Golden Crosses Not So Golden

On Friday, Microsoft stock had a golden cross.  A golden cross is when a stock's 50-day moving average crosses its 200-day moving average from below.  This is usually interpreted as a positive sign for the stock.  Conversely, a death cross is when a stock's 50-day moving average crosses its 200-day moving average from above.

I decided to investigate whether or not this would be an effective strategy when trading Microsoft stock.  To do this, I looked at its twenty year historical data and calculated each time the stock's 50-day crossed its 200-day, both from above and below.  I used a mechanical entry system - buy signals would be whenever the Microsoft had a golden cross based on the closing price, I would buy on the next day's open.  Exit signals were when the stock had a death cross, I would sell on that day's close price. 

As it turns out, the death and golden crosses pattern is not so golden.  Out of the 15 entry signals given, only 6 were profitable.  Moreover, 6 of the past 7 golden crosses with death crosses as the exit have had negative returns.  However, had you invested $10,000 in Microsoft on December 15, 1986 (the date the first 50-day moving average was above the 200-day moving average for which we have data) and took all the entry/exit signals (the last being March 10, 2006), you would currently have $648,313.  However, your account would have steadily decreased since April 18, 2000.  There, the peak of your account would have been $1,083,188.  Using this market timing strategy you would have lost 39%.

Furthermore, this strategy underperforms a straight buy-and-hold strategy.  If you bought $10,000 of Microsoft stock on December 15, 1986, you would currently have $1,771,982.  That's nearly triple the market timing strategy.  Your peak on April 18, 2000 (vs the market timing strategy) was $2,409,309.  That is a loss of only 25%.  So, buy-and-hold on Microsoft stock beats using golden/death crosses as entry/exit signals.

Compared to the entry/exit points of the market timing strategy, the average drawdown in the nine negative periods is -9%.  For the buy-and-hold, there are only two losing periods but their average drawdown is -30%.  That's huge.  So, the timing strategy returns less but is a smaller roller-coaster ride (ie, your losses are smaller between timing periods).