10/6 – George’s testing & trading journey continues

Well, it’s about time for another blog post.  Apologies for not blogging as regularly lately.

Testing Update

I have continued to work diligently on testing my 1H-timeframe trading strategy, and have more or less finished this piece of work, including the compilation of a trading strategy report that contains a description of the setup, scope of the testing, testing results, performance statistics and self-suggestions for optimising the strategy further.  In compiling the report, I got some helpful advice and pointers from one of the traders that has been mentoring me of late, as well as a couple of guys from a City-based hedge fund. Thanks Di., De. and Be.!

Meanwhile, I have continued to put on live trades throughout this week and the week prior.

FYI, I came across a good explanation of the difference between backtesting and forward testing on the Investopedia website.  I think this explanation (here is the link) is much clearer than my attempts on an earlier blog post!

What happens now that I have reached this point?

Well, I guess what should happen, is that i feel very comfortable with my trading strategy and that I have some decent statistics to base my confidence on – off I go risking appropriate amounts of my trading capital on my trades.

Instead, what is actually going through my mind is the following:

Some of the strategy’s performance criteria simply seem too good – to illustrate:  the annualized Sharpe ratio (what the heck is the Sharpe ratio?) is 4.1, the total return for the 12-month testing period is 101%, the maximum drawdown at any point is 6%.  And all the graphs showing the cumulative performance, and/or the distribution of the results just look awesome! Basically it just looks too good to be true.

Live trading results – in stark contrast to having an amazingly profitable strategy, the live results in the past fortnight have been pretty dire to say the least.  These two weeks alone I have already exceeded the maximum drawdown implied by the testing.

My gut feel is that the strategy’s performance is overstated.  By this I mean that my average expected profit for every executed trade is not as high the testing data implies.

Secondly, my gut tells me that the strategy’s trading frequency is understated.  The testing data for 2015 implies an average of 1.3 trades a day (and that’s actively monitoring 32 different instruments).  However, even this past week, which has been rather quiet and tight, there have been 15-20 valid setups i.e. 4 trades per trading session.

So in a nutshell, I feel I can’t rely on the testing data – I don’t have confidence in the figures.

I believe I know what I have done wrong – I have simply not done the testing correctly.  In both forward-testing and back-testing I have suffered from the hindsight benefit – in identifying whether or not a test trade has been placed, I have incorporated future price action into my decision – how does this happen? On the MT4 platform (that i have used for back-testing), all the future price data is already there on the chart – but that’s how it is meant to be with back-testing so that’s ok.  However with the forward testing, or my out-sample testing (scroll down the page on that link a bit to get to the explanation) using the ForexTester2 software,  I have also suffered from this – purely by sometimes rolling forward too much price action at a time – and then going back a few candles and actually taking a trade.  I have tried to avoid this – but i guess the human brain will quickly come up with logical explanations as to why a trade should or should not have been taken – in order to get into a winning setup, or to avoid a losing setup.  Basically I should have done this much more strictly and with more discipline.

Additionally there are of course the other testing-related pitfalls I mentioned a few weeks ago on this blog entry.

Additionally, In executing the strategy in the past fortnight, I have actually been tinkering with the strategy parameters – where to set targets, and when to tighten my stop – all in all this has not made a big difference except for a GBPAUD trade where it produced a costly swing of 2.1R!

So what is my next course of action?

Let’s take stock for a minute:

  • The testing process and structure itself has been very good and solid.  I can apply the same process again in the future.
  • I have spotted changes I need to improve the quality of the testing.
  • Having the testing under my belt – quality or not – allows me to review charts and say in black+white what the various setups were, and what the results of the setups were.  I have not been able to do this in the past.
  • All in all, the testing experiences are giving me another skill as a trader.
  • But I am still not in the spot where I can enter the markets with a “proven/tested” strategy – this feels frustrating – time is passing, I am continuing to work on this stuff full-time – but I am not making any money!
  • What the heck have I done since August 2013 (that’s when I went full-time)?  Why the heck am I still not a profitable trader?

What could I do?

Option A – Test 2015 again – with more strictness – then update my strategy report with more meaningful data – but how long will more testing take?  I could reduce the number of instruments, say from 32 to 10.  If the strategy looks profitable, then I go ahead and trade it.  If not, then what….?

Option B – Think of a different strategy that allows a quicker time to “ready to be traded” status, and start working on that one.

Option C – Reduce time allocated to trading and start working on non-trading projects (property investments, and maybe return to playing poker?)

Option D – Review the live charts for setups, but stop taking live trades. It’s cutting into my productivity and it is costing me money.  I don’t know whether I am trading with an edge or not.

Option E – a combination of the above? Anything else?

It seems I have already come a long way with this strategy.  I have tested diligently for a good two months already – surely it would be silly to drop it all now!  So Option A seems to be the most sensible choice!  But at what point will I say “enough is enough” on this trading project and move onto other (hopefully greener) pastures?

FYI…..

Writing this blog post is an excellent example of how blogging is/has helped me to keep making progress.  When i started writing this particular post, I was all confused – spelling out my thoughts has given me some clarity.  Incidentally, there are quite a lot of benefits from blogging – I have talked about a large chunk of them here.

This entry was posted in Testing. Bookmark the permalink.

One Response to 10/6 – George’s testing & trading journey continues

  1. Pingback: 15/6: Strategy Development & Testing: Nitty Gritty Stuff | Trick or Trade - the story of a currency trader...

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