TSX TrendWatch Weekly

Last week I read an article somewhere (on a topic that isn’t particularly relevant here), where the author tagged himself at the bottom as a “Discretionary Investor” for the firm he worked for (or possibly owned).

Perhaps you’re already familiar with that term; but I wasn’t, and it took me a while to find a more-or-less accepted definition.  After all, it could mean someone who’s company lets him invest or not invest at his own discretion, or someone who is discrete about his investments so as not to tip his hand. It could refer to a mutual fund manager who actively manages a portfolio of stocks bought or sold at his own discretion, or a private investor who won’t admit his investment strategy at all to anyone.  As it turns out, none of these alternatives are quite accurate.  I’ll tell you what I learned below and how it relates to the relative trend analysis™ (RTA) approach I use and discuss in this newsletter.

But first, let’s look back on another week of trading in Canada.

WEEKLY REVIEW… The S&P/TSX Composite Index rose +1.4% last week… positive, but not outstanding. The day-to-day changes once again were quite small.

The S&P/TSX Venture Index  appears to be losing stem with just a +0.2% gain on the week.  Meanwhile, the US indexes fared a little better.. DJIA +1.7%, S&P500 +2.1%, Nasdaq +2.0%.

Volatility continues to decline… the VIX is down -9.5% last week and -15% year-to-date. Complacency about the markets has apparently set in.

All of the S&P/TSX sector indexes were up over the week, except for INFORMATION TECHNOLOGY… led lower by RIM.  FINANCIAL SERVICES and ENERGY are currently the best performing sectors on a trend basis. Just CONSUMER STAPLES has a small negative trend value at this time.

The %-uptrending figure rose 10% last week… from 64% to 74%. The median weekly trend value is +0.5% for the S&P/TSX Composite Index stocks.

Be sure to visit the DATA & CHARTS workbook for details and the best performing individual stocks.

DISCRETIONARY TRADING… It turns out that discretionary trading, as defined in the Traderpedia section of the Trade2Win.com web site, is a “method of trading that relies on subjective entry/exit criteria”. That extends well beyond the possible definitions I had preconceived above.  The significance of the term is apparently that it is the opposite of mechanical trading.

Mechanical trading is certainly a better understood concept. You set up some buy/sell rules based on popular technical indicators. A mechanical system might be as simple as “buy when the price is above it’s 10 week moving average, and sell when the opposite is true”.  Most systems you read about are much more complicated than that, but the central theme is that a computer could execute a mechanical system with no human intervention… once it has been designed and implemented.

Designing such systems, then, is where human ingenuity is involved.  Back-testing (i.e., applying the rules to historical data) is a key step in “proving” whether one system is better than another. However, there are two fundamental problems with back-testing…

1. As the caveats always point out, “past performance is no guarantee of future returns”. In short, history never repeats itself, because so many extraneous factors are constantly changing. A system may be effective under one set of circumstances, but not another.
2. Human pattern recognition is inherently almost impossible to break down into simple rules that computers can mimic. An expert trader will “know” when conditions are right for a trade, without necessarily being able to explain why the timing is right. If badgered into explaining his move, he’ll make up a plausible explanation that may or may not reflect his actual thinking and behaviour.

That’s not to say that nothing can be learned from back-testing, but you just need to accept these two formidable limitations.  What’s more, back-testing allows you to “tweak” the parameters in your model and keep repeating your testing until you have results that appear fool-proof. In short you can develop a system that predicts the past perfectly… but wait until you try it moving forward in time.  You may be shocked at how poorly it works.

Amateur investors are often obsessed with mechanical systems, unless they choose the path of fundamental analysis. Even there, “fundamentalists” often want a “magic box” solution to riches. Alchemy doesn’t work in the equity markets any more than it worked turning lead into gold many centuries ago.

So that leaves discretionary trading, which is apparently everything else. Traders who rely mainly on fundamental analysis are inherently discretionary traders for the most part. Even though they look at prices relative to revenues, earnings, cost structures, etc. (i.e. the numbers), they also have to factor in their views of company management, employee satisfaction, the sector within which the company operates and other things that are very difficult to quantify. The net result is that they buy and sell on an overall subjective feeling.

I personally think that I can get by quite nicely without all of that detail; but I respect the investors who have chosen that approach, because I know it can be as effective (i.e., profitable) as my own methodology.  I also never ignore what I consider fundamental fundamentals.

I know, for instance, that if Apple CEO, Steve Jobs,  were to die tomorrow, Apple stock would plunge by 30-50%. I can’t demonstrate that mathematically, but I just know it. In one sense the fundamentals of the company won’t have changed at all; but the man is so much equal to the company, that if he sneezes, the company gets a cold. We’ve seen that before when he was hospitalized for some major surgeries.

That’s the kind of subjective information that successful investors are extracting from the markets all the time, without even realizing that it’s happening.  Try building the “Steve Jobs Factor” into a mechanical system. It won’t work.

This brings me back to relative trend analysis™ (RTA), the approach that I’ve developed, and which is the core of my investment strategies.  Superficially, RTA could be used as the basis for a mechanical trading system; and I constantly get questions about why I haven’t created one. I also get asked for back-testing results.

In fact, I’ve done both… and, no, I won’t share the results with you …for the reasons I’ve mentioned above. None of my best trades have ever resulted from RTA alone. I put everything I learn from the RTA results into a broader perspective that can’t be quantified (at least not easily). That’s the relative in “relative trend analysis”. Every investment decision needs to be placed in the context of alternative investment decisions.

That hasn’t stopped me from offering trading guidelines from time to time, but I always try to put those in perspective too. For instance, I personally like to look at stocks that are climbing at least +1% per week with 70% consistency or better. That’s just an arbitrary screen… not the basis for a “buy”.  From there, other things come into play. Attractive trend/consistency numbers would  meaningless if there was a take-over bid for a company at a 30% premium to previous trading prices. Those already holding the stock would do well, but anyone buying solely on the basis of the amazing trend/consistency pair of numbers, wouldn’t make a penny. The stock price would immediately go to the takeover price and stay there (barring any highly unlikely circumstances). You’d be buying at or slightly higher than the price where you’d have to sell if the bid is successful.

So, I guess that makes me a discretionary trader and I hope that you are too. I know that the definition isn’t very useful, because there’s no “how to do it” formula. All the same, as flawed as human judgment can often be,  you’ll always be at least one step up from any simple mechanical trading system.

US STOCKS… For the past several years I’ve only referred to US stocks and indexes to make the point that Canadian stocks were almost always outperforming American counterparts.  And, even if you found some US stocks that did particularly well, you were losing much of your profit  to currency exchange rate changes.

Well, things may be different now.  We have always kept a database of 3000-4000 US stocks (complete with our trend/consistency ratings) for comparison purposes. As of Friday, 89% of them had positive trend values.  That’s substantially above the 74% with positive trend values for  S&P/TSX Composite Index stocks and the 64% in our much larger Canadian database (around 2000 stocks… available through our Premium Subscription Services).

We’ll be keeping a closer eye on this going forward. With our strong Canadian dollar, US stocks now look a lot cheaper. We’d still lose potential profits buying US equities if the CAD went to say $1.20 from $1.00 now, but if the CADUSD pair stays constant at recent levels (or even declines) we’d have a distinct advantage by owning some American well-performing stocks until circumstances change.

PREMIUM SERVICES… Within the next few weeks we’ll be making our US stocks database available as a premium subscription service. Watch for it in our catalog of products.