Previously I have only daytraded stocks and traded SPY/ES. Over the last couple of months I have also had a look at GLD, and I’ve been trading this ETF. GLD resembles the price of gold bullion. I’m not a gold bug, instead I prefer to invest in real estate to hedge for monetary inflation, but GLD has a low correllation with the stock market and that’s what attracts me.
Because gold is a 24 hour market, and GLD only trade on the exchange in regular hours, there are a lot of gaps up and down. Unlike SPY, it has a tendency to not revert to the mean.
First, let’s look at some characteristics of GLD:
- c2c: accumulated profits from close to the next days close
- c20: accumulated profits from close to next days open
- o2c: accumulated profits from open to close on the same day (daytrade)
The period is from its inception in 2006 until present. We can clearly see the gain coming from holding overnight. However, when GLD gaps up significantly and also opens above the high of the last several days, it tends to rise even more during the day towards close.
Over the wholde period GLD has on average gapped up/down 0.64% (ignoring if it’s up or down). This is quite much and more than double the average gap compared to SPY.
Here is a potential strategy:
- GLD must open above the high of the last 5 days
- GLD must gap up at least 1.5 times the 25 day average of the opening gap
- Buy on open, exit on close
A very simple strategy which gives this result:
|P/L in %||#fills||#wins||Avg|
Average winner is 0.58% and average loser is -0.45%. There are 20 winners above 1% and just 10 losers worse than -1%.
Average gain from open to close over this period is 0.01%, so this a lot better.
Here is the profit curve:
Actually, GLD has a bigger chance of gapping yet again the next day. On avrage GLD has gapped up 0.07% over the whole period. The next day GLD has gapped up on average 0.12%, a lot more than average.
The P/L improves if we hold GLD longer, but in my opinion with a lot more at risk. Perhaps holding until next days open is worth it, it boils down to the risk appetite of the trader. In this case the average gain is 0.29% per trade.
If we turn it upside down and go short we get this profit curve:
This has been a bull market, so clearly there is something value in this strategy/setup (?).