Thoughts on the savings glut hypothesis

The “global savings glut” thesis was coined by Ben Bernanke in 2005, and refers to the notion that downward pressure on US interest rates was predominantly caused by excess savings in foreign trading partners such as China. One way to view this debate is that it is commonly believed that a countries current account is driving the capital account (i.e. that the latter is the source of financing for the former). Bernanke essentially pointed out the possibility that the causal arrow is stronger in the opposite direction – that strong demand for US assets (i.e. a capital account surplus) was driving the current account deficit.

In their case for the global savings glut thesis, Henderson and Hummel (2008) acknowledge that US interest rates were below their natural rate, but claim that Greenspan’s policy was “tight”. Selgin (2008) responded by downplaying measures of the money supply in favour of the interaction of the money supply and the demand for money. His argument is that if the PY side of the equation of exchange is volatile than so must MV by definition.  Interestingly, The Economist considered this in 2005, using basic IS-LM analysis to demonstrate that an increase in global savings would be revealed as a downward shift in the IS curve, cutting interest rates and reducing output. By contrast a monetary expansion would primarily affect the LM curve, also reducing interest rates but also increasing output. They conclude, that the latter case “seems to fit the facts more comfortably.

This blog post is simply a few fragmented thoughts on the claims made by Hummel, Henderson, and also Justin Reitz. For simplicity I'll refer to their collective position as HHR.

The HHR argument is along the following lines: the Fed only directly controls the monetary base, and it does so to ensure a stable US domestic price level. Since dollars that are held overseas have little impact on the domestic price level we should exclude them. Hummel and Henderson argue that since the ratio of currency to reserves is determined endogenously by the preferences of banks, it is better to focus on the Fed’s control of reserves, and they point out that reserves have approximately been frozen.

They do suggest that in theory one should really focus on the domestically held monetary base, and Rietz proceeds along these lines. He argues that since the currency held overseas has no direct impact on the domestic price level this should not count as evidence against a loose Fed policy. Indeed when foreign holdings of currency is stripped out he shows that the monetary base expanded by merely 2% per year. This supports Hummel & Henderson.

HHR makes an intriguing claim – not only that the Fed has been pursuing a de facto free banking monetary policy (i.e. keeping the base essentially frozen and then allowing fractional reserve banks to expand and contract the money supply on top of that to respond to changes in the demand to hold money), but that this is the true cause of the Great Moderation. No wonder they are surprised that Selgin doesn’t want to take credit for it! Rather, the bit that Greenspan was directly responsible for resembled a Free Banking environment.

But of course the incentive system within the existing regime (e.g. deposit insurance etc) means that the demand for money does not reflect voluntary behavior. HHR would say “but that’s not the Fed’s fault! Blame congress” and I sympathise. But regardless of whose fault it is it is at best an approximate free banking environment, and no one has argued that by resembling a free banking regime on one margin (whilst other margins are non free banking) is an improvement. Again, this comes down to whether you treat the regime as fixed, or expect the central bank to try to use what powers they do have to compensate for problems with the regime.

HHR don’t argue that the Fed didn’t play a role in generating the crisis. They merely show that it wasn’t within their narrow remit to prevent it. In isolation, a Fed that tries to limit its sphere of influence and sticks to approximating a free banking policy is a good thing. But in a world were governments create even bigger errors – whether foreign (in terms of their currency manipulation) or domestic (in terms of their addiction to debt financing) – maybe we should hold central bankers to a higher standard.


Special report on natural interest rates

I've made two previous efforts to calculate the natural interest rate for the UK, and you can find them here:

I've just written a short paper talking about why the natural rate is important, and a (very) brief summary of some recent efforts to estimate it. You download it here. The charts are up through 2015, but I thought it would also be interesting to incorporate it in our data section. The chart belows shows the present situation, and as of 2016 Q2 the natural real rate was 1.8%. Using the GDP deflator as an inflation measure, this implies a nominal rate of 2.34%. With market rates (I use SONIA) at 0.47% this implies monetary policy is too loose.


The lethargic recovery

I was talking today about why I thought the UK's recovery was lethargic, back when we thought that it was (we now think it was signficently better). There's obviously  a number of factors, and I wouldn't paint government policy as the sole determinant. We can split them into 6 main categories:

  • Falling real incomes: Household consumption has been squeezed via low wage increases and high inflation. CPI has been above the 2% target since December 2009 reaching 5.2% in September 2011. As already mentioned huge open ended tax liabilities and tax uncertainty has dampened spending with signs that consumers are factoring future debt burdens into their present consumption choices, “indebted consumers seem more interested in paying down what they owe than splashing out on flat-screen televisions” 
  • Low business confidence: not just because of low demand, but also due to Robert Higg’s concept of “regime uncertainty”. This results from the erosion of investor’s confidence in private property rights, and there is evidence that this occurred in the UK following the crisis.
  • Distressed export markets: since the financial crisis the trade-weighted value of the pound has fallen by about 20%, however 45% of exports go to countries within the Eurozone meaning that the UK is highly dependent on Eurozone growth.
  • Breakdown in financial intermediation: Bank lending has been weak, in large part due to new Basel rules that intend to encourage banks to hold more reserves. Evidence suggests that any gains from quantitative easing were almost totally offset by stricter capital requirements imposed by regulators.  In addition the recession directly followed a banking crisis that resulted in the government nationalising the four largest lenders in the country.  According to Reinhart and Rogoff, “the aftermath of banking crises is associated with profound declines in output and employment” , and the conventional view is that balance sheet repair takes time.  This may especially hold if accompanied by a housing bust – and the UK had one of the largest housing bubbles.
  • Regulatory problems: not only had much of the productive capacity of the economy been hollowed out prior to the financial crisis, there has been little supply side reforms as part of the recovery. The list of regulatory reforms in recent Budgets is underwhelming, treating government investment in infrastructure as being synonymous with supply side growth. Meanwhile airports are constrained by planning laws, housing developers can’t build new housing and small businesses are stifled by red tape.  In addition high marginal tax rates across the tax schedule dampens incentives and hinders growth.
  • Monetary policy mistakes: Similar to the US, interest rates in the UK were kept artificially low in the period building up to the financial crisis creating distortions in the economy. This “malinvestment” sowed the seeds of an inevitable correction, but these problems were compounded by additional monetary policy failures. In terms of the crisis period itself, nominal GDP began to collapse in early 2008 and didn’t reach its pre crisis growth rate of around 5% until late 2010. Some argue that the Bank of England was slow to respond to this – interest rates were 5% in February 2008 and they only began cutting in October (to 4.5%). Quantitative easing didn’t begin until 6 months after the collapse of Lehman Brothers, in March 2009.

In short, there has been a combination of reasons why the UK economy recovered thw way it did. It was vulnerable to a recession and monetary mismanagement compounded fiscal folly. 2008 wasn’t a temporary, irrational pause in spending but a permanent wealth shock.


MA growth accelerates

Some important monetary aggregates have shown growth recently, and MA is no exception. In July 2016 it was growing at 10.28% (compared to the previous year), up from 9.27% in June. The chart below shows the growth rate over the last calendar year.


UK Gross Output grew by 3.49% in 2014

Last month the ONS released the Supply and Use Tables for 2014. You can find them here. They are interesting because they provide a measure of intermediate consumption, which many Austrian economists - who care about the entire "structure of production" - believe is an important missing ingredient of typical national accounts. Indeed it's somewhat odd that Gross Domestic Product strips this economic activity out, making it more of a Net concept. As Sean Corrigan puts it, we want the Hayekian horse (of production) in front of the Keynesian cart (of consumption).

Mark Skousen has advocated a measure that he refers to as "Gross Domestic Expenditure", which incorporates all of the production side of the economy. A close substitute for this is "Gross Output", which, as I've previously mentioned, is now published by the BEA (Skousen adjusts Gross Output by adding Gross sales at a retail and wholesale level, to incorporate even more business spending).

I've made previous efforts to measure Gross Output in the UK, using the Supply and Use Tables. This time, I'ive modified the method. I think the simplest way to approach it is to simply combine NGDP with intermediate consumption. Doing so provides the following side-by-side comparison:


Similar to US estimates, we see that Gross Output is around (in fact just under) twice as large as nominal GDP. The interesting comparison is the growth rates, and Gross Output is more volatile than NGDP:

In 2006 it was growing at 7.88% which indicated an even larger boom that was being shown in NGDP data (which grew by 5.52%). It then contracted by -2.31% in 2009 before picking up again. As with NGDP the post crisis growth rate seems enduringly lower. My main interest was the 2014 figure, and we can see that whilst Gross Output grew faster than NGDP in 2013, this was reversed in 2014. Whilst NGDP delivered a robust 4.77%, Gross Output only grew by 3.49%.

The problem with offering an alternative to GDP is there's a burden to provide a better one, or a more theoretically robust one. I wouldn't claim that I've achieved that, but I think it's worthy of enquiry. And if the recent debate between Vincent Geloso and Scott Sumner (and Marcus Nunes and Vincent again) is anything to go by, maybe there's increasing interest in getting this right.