As with gold, the post-2005 linear fit (purple) looks very good (silver is in fact touching the line at this moment). Interestingly, that line depicts growth at $0.07 per week, which would mean silver takes close to 5 years to get to $50 again. Of course, that assumes no parabolic jumps going forward, which I don't consider likely. Nonetheless, these linear regressions are illuminating, not least because they show that silver did in fact regress to its mean after the 2011 fireworks, overshooting all regression lines to the downside this past summer. Since then, it has continued to coil around its main regression line (purple) that's right between what can be viewed as upper and lower boundary regression lines (green, brown). Nothing about this chart says silver is either overbought or oversold in the intermediate term.
Note I also drew an important lower trend line (light blue, dotted) that seems to establish a pretty strong floor for silver at ~$28.
And for kicks, I drew a Fibonacci fan (dotted blue lines) based on the April 2011 top and the June 2012 bottom, because it appears to be acting as tough resistance this week, just as it was during the last pop in late September.