One of the most difficult problems of tactical investing is using timing strategies that are not biased due to special market conditions of the past. However, most promoters of tactical investing are unaware of this problem and rely on hypothetical returns from the past included in popular books and academic papers.
This subject is broad but very important. Many traders but also professionals are deceived by simulated performance that includes market conditions that are highly unlikely to repeat in the future due to significant regime changes.
In the same way that market makers benefited from the widespread use of technical analysis by retail traders in the past, nowadays patient buy and hold investors are benefiting from the failures of timing methods that have been popularized in books and academic papers and touted based on hypothetical backtests.
My purpose in this blog post is not to present massive evidence of the tactical investing conundrum but only one example based on the popular long-only golden cross strategy and a brief discussion about the sources of bias.
Let us look at the performance of this long-only golden cross strategy in adjusted SPY ETF since inception. The strategy buys when the 50-day simple moving average rises above the 200-day simple moving average and sells when it falls below.
Performance appears excellent: CAGR is 9.5% for the strategy versus 9.33% for buy and hold and in addition risk-adjusted performance is better with Sharpe for the strategy at 0.75 versus 0.50 for buy and hold. This could be the ticket for traders with a large marketing department for starting a hedge fund. But it is really?
Let us look at the performance of this strategy since 01/2010 below:
Any apparent outperformance based on the longer-term backtest has vanished. The buy and hold investor is laughing at the biased market timer. CAGR for buy and hold is 13.1% versus 8.40% for the strategy. Risk-adjusted performance for the strategy is also worse with Sharpe at 0.74 versus 0.88 for the strategy. Why?
The answer lies in the remarkable 1990s uptrend. During that unique uptrend there was only one major short-term correction due to the Ruble crisis in 1998. On the other hand, during the recent uptrend there were three major short-term corrections, in 2010, 2011 and 2015. This is despite quantitative easing and central bank support. The net result is significant buy and hold underperformance of timing strategies.
And while price series momentum advocates show only those references that confirm their bias, even with 200-year old backtests from the times there were no telephones, electricity and cars and more importantly no trader used moving averaged, they avoid those references that do not agree with them, for example this paper. There is not only bias in interpreting results but also of different points of view. This is the result of the Ostrich effect some in momentum/tactical investing suffer from.
This by no means implies that tactical investing is not viable. It only means that ridiculously simple models based on moving averages and momentum are biased from not interpreting market conditions correctly to obscuring information that challenges their viability. As I wrote in Twitter the other day, these gurus and pundits come and go but their countless victims of their public relations teams remain.
— Michael Harris (@mikeharrisNY) August 6, 2017
One problem is that these gurus never learn because they block anyone who challenges their views (I hold a prominent place) while they concentrate on their public relations efforts to convince the public that if they use a simple formula everyone knows, they will outperform the patient buy and hold investor. This offers a simple explanation as to why most fund managers underperform benchmarks: they are not aware of the bias and even if they are, they are not willing to put the hard work required to develop unbiased timing strategies. As a result, their fate is determined.
If you have any questions or comments, happy to connect on Twitter: @mikeharrisNY
Charting and backtesting program: Amibroker
Technical and quantitative analysis of Dow-30 stocks and 30 popular ETFs is included in our Weekly Premium Report. Market signals for longer-term traders are offered by our premium Market Signals service. Mean-reversion signals for short-term SPY traders are provided in our Mean Reversion report.