Saturday, August 10, 2013

Walk-Forward? More like Crawl Forward


I've been vacationing in Bar Harbor, Maine this week and driving to Quebec today for some sight-seeing. Around the edges I have been cutting my teeth on using the TradeStation Walk-Forward Optimizer on the strategies of which I have recently written. It's a humbling experience, be sure of that, but I like a good challenge - the more impossible, the better.

But for what its worth, I have had some modest success - both in terms of figuring out how it works and getting something to pass the seemingly impossible test.

Here is a graphic of my first conquest using the SPY-1 strategy.



I am finding there is a particular mind set or skill to writing strategies and I seem to have mastered this challenge, at least for the most part. But the skill to 'think' like the walk-forward optimizer, to correctly anticipate what will work, what will fail and know how to get the desired outcome is new to me and will take time to figure out. Lots of time.

I've also been trying to keep an eye on those rascals on Wall Street and made a couple of charts to share with you.

First we'll look at the Gold Miners ETF (GDX) followed by Gold Futures (GC).

The bottom in the gold miners came on Wednesday June 29 (as Fibonacci himself told us) and was confirmed with the gap opening of Monday July 22 which leaped the miners into freedom. It turns out that this important trend line emancipation was successfully retested Tuesday and Wednesday of this past week and the miners should now be ready to rock 'n roll.



Speaking of trend lines, several posts ago in the comment section I noted that gold could well have concluded its daily cycle but the nagging thing was that the price was going to leave the angle of this first daily/intermediate cycle noticeably steeper than the comparable at the 2008 bottom.

I drew a trend line on my chart at home and mused in my comments that gold would have to trade down to about $1250 to reach an equivalent trend line angle. The next chart shows you the trend line I drew in blue and as I was 5 or so days early with thinking gold had bottomed, look where price did bottom.




Yup, right on that trend line I made weeks earlier. Actually, this recent trend line is .1% steeper than the 2008 specimen. Close enough for me.

Symmetry. There is always a way to find symmetry and balance in the way gold price moves - both in terms of price and time. This daily cycle peaked on day 18 of a 28 day cycle - practically nailing the 62.8% measurement in terms of time (days, not price).

We've had a couple of longish daily cycles at 28 and 29 days. Perhaps this current new daily cycle (Monday will be day 3 of the average 24 day duration) will be a shorter cycle and pack a powerful punch, too.

I made it to Quebec - off to find some crepes, coqauvin, croissants and whatever else.

Keep the faith,

John
tsiTrader@gmail.com


4 comments:

  1. John, how high do you have us going in the near term? I have us topping out at 1450-1500 around mid September, then a nice decline and higher low set in early October (1300), then a much larger rally from 1300 to 1650 (Oct 13 to Jan 2014)
    Im currently Long Gold and Silver options for Sept Q3.

    ReplyDelete
  2. Beats me James. I'm sure your crystal ball works a lot better than mine and you may well be correct.

    I guess I would tend to look at answering your question in terms of what I know about the past which, of course, may not have a thing to do with the future. But based on the past I recognize that the 'typical' intermediate cycle is roughly 5.5 months in duration. It is made up of 4 - 6 shorter daily cycles that average 24 trading days. Each daily cycle is right translated until the final one which is left translated.

    Using this information as a template of sorts, and using the first intermediate cycle off the 2008 low as my guide, a top just above $1800 in early November makes complete sense to me, with the conclusion (low) of the intermediate cycle around late December.

    This intermediate cycle should be right translated (the top should occur after the midpoint day of the cycle) and your mid September idea puts the top at about 10 weeks.....that's early. A January 2014 top is quite late...at 6 months.

    But heck, what is that disclaimer wording everybody likes to use? 'Past results do not necessarily predict future performance', or something like that. Your long gold and silver options for September Q3 should do very very well. I tend to think pushing them out to October or November may do even better. I most sincerely hope they do you right! Good luck.

    ReplyDelete
  3. Lol,I like your scenario better :D
    Thank You for your honest opinion.

    ReplyDelete
  4. GLD 20/50 dma bullish cross very close to fruition. Last cross we had was 12 mo ago and the subsequent rally was $190.00 (FROM THE CROSS), if history repeats itself, we should be looking at $1526.00 in 6 weeks.

    ReplyDelete