To figure out where the FX market might lead us in 2013, I thought it would be useful to identify (and quantify the impact of) the main common drivers of the most traded pairs. I implemented a principal component analysis (PCA) on the following basket (for the sake of consistency, all pairs are represented as USD/CUR where CUR stands for Currency): EUR GBP AUD NZD JPY NOK SEK CAD. I did not include the CHF because of its peg.
For those not familiar with PCA, it models the variance structure of a set of observed variables using linear combinations of the variables. These linear combinations, or components, may be used in subsequent analysis, and the combination coefficients, may be used in interpreting the components.
I extracted the 3 first components. They explain 90% of the total variance of the weekly returns (carry) of all the pairs.
Identification of the New Measure (Component)
1. The first component (66% of total variance) is clearly linked to risk aversion. As can be seen on the left chart below it is strongly correlated to the VIX. The fact that JPY has a negative sign reflects its traditional "safe haven" role.
2. The second component (13% of total variance) is Europe-specific as it assigns a negative signal to EUR, GBP, NOK and SEK. There is a close link between the 13-week change in the Sovereign Europe composite CDS *(SOVX) and this component. The weakening of the link in late 2012 can be explained by the fall in the euro breakup risk (cf. the latest Sentix Investor Confidence survey).
3. The third component (11% of total variance) is pretty straightforward: oil currencies (AUD, NZD, and CAD). Both charts below show a strong link between oil (or CRB) and the third component. The disconnect observed in late 2012 can be explained by the disconnect between oil and other risky assets and/or the monetary policies carried out in countries such as Australia.
Interestingly enough, gold prices managed to track the 3rd component in the last period (left chart below), but the same gold failed to track the surge in USD/JPY!
The late-2012 disconnect is mostly due to the Yen. If we look back to the table above we can see a very high coefficient (0.8) assigned to this component. In a sense, Japan's idiosyncrasies skewed the reading of the 3rd component. This will not last.
Implications for 2013
1. FX markets will be driven by risk aversion. My view is that the economic recovery (albeit fragile) should increase risk appetite, though political risk (U.S. debt, elections in Italy and Germany …) will continue to dominate in the short-run.
2. Euro risk is not over, as Spain will ask for the OMT, and Cyprus will also ask for help. It is not clear though if the OMT would be that negative fort the EUR. In addition, if mild reforms are implemented (resolution scheme for the banking union), the second component may not reach its 2012 highs, as the risk of breakup is clearly low.
3. Commodities had a tough year in 2012 (China slowdown, European debt crisis). In 2013, non-OPEC supply is expected to grow more rapidly than demand, hence a decline in the daily call on OPEC's production. Crude prices are thus expected to remain soft against this backdrop. My view on gold is still bearish with an average target price of $1600 in 2013, a forecast that is in danger of being revised downward. Demand from India, China, and EM Central Banks may not be as buoyant as it has been over the past few years. Basel III-related demand of gold is not enough to bring the market to equilibrium (the liquidity requirement deadline has been postponed, with a "phase-in" period enacted for 2015-2019).
The FX spectrum will continue to be driven by risk aversion in 2013. The euro risk will probably be slightly less relevant whereas the potential for a huge shift attributable to oil prices will remain limited. One major uncertainty is the USD/JPY, as Abe and the new governor may prove less inflation-prone. Restarting the nuclear program in Japan would ease the current pressure on external accounts hence alleviate the downward pressure on yen.