Lately I have been reading 'Simple Economic Concepts for Complicated Financial Markets' by GaveKal. In the section on 'Gold' the following paragraph appears, taken from an August 2nd 2004 article:

By most accounts, over half of global jewellery sales occur in November (Indian wedding season), December (Christmas and Hanukah), and January (Asian Lunar New Year). This simple fact might explain why nearly every peak in gold prices has been made in January-February and nearly every trough in July-August. Surprisingly, few investors take Gold's strong seasonal factors into account.

I looked at gold prices (continuously backward adjusted futures) from 1975 to 2007. I defined 'Sell' as selling on last day of Jan and 'Buy' as closing out short on last day of Jul.

Park1The 23/33 win ratio looks interesting but I thought about checking that the gains from selling Jan and buying Jul are significantly different than picking any other arbitrary month combination.

Looked at all other six-month points changes, e.g. Feb - Aug, Mar - Sep, over the period 1976 - 2007 to compare against selling Jan and buying Jul:Park2
So it appears that Jan/Jul returns are not significantly different from any other six-month combination.



Backtest , a new R package I've been exploring:

The backtest package provides facilities for exploring portfolio-based conjectures about financial instruments (stocks, bonds, swaps, options, et cetera). For example, consider a claim that stocks for which analysts are raising their earnings estimates perform better than stocks for which analysts are lowering estimates. We want to examine if, on average, stocks with raised estimates have higher future returns than stocks with lowered estimates and whether this is true over various time horizons and across different categories of stocks. Colloquially, "backtest" is the term used in finance for such tests.


Resources & Links