In yesterday’s IKN Weekly, IKN242, we ran a short piece on the predictive powers of the Kitco News Gold Survey, a weekly exercise run by said website that polls a group of experts and gets them to call the next week in gold (here’s the link to the latest example, from last Friday, to give you an idea).
A couple of readers have mailed in this morning to say that it would be a good piece to see on the open blog but I don’t really want to share it all, so by way of compromise (and hey, to put something easy on the blog this AM) here’s just the table that went with the article, plus the explanatory notes on what the table shows below. From here you can draw your own conclusions, methinks.
Notes on
the table:
the table:
- On
the left the date of each survey. Simple.
- The
next five lines are the raw data from Kitco, including a) the number of
people asked (part), b) the number of responses garnered c) the number of
bullish calls d) bearish calls e) and neutral calls on the week ahead in
gold.
- Next
the numbers were crunched into the percentage of bulls bears and neutrals,
which you see in the centre of the above table (all cute with a colour code, too). Also in this section, I’ve
highlighted in peach the most popular call of the week, i.e. the direction
in which the overall survey is pointing for gold.
- Then
comes paydirt. First the price of the gold bullion ETF (GLD that weekend
and then the percentage change in the price of gold that week (result %).
- The
final column of the table is the answer to a simple question: Was the
survey accurate that week in its prediction for the price of gold. To
decide this, I assumed any plus/minus movement on the week of under 0.5%
was neutral, then above was a bullish result and below a bearish result.
