gold shares set to run hard ..., page-29

Bruce,

See, now, here's your problem: you don't understand the statistics of correlation.

As a consequence, when you look at charts like this (or the Dow/Gold movements which I discussed on another thread in the past few days), you don't really understand what you're looking at.

At face value, your BGMI/Gold chart looks impressively like the equities did diddly squat during the 70's. The thing is, whenever someone who understands correlations sees a chart like this (and this chart is exactly that: one which tracks correlations), they ask themselves three questions, before they get started:-

1. Why were the initial starting conditions (prices) selected? Bear in mind that, at the starting point, we have a one-to-one relationship. Was there anything unusual about the condition of one indicator or the other, at that starting point, which meant it was more likely to produce a graphic that met a "need" to prove a point.

2. Why was the end point selected? The exact same considerations apply to this, as apply to point 1.

3. Even bearing in mind points 1 and 2 above, that does not mean that the "underperformer" didn't perform well - it just means that the "overperformer" performed better for the selected timeframe. A far fairer graphic is one which starts off with each variable's starting price and overlays (rather than compares) them, if you cannot give the entire series for consideration.

Now, I know that these points are a bit technical, so let's get to the two, final things which I think are leading Bruce astray. At this point, it might be useful to reproduce Bruce's graph:-



First, you will note how Bruce's graphic shows gold and gold shares coming back to a one-on-one relationship at 1980. All this means is that they came back to the same level that they were in 1960. So, if in 1960, gold equities were twice, or three times as great as gold, all it means is that gold equities went way above that and then came back to be at the same ratio (2 or 3 times) by 1980.

On the flipside, if gold equities started 1960 at an unusually low point in their cycle (say, one-third of their historical relativity), the big rise in the line could simply mean that they recovered to be at or above their average, then declined, gradually, to be undervalued by 1980.

Now, I know - I'm still being as clear as mud to many. I never found statistics to be intuitive, when I studied it.

So, for those who haven't already gone to sleep, here's the crunch question: How did the BGMI actually perform, by itself, over the period that Bruce is talking about?

And the answer...is bloody well!