Compared to understanding some areas of financial markets, the principle that interest rate differentials should drive movements in currency valuations is a simple one. The more interest one can earn in a particular currency, the more money is attracted to it.
Usually, and certainly for almost the last 25 years, it was short-term rates that FX traders paid most attention to -- and that means anything up to the 2yr bond yield. It's been a totally logical way of approaching things, and for two main reasons . The first is "Money Flows" : Higher interest rates attract speculative cash chasing the "carry". Remember that fella ? Essentially, it's the extra interest available in one currency over another. The second reason one might describe as the implications for fundamentals -- if rates are moving higher, it's usually an indication that either the economy is doing well or that inflation is on the up. Either one would be a cause of a stronger currency.
In 2017 the natural order of things has changed a little. After a post-recession period during which central banks globally have bought a total of over $12 trillion worth of bonds through quantitative easing programmes, FX markets are now more influenced by longer-term yields (say 10 years and above) than by short-term rates, and that's made them trickier to read . Even if they are not always able to match up to their own predictions, central banks these days do at least try to communicate what their intentions are with regard to monetary policy. But yields further down the maturity scale are not moving in tandem with short-term rates.
That makes it more difficult for traders, yes .... but also for central banks. Managing exchange rates directly is seldom part of a central bank's remit (at least, not publicly), but they are a big factor in financial conditions ..... which certainly do form part of their policy thinking. With the Fed beginning to cut its balance sheet largely made up of its QE purchases, and the ECB about to taper its own purchases, this is a delicate moment and ideally not the time for central banks to become less influential in either exchange rates or financial conditions.
But that's what's happening ..... the correlation this year between 10yr yield differentials and the value of the Dollar against the Euro, Yen and Sterling has recently hit its highest mark since the early 1990s. At the same time, even though the Fed has reacted to the US economy being further along the economic cycle by hiking rates (almost certainly three times by year-end and in contrast to the actions of other central banks), the Dollar has fallen by about 12% against the Euro, for example.
If the focus switching from the short end to further down the yield curve is unhelpful to central bankers, you could argue that they've only got themselves to blame. Global investors, faced with ultra-low or negative returns on short-dated domestic instruments at home, have been buying longer-dated US Treasuries instead. And if you're talking about central bank actions distorting the normal market dynamics that might have kept the focus at the short end, what are we to make of the ECB's pledge to keep rates at zero and below even as Eurozone growth rallies strongly ?
Short-term rates are just not giving the leads to FX traders that they have done traditionally, which is why the 10yr yield differentials are now a bigger factor. Even the recent bounce in the Dollar occurred as the yield on 10yr Treasuries rose more than those on 10yr German Bunds. Mind you, it has at times been a painful 2017 for traders slow to adjust to the new way of looking at things. Despite two Fed rate hikes, the yield on the 10yr Treasury fell from 2.5% in late Dec 2016 to a low of 2.05% in September, and the Dollar moved in similar fashion ..... not good news if you were predicting that rate hikes would support the greenback.
For the Fed, the change means that their ability to tighten monetary policy has been undermined by the loosening effects of a weaker dollar on financial conditions, which they can no longer influence so directly. As it happens, that's not been too much of a problem with inflation confounding many economists by staying so low. But if that should change, there are uncomfortable precedents of Fed policies not having the desired effect. In 1994, one year after a similarly closer correlation between long term yields and the Dollar, the Fed was forced into surprise rate hikes that not only shocked investors but forced FX traders to focus once again on the short end (even though bond yields had to follow). It was an unpleasant period then, and it's not likely to be much prettier if they're forced to do the same thing again.