there were many losing trades where I would have “lost less” (a saving of 2.4R), there were also a number of trades that ended up with scratch or winning results after an unfavorable XO occurred (additional losses of 1.29R and 4.77R).  The cost of the adjustment in those cases outweighs the savings from exiting early the losing trades.  The table here shows the data (note that I have had to approximate the numbers for each trade):

DAX 2013 early exits on basis of EMA XO

This implies that my stop management approach actually works better than using that approach in conjunction with EMA XO’s.  This was an unexpected finding.

6. Adding to winning positions using the EMA framework – I did not figure out a workable approach for that, after reviewing around two thirds of the trades.

7. Using EMA’s for position management (particularly portions remaining after T1 – given the findings in points (5) and (6) above and the amount of time already spent looking at the trades, I did not consider that this was a viable option, and consequently did not consider it in detail.

What changes am I going to make to my trading strategy as a result of this testing and analysis (for both future testing and live trades)?

(Note that these insights are based on the DAX testing for the 2013 data.  Maybe conclusions might be different is I used 2014 data, or 2012 data.  Maybe these conclusions will not work at all if I apply them to Oil and Cable)

1.  My approach at the beginning of each session will change – specifically how I review and comment on the 1H charts for my three instruments.

2.  I will make heavy reference to my updated strategy checklist.

3.  Subject to findings from further investigations, I will exit the entire positions at T1.  Thus I no longer need to enter multiple orders, unless I am entering at varying price levels.  This adjustment will reduce my trade durations and the amount of time/effort required.

4.  I will not pay attention to any EMA XO’s against me – I am better off just sticking with my current position management approach.  I will also not use EMA’s for adding to, or managing, positions.  On this basis, why do I bother having EMA’s on my charts?

Part 2 to follow shortly – I still want to look at how to catch those favorable moves that occur beyond T1 (including potential re-entries), as well as considering what the best entry strategy is – balancing aggression with conservatism, using partial stop-entry orders and waiting for clear confirmations.

This entry was posted in Uncategorized. Bookmark the permalink.

2 Responses to 22/2: DAX Testing with 2013 data – findings & insights – Part1

  1. all looks good and always been an avocate of exiting at T1 – unless you are trading higher timeframe as you suggest or macro trading. I don’t think there is any need to overcomplicate things by finding exceptions. Far better to spend time looking for good setups, watching your levels and generally being in tune with the markets and the driving forces (if any)

    regarding the last post on stops and targets, fair comments i think. on having open positions, i’ve never been a fan of monitoring out the office, you could have your software on your phone. far better is to just close positions if you really really need to leave the office – otherwise you just get the worst of both worlds

    Liked by 1 person

  2. Pingback: 2/3: Insights from Testing & Analysis – Part 2 | George's FX trading blog

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s