Silver vs. SPY

Well, silver did exactly what I thought it was going to do. Unfortunately I didn't get back in. It fell almost precisely to the point that I thought it was going to (the blue bar on the graphic). I was nervous though because silver and the market seem to have become inversely correlated and I thought the market was going to keep going up today. Instead it retraced it's gap, which of course is another thing I typically expect. Frankly, I was just too busy with household duties this morning to trade the minor trend and as such I ought to have just stayed long silver. If one doesn't have the time to trade the minor trend then only trade the intermediate trend and live with the consequences!

Anyway, this chart is a 30 minute comparison of SPY (S&P) and SLV (silver). I'm most baffled of late by the sudden inverse correlation. The green bars on the chart indicate periods where the two assets were positively correlated. The red bars indicate periods where they are negatively correlated (in regards to direction rather than price level). The strength of silver has mostly been in that it did better in the overnight than the S&P. Since the S&P began rallying however, silver has suddenly take a different course. One had the feeling that this was going on for a while but to see it starkly gave me pause. Since I think the market is going to rally does this mean silver is going to drop? Not necessarily and since it's a lottery ticket I suppose I ought to just have stayed long and watched it play out.

I like looking at the 30 minute chart to analyze minor trend and look for potential entry points. The stochastics are particularly interesting on the small chart. One less is just never to go against stochastics when they are running the middle. They tend to slide back and forth rather quickly and you want to make your decisions at the end points. Here we see that the S&P stochastics are falling fast (brown line) and we should expect them to bottom out at the same point that they did on previous corrections. Note that it is not necessary for the price to drop in order for stochastics to "recharge", but better safe than sorry.

It's also worth noting that stochastics can stay nailed on one end for a long time while a market is trending so one should either turn to other analysis or wait for the stochastics to make a more certain turnaround before re-entering.

The composite RSI meanwhile gave us really good exit and entry signals for silver (orange lines). The RSI did as well be bouncing off of it's moving average and then slicing right through it. A possible re-entry for silver will come when stochastics dip down and when the RSI bounces off of the top side of it's moving average. Are the markets actually inversely correlated or is silver merely working out a wave pattern that coincidentally is inversely correlated? I don't know. I do know that treasuries and the S&P have suddenly become correlated so something fishy is definitely up.

Let's look at USO (oil) vs. TLT (treasuries) at the same 30 minute chart.

Note how treasuries and oil are now correlated whereas they were clearly uncorrelated prior to July 14th. The stochastics on TLT are at their support level  (black line) with the fast stochastic but the slow stochastic has a little ways to go. I've been expecting TLT to dig deeper and fill the gap it left behind but when this happens I don't know. One usually doesn't expect a gap to be left so close to a consolidation area so at some point I think TLT is going to reach down. stochastics could stay nailed to the bottom long enough for this to happen. If RSI or composite turn up though then that would be bullish divergence and probably indicate that treasuries (and the market) are heading higher imminently.

Meanwhile the minor trend trader could certainly have made a lot of hay trading oil based off of the stochastics signals (orange lines). There was even a wedge which I observed in real-time but chose to ignore since I wasn't playing the minor trend. I'm a little surprised that oil didn't dip down to fill the gap.USO is not a religious gap filler like TLT or SPY but I still wouldn't be surprised to see a gap fill and a stochastics pattern like I've drawn in.