The past week for gold has been fairly flat, after the early November drop. Looking at the results from different backtested algo strategies may give clues as to further direction.
Overall, various strategies for gold suggest a bearish bias for gold in the week ahead.
An algo based on the summation of moving averages suggests a downward trend will continue. This model tends to be longer term, so short-term spikes could invalidate it.
A simple momentum model, verified by testing over the last 5 years, also suggests continued downward pressure.
The final trading model for gold consists of eight unique algo strategies, incorporating trend following and mean reversion elements. The overall correlation between the models is low, which many times leads to signals from competing models cancelling out.
When the 8 algo models are aggregated, there is a very slight bullish bias. This disagrees with the moving average and momentum algos, which indicates that caution is warranted with any down movement.
Typically, when the model are not all in agreement, the market may continue its downwards trend, but is highly susceptible to upward spikes.
It should be pointed out that each gold algo has been historically tested, with profitable results. Each model has also performed well in real time. That said, each algo model can be incorrect at any point in time.
Summing the models together helps with the overall accuracy.
Wrapping up, algo model suggest continued downward pressure for gold, with an increased likelihood for upward spikes. Stop losses above the recent 1 to 2 week highs are suggested.