Don't Be Fooled by Randomness

Don't Be Fooled by Randomness

Posted on November 04, 2013

Canterbury Portfolio Thermostat - Weekly Update - 11/04/2013

Current Market Environment:
Market State 1: (last 15 trading days): Bullish/Rational - Market State 1 is the most predictable of the Portfolio Thermostat’s 12 Market States (environments). The risk, while in MS 1, is typically around -2% to -4% from the previous market peak. The S&P 500’s peak was registered on the close Tuesday at 1771.95.

Canterbury Volatility Index (CVI) = 54 (rational market environment) The CVI was down 3 points for the week and was also down another 3 points the previous week. That would be a total 10% decrease in volatility over the last two weeks. Volatility is low and declining which is characteristic of a healthy Bull market.

Market Comment:
Our short term overbought/oversold indicator is currently 89% overbought up from 83% last week (market is over extended). It is likely to see some weakness as the market works off its overbought condition. Any expected weakness should fall in the category of normal market noise unless we see some major changes in the Portfolio Thermostat’s indicators.

The S&P 500 was up about 2 points, or flat, for the week. The good news is that the market was able to hold on to its substantial gains from the previous two weeks. The month of October finished up 4.47% and September was up 2.97%. The substantial out performance of the stock market is particularly impressive because it flies in the face of traditional lore that says markets are expected to do poorly in September and October.

Historically, September has had the worst stock market performance of the 12 months and the following month has had the worst crashes. October’s stock market disasters include the Panic of 1907, the October 24, 1929 and the October 19, 1987 stock market crashes. Don’t forget Halloween is also in October. That is scary.

September and October’s poor track record has not gone unnoticed by market prognosticators.Many of these "gurus” always seem to find ways to explain why markets do what they do; of course they aren’t nearly as accurate before the fact as they are afterwards.


Published: Monday, Aug 26, 2013 | 11:35 AM ET
By: Patti Domm | CNBC Executive News Editor - Headline:Look out! September market headwinds are looming
September is historically the worst month for stocks, but this year, the calendar is a minefield for markets.


Published Thursday, Sep 12, 2013 12:42 PM EDT
By Morgan Korn | Daily Ticker - Headline Stocks Heading for October Crash: "Take Some Chips Off the Table,” Merk Says Axel Merk, president and CIO of Merk Investments:
A stock market "crash" is highly likely next month. "It’s not just because October is crash season.” "The stock market has been going up month after month.”


Much has been written about the historical relationships between obscure events and the stock market’s performance, such as:

  • The January Effect
  • The first five days of the year
  • Presidential cycles
  • Sell in May and go away
  • The lengths of lady’s’ skirts and many more

I bet you didn’t know that, since 1950, stock market returns have been about three times higher during the three days following a new moon than it is for the three days following a full moon.

Observation:
Humans, by nature, have a problem distinguishing the difference between random events verses true cause and effect. I am sure that many of the Boston Red Sox players truly believe that their scruffy beards had something to do with them winning the World Series. The truth is that it is easy to be "fooled by randomness” because randomness does not always appear to be so random.

For example, the famous Super Bowl indicator correctly predicted the direction of the stock market for 28 of the first 31 years. This incredible predictor says that the stock market would have an up year when the NFC, or a former NFL team, like the Colts, wins the big game. The market would go down if the AFC, or a former American Football League team, like the Denver Broncos, win. In 1998 the Broncos (AFC) won the Super Bowl and somehow the market managed to close UP, instead of DOWN, for the year. The Bronco’s win kicked off a 4 year straight losing streak for the Super Bowl indicator.

By the way, the NFC won the Super Bowl coin toss 14 years in a row until the string was broken last year. The odds of that happening are 1 in 16,384. Fortunately, I am not aware of any guru who was able to tie the NFC’s coin flip prowess to a market prediction.

Bottom Line:
Any investment management or portfolio strategy should be composed of a closed set of testable rules. These rules should be programmed into a testable model that will produce specific actions and the timing of those actions. Models should be extensively stress tested through all possible variable environments. Observations and conclusions regarding predictive value should show statistical valid probabilities that exceed luck and the possibility of "being fooled by randomness.”

We remain in a long term Bull Market with generally low and declining volatility. Based on where the extensively tested Portfolio Thermostat’s indicators stand today, we can expect a continuation of the Bull market including the normal noise from the financial media and normal market fluctuations.

Canterbury Investment Management: Tom Hardin

More About Tom Hardin

As Chief Investment Officer, Tom Hardin, Chartered Market Technician (CMT), makes all the final decisions on all investment and portfolio management decisions for Canterbury Investment Management. Tom has more than 30 years experience in the investment management industry and has broad breadth of knowledge. He is known as an innovator, educator and been revolutionary in the advancements in portfolio and risk management.

Every effort was used to provide accurate data and mathematical calculations to provide, what we believe to be, accurate results. Canterbury Investment Management, LLC, and its principal owners, make no guarantee of completeness or accuracy of data or calculations as well as conclusions of any statistical data or information contained in the simulation illustrated on this page. Past results or performance is in no way a guarantee of future results.