Friday, March 16, 2012

Meta-Trader - Going parabolic

Welcome back Meta-Traders.

It was yet another good week in US equities with new all-time highs in AAPL, IBM, and YUM Brands. We also had meaningful breakouts in the financial and the transportation sectors with broad participation which lends some credibility to the broader stock rally.

Apple had another spectacular week, tracing out the full distance between last Friday's close at 545.07 and 600.01 which it tagged briefly on Thursday before closing the week in the mid 580's.

Study the chart on the left carefully to notice the gaps. Of this 55 point range from 545 to 600, we had upside gaps in the morning on 4 out of 5 days. If you add up the entire point movement of those gaps, it adds up to 20 points or nearly 30% of entire movement! This means that day-only traders left about 30% on the table and you had to hold this stock or the options overnight to capture the full movement. This type of action is exceedingly rare, and even more exceptional given that AAPL is now the highest capitalized stock in the US market, exceeding is nearest rival Exxon Mobil by 20%!

Regarding the upside gaps, notice that none of those gaps were filled, except for the Thursday opening gap where the stock opened with an 11+ point opening gap, printed at 600, printed again at 600.01, the fell back. That price action appears to the right of the chart above which is up to the first hour of trading on Thursday morning. Thursday's gap was filled later that day and Friday's action was well within Thursday's range. So this means that AAPL has found a solid value area, at least for the time being.

That said, I will be buying breakouts of the $600 area. One brokerage raised their price target to $720 this past week. Given the price, earnings and product momentum, AAPL could easily reach $1000 per share in the next 12 months. This is a flash back to Qualcomm back in the late 1990's which went from about $4 a share to over $100 per share in about a 13 month period. Take heed and profit fellow traders, these moves don't come along that often.

All this bullishness led my equity portfolio to a 3-year high, taking out the highs of last spring. I'm not that great an equity trader, but its easy to make money with a stock like this. Simply buy when it exceeds the old high, and changes are, you will make money. But watch carefully for topping action, and fortunately APPL is very easy to read where tops are concerned.

One more observation. Due to weekly options, every week is now options expiration week. We tend to see most of the big moves early in the week with flattening and base building on Thursday and Friday and AAPL's price action reflects that perfectly this past week.

Here are some developments on the Forex side:

- All demo accounts have been shut down including Amachay which was down about 10% at that time and a dismal failure in my testing. Additionally, I could not replicate the results on Asirikuy which were somewhat better but not spectacular if I recall.

- Megadroid and FX-Regression accounts have been liquidated with the proceeds moved to Atinalla #3 and Atipaq full portfolio respectively.

- I stopped trading Coatl H1 and closed all open positions. The account was down about 11%, well short of the 49% worst case scenario. My reasoning was that compared to the 1 other Coatl H1 account on Asirikuy, it was not trading properly. Case in point as follows. Coatl opens up positions that sometimes last days and even months.

When comparing my Coatl H1 portfolio to the other one on Asirikuy, the other one had positions still open from January and February. My portfolio - which also traded live since late 2011 - had all of its positions opened in the month of March! So clearly, either the order opening or the order close mechanism has gone awry. How could this happen after I went though all the proper steps in vetting this account for live trading?

After some reflection, the only plausible explanation I can come up with is an upgrade to the experts to move to a new, internally developed library in early 2012. The other systems I'm trading also took that upgrade, and they are trading properly. So in the end, I don't really know what caused the differences in behavior. But is was clear that I did not need to let the portfolio go to its worst case scenario to find out it wasn't working properly. Also, trading is about limiting losses and letting profits run, and I was not willing to take any additional losses in the face of a misbehaving expert system. I would be curious to know of the other Coatl H1 account took the experts upgrade so see if that either confirms or refutes my hypothesis.

That's all for now, enjoy your weekend and the fruits of your labors.

5 comments:

  1. Hi Chris,

    Thank you for the update :o) All the Coatl accounts were updates to the latest library. Your problem probably stems from startup point dependency, it is exceedingly important for Coatl derived strategies and syncing to the same trade chain can take even a couple of years. If you run several back-tests with different startup times you'll notice this issue.

    This dependency has been eliminated on F4 but of course we still need to test this framework before making it available for live trading.

    I would advice you to read some of the topics within the Asirikuy forum concerning robustness and watch the videos on trade chain dependency and the new perspectives on portfolio design which are being developed within the website. Keep up the great work :o)

    Best Regards,

    Daniel

    ReplyDelete
  2. I mean "were updated" on the first sentence

    ReplyDelete
  3. Hi Daniel and thanks for commenting.

    I went back and compared the trading action with the Coatl 1H instance on Instaforex on Asirikuy versus my Coatl 1H instance on FXDD. To keep it simple, I focused on a handful of trades opened on 1/30/2012 and still open as of today on the Instaforex side.

    I located those same trades being opened in my FXDD account and found the same trades, opened on 1/30/2012 in both accounts: NZDUSD buy and sell. There was also and AUD/USD buy and a GBP/USD sell. Both accounts show those trades, with the same entry prices almost to the Pip! This shows that the entry signals were the same and should eliminate any issues regarding startup time and trade chain dependency.

    Regarding the trade closing times, those same trades on the FXDD side were closed within hours versus on the Instaforex side, those trades are still open!

    If that's not convincing enough, take a look at the Average Trade Length metric on the trades tab on Myfxbook for both accounts. Average trade duration for the Instaforex account: 9 days. Average trade duration for my FXDD account: 1 day!. Scrolling thru the trade details, you can see this affect, trades in the FXDD account are almost always closed within hours while trades closed in the Instaforex account nearly always exceed one day.

    Next step is to open 2 FXDD demo accounts, one with the current library and another with the F3-68 library with the identical startup time. If there are any logic issues around trading closing times, this should reveal any issues regarding closing logic.

    We'll see how it goes and take care,

    Chris

    ReplyDelete
  4. Hi Fd,

    I'll look into the InstaForex side then, possibly there is something wrong with this account instead (potentially something wrong with instance ID values or something of this sort which has left these positions unmanaged). I'll let you know!

    Best Regards,

    Daniel

    ReplyDelete
  5. Discover how 1,000's of people like YOU are working for a LIVING from home and are living their dreams TODAY.
    JOIN NOW

    ReplyDelete