While I consider myself a fundamental investor, I have an appreciation for technical trading. Through various valuation techniques such as Earnings Power Valuation, I can ballpark whether a stock is fairly valued or not. However, the best price to enter a stock that I feel is undervalued can be a challenge. This is where technical indicators come into play as they help quantify a near-term entry price. Sometimes I trade based purely on technical indicators, however, when the trade does not go well, it can be a harsh feeling. At least with fundamental investing, there is a safety net knowing that the investment is sound (assuming the homework was done correctly), but perhaps the investment was poorly timed. With pure technical trading, there's only the reliance on prior price performance or 'what has happened in the past has a high probability of happening again.'
With that introduction, I wanted to review some lessons I have learned the hard way from technical trading over the past decade. This is not meant to be an anti-technical trading post, but rather a post to offer up some lessons learned so that common trading pitfalls can be avoided. In fact, I believe that sound technical trading simply quantifies good economic sense (to be discussed below).
For this analysis and discussion, I will use a common strategy based on Bollinger Bands using the S&P 500 SPDR (SPY) from Jan 4, 1999 to April 30, 2012. Closing prices are used without accounting for dividends. For reference, the BB formulas are provided in the Appendix. The strategy is defined as follows:
The economic thesis addressed by this strategy goes like this; ""The S&P 500 has dropped substantially more than recent history suggests it should. Therefore, I believe the stock is oversold and will return to normalcy or even an overbought condition in a few days." Is this a profound, Nobel Prize winning thesis? No, but it's more insightful that you'll probably hear on TV or when hanging around the water cooler. The strategy simply quantifies this thesis using standard deviations, trading time frames and overbought conditions. With that said, the results for this strategy are shown in the table below:
Profit per Trade (no commission) | 0.75% |
Number of Trades | 186 |
Number of Winning Trades | 123 (66%) |
Number of Losing Trades | 63 (34%) |
This strategy is not a 'knock it out of the park' strategy, but is on par with good trading strategies - a 2:1 win/loss ratio with a decent return in 10 or less trading days. Now let's dig a little further, first by plotting out the profit curve in the figure below using $10,000 per trade.
click to enlarge
Three points to make from this chart; two are obvious and one is not. First, it is assumed that trades are made with a low cost broker at a flat rate of $7 trade ($14 round trip). So with $10,000 used per trade, this equates to 0.14% - a noticeable impact for this scenario. Note that bid/ask spreads are not included which will degrade the results further. However, for the SPY, the bid/ask spreads are very narrow so not much of an impact will be felt. This is not the case for other securities that are thinly traded. Just be careful, as many trading books will not consider either commissions or bid/ask spreads.
The second point is that the strategy performed poorly from February 2001 through May 2002. That's forgivable considering the overall market performed poorly during that time frame as well. In addition, it's at best difficult if not impossible to come up with a strategy that performs well all the time.
The third point is that there looks to be a smooth and consistent rise in profits from March 2003 to present with the exception of a slight hiccup in 2008. This chart will make the potential trader very interested in what appears to be an almost perfect profit curve. Not so fast…let's examine a subset of the data around the 2008 time frame below.
Simply put, this chart is a trader's nightmare. In March 2006, the trades are going well with a slow, but steady increase in profits. This is followed by a year of low activity. There just were not many trading opportunities where the SPY was down hard during this time, but profits are still being made. This is the first challenge for a technical investor, the need for patience. Waiting for up to four months for an entry to trigger can be hard and often causes traders to adjust strategies to get some 'action' or to simply quit.
Well, the trader sticking with this strategy got some action in Jan 2008, unfortunately with two significant losing trades. Fortunately, some of these losses were recovered in the weeks to follow, but then the summer of 2008 came along with a series of large losses. When it was over, two years of profits were lost in very short order. That will test the nerves of even the most ardent technical trader and often leads to an exit altogether. In the fall of 2008, the strategy gave more entry points. I vividly remember that time. While I was not curled up in the fetal position just yet, I admit to taking a 'you first' attitude when it came to new trades or investments. And guess what? This is when the strategy sky-rocketed, generating the highest rate of profits in the 13 years tested. However, most people exited this strategy (and many like it) and did not experience that rise. In fact, most traders would probably exit at the right at the low point shown on the curve. The pain and frustration was just too much. That is the definition of a nightmare which is hard to see by looking at the long term chart or the summary table of results.
In summary, long term charts can often disguise trouble spots for traders. That March 2006 to March 2009 time frame looks benign on the long-term chart, but deadly when analyzed in detail. This is a problem regardless of the trading strategy used or the underlying asset being traded. I ran this strategy for the PowerShares QQQ (QQQ) with similar results.
So how does one deal with these challenges when they occur? There are books written on this vary, but I believe it boils down to common sense. First, any strategy or investment should be considered in the context of an overall financial plan. A single strategy or even technical trading as a whole should only be a piece of that plan. If a trading strategy is underperforming expectations, it is much easier to make sound judgments when it is only a small element of an overarching financial plan. I would think that technical trading does not fit the financial profile of most people. If it does not, there are other ways to enter the market. Some other comments on technical trading:
If trading multiple strategies make sure they are uncorrelated: Often times multiple strategies give the same entry points for the instrument being traded. If so, perhaps the strategies are correlated. I have fallen victim to this problem. Taking a step back, I realized the obvious; using multiple strategies that require a large price drop for entry will almost always trigger in unison when there is a large price drop. James Altucher's book entitled, "Trade Like a Hedge Fund - 20 Successful Uncorrelated Strategies & Techniques to Winning Profits" is a good starting point for uncorrelated trading strategies. This book was is one of the few in the past decade that not only concisely defines the strategy, but provides statistics through back-tested results.
Use trade management techniques: If technical trading fits in your financial plan, always use good trade management techniques. Elder's book entitled, 'Entries and Exits' is a good example of limiting trading amounts.
Always have clear exit points. The trading strategy has both a timed exit (10-days) and a conditional exit (profit > 1.5%). Know how and when to exit before entering.
Keep it simple. Adding conditions to strategies can squeeze out additional profits, but at the risk of 'data-mining'. For instance, you may find out that stocks falling below 1.75 standard deviations on Tuesday's have a 93% chance of at least a 3% gain in 4 trading days. I don't know of an economic argument that says Tuesday is the best day to enter a BB trade. Your results are most likely a trap that will result in future results not matching historic results.
Don't trust and always verify: With free tools available on-line or the use of Excel, most strategies can be tested. If the strategy is vague, don't bother. I can't tell you how many books I have returned to Amazon.com because a trumped-up strategy offered little to no substantiation. Altcher's book does not fit this category. He provides concise strategies and a snapshot of historical results. However, this book was published in 2004. Altucher correctly stresses that markets are constantly evolving, finding those 'pockets of inefficiencies' will require adapting strategies over time which can be done with some effort.
I have been seriously investing and trading for a little over a decade. That makes me a mere novice so I'm interested in your feedback.
Appendix: Bollinger Band Formulas (www.bollingerbands.com/services/bb/)
Disclosure: I am long SPY.
Additional disclosure: I have a covered call position in the SPY.
