While reading and thinking about trading systems, I wrote some notes at the beginning of the year based on an interesting discussion in the Trading Blox forum. Trading Blox is a company that produces software for testing mechanical trading systems, thus the forum is largely composed of traders favorable to the mechanical approach. There were a number of comments by traders who emphasized that they use discretion to supplement their mechanical systems. These comments assisted me in consolidating some ideas for adding discretion to my cybernetic systems.
One idea that comes to mind is recognizing current secular trends in the various asset classes. US equities tend to have secular bear and bull phases that last about 14-18 years (See Unexpected Returns: Understanding Secular Stock Market Cycles – by Easterling, Anatomy of the Bear – by Napier, Hot Commodities – by Rogers). We are currently in the midst of a secular bear market that began in 2000 which may last for a while. This does not mean that the US stock market has been or will be in a steady decline, even in real dollar terms. What it does mean is that there will be primary bear and bull trends within the secular bear market. We saw this in primary bear trends in the spring of 2000 to the spring of 2003 and the fall of 2007 to the spring of 2009, while the US stock market was in a primary bull trend otherwise. After the strong primary bull market that began in the spring of 2009 that began to falter a few months ago, one has to wonder how much upside is left. (See the following links for the current secular bear market in nominal and real dollar terms.) The realization that we are in a secular bear market leads me to believe that the reward/risk for a long position in stocks is poor even though stocks are still rising. Also, the next big move is likely to be on the down side, however, due to recent volatility, it may take some time for this trend to appear. Applying this thinking to my cybernetic systems leads me to go flat for US stocks and allocate funds for this asset class elsewhere.
Jim Rogers cites 4 commodity bull markets: 1906-1923, 1933-1953, 1968-1982, 1999-present . With every government in the world busy printing up as much of their fiat currencies as their printing presses will allow, coupled with rising demand from China, India, Brazil, Indonesia, etc., the price of “real stuff” should continue to rise for some time, regardless of whether they are denominated in US dollars, Yen, Euros, etc. A secular trend coupled with structural economic imperatives (countries attempting to inflate away debt) and favorable demographics (in emerging markets) leads me to believe that commodities should be overweighted as an asset class. Also, if trading a commodity index, then only long positions should be executed. A short signal will never be ignored, it would mean going flat instead of going short. The rationale is that this short signal most likely indicates a correction rather than the reversal of the secular trend. The only caveat is if there is another parabolic rise as in 2008, in which case a short signal would mean the start of a primary bear trend as opposed to a correction.
Regarding currencies, for reasons cited above, all fiat currencies should continue to depreciate versus gold for years. Thus gold should be traded in the same way as commodities. The US dollar index is a measure of the US dollar versus a basket of other fiat currencies. As such, it should be traded both long and short, as it behaves more like a stock index than a commodity index.
US REITs have been rising even more strongly than the overall US stock market. I find this completely baffling as both the commercial and residential real estate markets are declining. However, due to the strength of the move and my lack of conviction about this asset class, I will take the signals that my systems generate.
For international stocks, I believe that developed markets are in the same situation as the US stock market, limited upside with significant volatility. Emerging markets should present more favorable opportunities, both on the long and short side. Thus I will only trade emerging market stock index ETFs and will follow my systems.
US bonds present a very interesting case of the potential pitfalls of fundamental analysis. The blogosphere is full of well written and reasoned articles predicting an end to the almost 30 year bull market in US bonds. It is certainly compelling to read such arguments and then look at the yield curve and conclude that bond yields can’t continue to decline. However, all such arguments have been made for Japanese government bonds for 15 years and have led to continuous trading losses for those who tried to short them. This is a classic example of the old trading adage that markets can remain irrational longer than a trader can remain solvent. While I agree with many of the arguments for why the bull market in US bonds will end, neither I nor anyone else has any idea when this will happen. Thus, I will continue to follow my systems. However, when rates do rise, there will be a long secular bear market in bonds and I will be overweight on the short side.
To summarize my views about combining discretion with a mechanical system, I will first note that one trading rule will be to never go against a signal generated by my systems. If I don’t like the reward/risk, then I will go flat, but never execute a trade in the opposite direction of a system signal. Since my systems are long term systems, I am not interested in short term market moves or worried about important policy issues (for instance if quantitative easing part 2 happens, if the Bank of Japan will continue to intervene in the forex markets, etc.). I am only interested in identifying primary and secular trends to enhance returns by changing the weighting schemes of asset classes and avoiding potentially poor reward/risk situations. Will such a strategy outperform my purely mechanical cybernetic systems? To answer this question, I will create a discretionary version of the small portfolio cybernetic system beginning November 1st and add this to my two existing systems on Collective2. As a preview, here are the positions of the system if I had started trading October 1st:
RJI, commodities, long,50%
GLD, currencies, long, 10%
UDN, currencies, long, 10% (This is an inverse US dollar index ETF. Although there are tracking issues with inverse ETFs, using them avoids margin costs.)
VWO, emerging market equities, long, 10%
VNQ, US real estate, long, 10%
TLT, US bonds (20+ year), long, 10%
 Rogers, Jim. Hot Commodities: How Anyone Can Invest Profitably in the World’s Best Market. New York: Random House, 2007.