In our session on exchange rates, we introduced the concept of interest rate parity. The basic idea is that efficient market forces will ultimately combine to produce a global benchmark for safe real interest rates. Safe usually means the short-term rates paid by State institutions such as the Fed, the US Treasury and their equivalents in other parts of the world. And short-term means just that – the yield paid on debt maturities that are often significantly less than one year. For such financial instruments, any interest rate divergences will prove temporary and associated with real exchange rate disequilibria.
Once we move beyond these short-term official rates it gets more complicated. Market rates, the longer-term interest rates – both official and private – that drive consumer and investment spending, typically involve risk. However, uncertainty is variable, not fixed, and depends on circumstances – not least the business cycle and mercurial expectations.
Moreover, beyond the short term, uncertainty is not confined to credit risk (where payment promises – if they exist – might not be honoured). Market rates will be buffeted by varying risk premia connected with defaults, market sentiment, inflation and economic prospects. As we learnt in session 7, asset prices – even of high credit quality US Treasury Bonds – can change simply because of fluctuations in policy or the inflation outlook. For most advanced economies, governments are hardly likely to default on their borrowing. But, US Treasuries and other government bonds can be highly volatile if investors alter their views about the likely path of short-term policy rates, perhaps because the inflation target seems less likely to be achieved. Even prospective moves in the regulatory environment or tax treatment of government debt could shift market prices, creating headaches for investors despite secure cash flow promises.
In Session 13, our penultimate teaching week, we explore the interactions of mobile capital, floating exchange rates, policy choices and international economic performance. As with our domestic AD-AS model we soon discover that fickle expectations play a key role. However, unlike goods and services, global financial asset prices are far from sticky. Valuations are highly sensitive to changes in beliefs. Moreover, such expectations are not always firmly rooted or classically rational. Bob Shiller’s article on potential reactions to the Fed’s widely-expected rate hike in December 2015 is highly recommended. His references to behavioural finance are worth noting with the theme of human psychology prominent in the literature on rational inattention. And take a look at the worked examples especially the topic of risk-on and risk-off dollar trading. Our session touches on the point that, when analysing shocks, it is important to assess whether that shock is expected to be temporary or permanent. Similarly key is whether markets see the “”shock” coming (an anticipated shock) or whether it comes completely out of the blue.
The November 2016 Presidential election heralded the adoption of both looser fiscal policy and tighter monetary policy. The standard prediction, as we confirm with the aid of a Mundell-Fleming model, is that real interest rates rise and the US dollar jumps sharply higher (creating expectations of future depreciation). Currency overshoots frequently occur and are in line with interest rate parity ideas. Given efficient arbitrage, high US interest rates can only co-exist with low European equivalents if the dollar is expected to weaken. What is gained on the yield roundabout is lost on exchange rate swings. That is exactly what interest parity means.
To digress, if only to underline earlier warnings about accounting pitfalls, do not necessarily assume that the much talked-about US corporate profit repatriations – prompted by the 2017 Tax Act – will necessarily boost the US$ through capital inflows. As noted here most of the corporate cash mountain is already in dollar-denominated assets!
In the wake of the 2016 US Presidential election, the dollar behaved pretty much in line with the standard model. After the initial post-Trump surge, the real broad effective US$ index fell by around 10% between end-2016 and end-2017. However, for a more rigorous examination of events, we turn to some empirical studies on policy divergence.
For example, IMF analysis confirms the importance of risk premia and interest rate differentials. A July 2016 study presents a plausible narrative explaining 2014-16 US$ strength in terms of risk premia and yield differentials. There is also evidence of a dollar safe haven effect in the context of heightened global risk.
A 2016 Fed study on international spillovers of monetary policy tackles another important theme. Specifically, changes in US monetary policy can significantly impact foreign economies. Taking the example of GFC-related cuts in US interest rates, the analysis shows the quantitative importance of three distinct channels:
a) exchange rate: a softer dollar, reflecting lower interest rates, will boost US exports and so weaken foreign GDP
b) domestic demand: but, in the opposite direction, better US GDP performance – on the back of a weaker dollar – sucks in imports, so providing offsetting benefits to overseas producers
c) financial spillovers: lower US interest rates will pull down foreign asset yields so benefitting foreign GDP (although note BIS concerns about potentially greater risk-taking in emerging markets)
The Fed’s empirical work suggests that a) and b) net out. In other words, the overall effect of US monetary easing, represented by c), is GDP-positive for the world economy.
Of course, it is not always a one-way street as our feature video makes clear. Foreign developments can also independently impact the US. America is a large and powerful country, but we should never underestimate the importance of global networks and feedback effects. The dollar can change not just because of what happens in the US but also because of overseas shifts in asset prices and risk tolerance.
In quantifying these monetary spillover effects the Fed makes use of its SIGMA model, one of several that US officials use in framing interest rate and other policy decisions. As such, the empirical studies cited above are a useful prelude to our final session that focuses on large-scale macro models – both conventional DSGE models widely in use amongst central banks as well as increasingly popular Agent-Based Models (ABM).
Throughout the course we have gradually expanded our modelling expertise. But you will no doubt be aware that macro-analytical techniques faced extensive criticisms after the GFC. Are these models too pointy-headed, too remote from reality, to be of any use? Hopefully, I can persuade you that your sampling of these quantitative techniques has indeed been worthwhile.
2 May 2019