If you download historical data into Excel. It is possible to calculate "correlation". A mathemtacal figure that tells how much one item "zig or zag" in the graph is reflected in the other graphs "zig or zag". 1 means they are equals. -1 means they are opposites. 0 would be "no correlation".

Correlation value change over time so the values provide here are only valid in the mid of 2006. A good example of this is Canadian dollar that sometimes correlates with gold, sometimes with oil, etc.

Use any correlation value that is above .5 or below -.5 as it gives a trading edge.

(Each correlation is only listed ones. I don't mention non-correlating items)

Australian Dollar: Understandable correlates to Gold (0.84) since Australia is an important gold exporter.

Corn: Eurodollar (CME traded index) (0.81). Five-year notes (Chicago Board of Trade) (0.77). Thirty Year Treasy Bonds (Chicago Board of Trade): 0.77. Weat (0.93).

Euro Dollar: Five-year Notes (0.96). Japanese Yen (Chickage Merchantile Exchange) (0.68), Thrity Year Treasy Bonds 0.91. Weat (0.71)

Five-Year Notes: Japanese Yen (0.74), Thiry-Year Treasy Bonds (0.97), Weat (0.75).

Gold: Japanes Yen (0.74), S&P 500 (0.62), Wheat (0.62)

Yapanese Yen: Thirty Year T-Bonds (0.71).

Weat: 30 year T-bonds (0.8).

Stock Prices

Here are some examples how S&P correlates with DEL (0.82), and DOW Industrial (0.978).


In my Forex trading I often use EUR, USD and YEN due to the excellent leverage I can get with my money on Yen.

For trading Euro, it is good to watch: Five-year Notes (0.96), Yen (0.68), Thirty-year t-bonds (0.91).

For trading Yen, it is good to watch: the Euro (0.68), Five-year Notes (0.74), Gold (0.74), Thirty-year T-bond (0.71).

Correlation Between Currencies

EUR: CHF (0.96), GBP (0,80), JPY (0.63), Aud (0.67)
JPY: EUR (0.63), CHF (0.65), Aud (0.75), Nzd (0.5)

Unfortunatly, the hour corrleation between even highly correlated currecnies like EUR, CHF can vary greatyl. However, EUR & CHF correlations are very high on the long term.

Here again with currencies, it is good to calculate correlations using Excel from time to time.

No comments: