Friday, February 13, 2015

2015.02.13 Gold Cycle Model Chart

Once again, gold prices are lagging the predictive curve, this time by two sigma units. This suggests that pressure for a reversal is building based on historical price cycles.  This is similar to what we observed at the end of October last year as you can see below and on the blog. Hopefully this is also helpful to readers learning to interpret predicted vs actual divergences by means of the Z-score. One other observation to take into account: at this point in time the predictive curve is rising, if the gold price were to drop below $1200 in the next few days, the Z-score would rapidly rise to greater than 3, an unlikely event by the model parameters, suggesting that $1200 +/- is a base price for gold at this time.

2015.01.16 Gold Cycle Model

2014.10.31 Gold Cycle Model


PayDay said...

Is Vix under similar pressure? Not sure what the Z score is or how you calculate it but seems it would be at or above 4 for Vix.


Paolo said...

It is over 4, and there is pressure for reversion to the mean. However, unlike commodities, previous cyclical factors seem to be important, but are not the only factors influencing VIX. My caveat for the VIX cycle chart has been that the model, as structured, cannot capture very short term spikes in either direction, but only the general trend. Ideally I could balance the information with a neural network model I use, but that would require daily updates to better capture behavior.

Anonymous said...

thank you for the gold update. much needed. it went down very heavily again... it quite surprised me. i have not anticipated such a fluctuatin. greets from germany.

Anonymous said...

Thank you, as always great work!

Permabear Doomster said...

So when Gold falls below $1200 this spring, what then?

Whats your downside scenario, because I only see a gold-bug upside line?

Higher USD will remain a problem for all $ asset classes for the next few years.

Paolo said...

I reckon the cycle model has picked up a gold bug. Seriously, the cycle model simply shows what might be expected based on past cyclical behavior. If the gold price drops below $1200 this Spring, then reality always trumps models. However, prior to that, we should see a Z-score that stays >4 for some time. At which point, reality trumps the model, much like reality trumped the DJIA index model. C'est la vie with cyclical models, or any model for that matter, I am amused because I mentioned a while back that both gold and silver were ready for a fall, and at that point, I had the opposite criticism - I must be a gold killer, etc. This type of tech analysis is relatively dispassionate. It is what it is, and fundamentals, if they can be discerned in an unbiased fashion, don't enter into it.