Looking at the other side of the coin, the long-term macro hedge fund specialist told CNBC that although the dollar is overvalued, the fundamentals which have produced that dynamic have not changed recently.
“It’s a very late cycle bull market in the U.S. dollar. We believe this is one of the dollar supercycles that we’ve seen typically over the last 40 years that lasts from 5 – 6 and sometimes up to 8 years,” he posited, adding that the currency’s strength reflected both structural and cyclical push factors.
“It will be a dollar overshoot by the end of the day and if the fundamentals remain in place, it is unlikely to resolve itself through a change in economic fundamentals. It could very well take another coordinated political intervention to reverse this dollar-up trend,” added Farrington, pointing to the Plaza Accord of 1985 as an example of such an intervention.
Highlighting the ongoing divergence in the interest rate trajectories of the U.S. and its main trading partners, the macro-focused investor unpicked the key factors driving the euro to its current valuation level.
“We think the euro is close to fair value but should be trading cheap to fair value because of a number of fundamental weaknesses and the unresolved stress on monetary union itself,” argued Farrington, noting the uncharacteristic rapidity with which the European Central Bank (ECB) intervened this month to reverse a run-up in the currency after the market interpreted the central bank’s message as more hawkish than it had apparently intended.
Farrington has a long and distinguished track record as a discretionary macro trader and while his fund incorporates computer strategies into its investing process as an aide to decipher and act on longer-term trends, he believes there continues to be a clear role for humans in the investing game, despite the rise of both passive and artificial intelligence trends in the asset management world.
“We’re still quite a long ways off before machine learning is advanced enough to replace the discretionary fund manager because everything that we do is forward-looking, we’re predicting structural breaks, political events,” he explained.
“By definition, machine learning is a momentum concept…so it is always lagging.”
