Yesterday’s price action in the precious metals market was very specific.
Gold declined very little, silver declined much more, and mining stocks moved a bit higher. When we see a mirror image of such action (gold pauses, silver rallies and miners decline a bit), it’s usually a sign of a top. Have we just seen a local bottom? Let’s take a look at the charts (charts courtesy of http://stockcharts.com), starting with silver.
In short, the self-similar pattern remains in place. At this time the pattern is proportionately bigger, but remains similar in terms of shape. Big declines in the precious metals sector have been very often preceded by silver’s short-term outperformance, even if this outperformance was preceded by a visible downswing in the price of the white metal. What we saw in February and March serves as a perfect example.
Why has silver underperformed recently? As always, there is no clear explanation behind any price move (other than because someone pushed the “sell” button), but it seems to us that it can be in a large part attributed to the sharp decline in the general stock market.
If we are just seeing the beginning of another huge decline, then we are still quite likely to see a sharp rally in silver, just before the big drop. In other words, what we have written about the silver market previously remains up-to-date:
The current situation is similar to what we saw in March. Silver declined after a local top was formed close to the turning point, then bounced a bit and then it moved a bit below the previous low. Back in March it rallied for a few days only to disappoint and plunge shortly thereafter. This scenario seems quite probable at this time, not only because of the similarity on the above SLV ETF chart but also because of the situation in the currency markets.
Turning to the yellow metal, we see no new low. Gold has basically remained where it closed on Monday, which is a sign of strength, given that the USD Index moved a bit above its previous high. Overall, what we wrote yesterday remains up-to-date: