Letters to the Editor ? Moving Average: Holy Grail or Fairy Tale?

by Various, 6/23/09

The following letters are in response to Ted Wong’s article last week, Moving Average: Holy Grail or Fairy Tale - Part 1:

 

Mr. Wong:

I found your piece in Advisor Perspectives very interesting. We have looked at systems like this in the past, as an overlay to our mainly fundamentally driven investment process.

One question – what were you trade signals, i.e., when did you “buy” or “sell”? Given that prices often trade through moving averages then reverse – when did you take your trades?

For example - if the index crosses DOWN thru the MAC - on a Tuesday - is it registered as a SELL by your system? What if closes above on Thursday - is that now a BUY?

Do you look at the data on a daily, weekly, or monthly basis? And if you are using a monthly basis, do you take the level of the MAC at the time of the cross or the closing price for the month?

Also have you ever done this real-time, or is this all theoretical?

Sincerely,

Jerry Jordan
President
Hellman, Jordan Management Co., Inc
Boston, MA

Ted Wong replies:

Jerry,

Thank you for your feedback.

The signals are defined by the MAC rules. You follow the signals mechanically even if sometimes they are wrong.  You will have bad calls during sideways markets from any trend following system. But the frequency of trading, as shown in my article, is quite low. The question one has to ask is whether the overall gain at the end of the journey is good enough to compensate for the occasional losses along the way. That is why I use both CAGR and drawdown to compare investment systems.

I use the trend-following systems I developed in real-time trading, but I didn't use the 6-month MAC. I am not publishing this series to suggest that MAC or any other system should be used as an active management system. I simply took the most basic approach known by all (the MAC system) to illustrate my point: buy-and-hold is not the only sound investment doctrine and active investment works as well as passive management based on returns, and much better based on risk control.

Back to your question - you buy in the same month that the index crosses above its MA, and sell if the very next month it drops below. Sell signals would be just the opposite.

My test results are based on same-month trading. You could argue that you can only execute buy/sell signals on the last day of every month, and you must project the last day's close before it actually closes. With real-time quotes and high-speed internet, that would not be impossible. Of course, you couldn't have done that in 1871.

One clarification: I used Prof. Robert Shiller's database which is based on the monthly average closing price (http://www.econ.yale.edu/~shiller/data.htm) of S&P500, not the last day's close. 

Hope this helps!

Ted