daniel@ETFsCanada.com

4/16/2013

How popular was portfolio theory during the great depression?




Investing is tough. Billion dollar hedge funds and super fast computers keep markets efficient. Trying to pick the right stock to own or guess the S&P500's next move is a gamble. Bonds barely pay interest and have big downside if/when rates rise. Hedge funds have been useless in bear markets. Inflation eats your cash. Where do investors turn? No, Bitcoin is not the answer.





For lack of better alternatives, many investors use passive indexing (Modern Portfolio Theory). The belief is stocks go up in the long-term and any trading or stock picking will likely reduce returns. Historically, stocks have outperformed bonds, but with more volatility (standard deviation of monthly returns). Passive indexing assumes this will continue. The best portfolio is created by finding the right stock/bond mix to meet an investor’s risk and return needs. Need more return? Dial up the equity. Less risk? Increase your bond allocation. Then just hold your investments for the long term and try not to get in the way.

Passive indexing is uncomfortable in practice. Hoping for 10% a year but with losses exceeding 20% and possible decadxes with no return is a slow and painful way to invest. I cannot imagine the discipline required to hold investments during the great depression or the recent mortgage crisis or not to chase tech stocks in the 90s.

The problem is that being passive only works in the right place at the right time. Luck of the draw with your start and end date will decide your investment success. An investor in the S&P500 starting in early 2009 will have twice the return of similar passive investor that started in mid 2007. That extra return will continue to compound and catching up will be impossible. Selling in 2009 instead of 2007 would lost 1/2 your investment in 2 years. Just one below average period can impact a lifetime of saving. It’s not like flipping a coin 100 times and expecting 50 heads. One extra toss can’t impact your results very much. An investment may only get 2 or 3 cycles to be successful and just one cycle could double (or half) your investment. 

Passive indexing is inaccurate in theory. This is for one major reason; the law of large numbers. It is impossible to ever observe enough economic cycles to make statistically valid projections about the risks of an investment. 

How long is one independent economic cycle? I’d argue its 25 years but let's be generous and say every 5 years you get a fresh economic cycle to try and invest in. That's only 20 cycles in the last 100 years (not a large number). 2008-2013 probably shouldn't be in the same sample as 1913-1918 anyways. Inconveniently, economic cycles are not independent. That's how falling prices in 2007 triggered a feedback loop of fear and deleveraging that lasted for 15 months and saw the Dow lose over half its value. The impossibility of observing a large number of independent trials makes standard deviation a poor risk management tool. The potential downside from a passive investing is so unquantifiable that I do not believe it is an acceptable investment strategy.  

I don't think people were worried about historical standard deviation during the great depression. Passive investing has only worked if you were in the right place at the right time. The next decade will be decided by things like the growth of China, survival of the Euro, profitability of new technologies, and even the weather. Guessing what, when, and how much something will happen is a gamble. I want to invest, if I want to gamble, I’d rather gamble on sports.

Hope is not a strategy. If an insurance strategy is not available (either through options, futures, ETNs, PPNs), do one of two things; have a plan to sell after a predetermined amount of losses (risk budget) or only invest what you’re prepared to lose (barbell).