Having published this article last Friday, we decided that the Night(Long)/Day(Short) strategy deserved a second look.
Our results are presented below, plus details on our current positioning in this market - we're in the black and averaging out above 1680.
Between 14/10/11 (1677.5) and 20/01/12 (1664.6) we have Gold broadly unchanged and actually down around 13pts. So a Buy & Hold strategy would be down, making it an unbiased and representative period for a backtest.
Executing a Buy at 22:00 GMT and a Sell at 06:00 GMT (every trading day for the period) would've netted 95.94 USD over 68 trades. Eureka then?
A few caveats:
A few caveats:
- Accounting for dealer spreads - anything up to $1 per trade - reduces this gain to between 30 and 60 USD.
- Our 68 trades are more representative than the article's test period as the Gold price was broadly unchanged. However, we don't feel that this has enough statistical power to warrant a forward test.
- Possible trade variations, such as including stop losses or take profits, are unlikely to have significant effect as the number of winners and loser are broadly even - i.e. few spikes down distorting the analysis.
Conclusion
The results are by no means meaningless. We will bear this market bias in mind when entering and exiting positions, just as we do the multitude of other technical factors that PMs need to be aware of but cannot trade off alone.
On XAUUSD (Gold), my main message would be "Buy on Dips" as ultimately the LTRO is quantitive easing by another name. As the full effects of this radical policy are yet to be felt, or even accurately estimated/discussed, the ultimate winner will be inflation assets, of which Gold is one.
DISCLOSURE: We are long Gold from 1649
DISCLOSURE: We are long Gold from 1649
No comments:
Post a Comment