I went to the excellent Barclays Inflation Conference a couple of weeks ago – although titled “inflation”, a lot of the conference’s content concerned the growing fears about the solvency of western governments. In particular, whilst the US Treasury market is currently seeing a massive flight-to-quality bid (10 year yields are now down below 3%) I came away worrying that it’s difficult to see that the US has any plan to avoid medium term bankruptcy other than some hopeful reliance on the American Dream to magic it all better.
Ken Rogoff (Professor of Economics at Harvard, and co-author of This Time Is Different) accused the US Treasury of “playing the yield curve”. With yield curves still extremely steep by historical standards, the authorities have skewed issuance to short maturities with the lowest interest rates (even though long dated maturity bonds would have historically low coupons despite the steep yield curve). Around half of all US debt will mature in the next three years – a tactic which keeps the US’s interest payment burden down in the short term, but which is a “classic way” of triggering a financial crisis when rates start to rise. This is the shortest debt maturity profile for the US since the 1960s. A huge burden of debt refinancing, coupled with higher interest payments was the trigger for Greece’s recent debt crisis. It’s another reason why we disagreed with Bill Gross’s “nitroglycerin” comments regarding the UK – the average maturity of the gilt market is around 14 years, compared with under 5 years for the US and 6 and 7 years for Germany and France. A “buyers’ strike” should be a little less problematic for the UK than it would be for the other nations.
Rogoff also talked about the prospects for financial repression as a method for governments to create a buyer for their debt when the natural, economically motivated, buyer has disappeared. Financial repression is the process of making people own assets they don’t want to hold – and the financial regulator is the important driver of this. In particular banks are encouraged (or forced) to hold more of their assets in less risky assets – i.e. government bonds – but also pension funds and individuals might find themselves being nudged into government bonds (in Japan individuals have most of their savings in the Japanese Post Office, which invests those savings in JGBs). Once domestic buyers are handcuffed, it becomes much easier to use inflation as a tool to reduce the real debt burden, especially in an economy like the US which has been steadily reducing the amount of inflation-linked debt it has outstanding as a percentage of the overall debt mix (although see comments below about the other inflation-linked government liabilities which stop inflation being the magic bullet policy tool).
Finally Rogoff said he’d be astounded if many Eastern European governments (and Greece) did not default, even with the IMF helping them. He pointed out that an IMF rescue package doesn’t always mean an economy is saved; in fact in 1/3rd of the IMF programmes since the 1970s default has ensued (including Argentina, Indonesia, the Dominican Republic and Turkey).
If you were nervous about the US keeping it’s creditworthiness after Ken Rogoff, a speech by Ajay Rajadhyaksha (Barclays Capital’s Head of US Fixed Income) piled on the anxiety. First the good news – the role of the US dollar as the primary reserve currency allows it to run excessive deficits far in excess of its economic rivals. Barclays have modelled the US’s AAA credit rating with an overlay based on the percentage of the world’s reserves kept in US dollars. On a stand alone basis, the US should have a AA credit rating, like Spain – but currently the US$ makes up 60% of global currency reserves, and this would allow them to run a 200% Debt/GDP ratio without losing their AAA rating, compared with the estimated 90% Debt/GDP level now. If the US$ became a bigger portion of global reserves (65%) then a 250% Debt/GDP ratio could be sustainable. Under current projections, only a fall in the dollar’s share of reserves to 50% would trigger the downgrade to AA. Even with continued diversification away from the dollar by foreign investors, this looks a long way off. However, once it happens the acceleration is severe – when Japan lost its AAA rating the yen fell significantly as a percentage of foreign portfolios, perhaps helping to trigger Japan’s further ratings downgrades.
That was pretty much it for the good news. Even at current low levels of interest rates, the US’s debt servicing costs take a step upwards in coming years, as the Debt/GDP ratio rises to 95% by 2020. If yields were to rise by 2% across the yield curve the percentage of US government revenues spent on debt service would rise from a troubling 17% now to around 33%! And inflating away that debt burden doesn’t work very well, as so much of the government’s outlays are indexed to inflation (although I guess you can always do what George Osborne did in last week’s UK Budget and change the inflation measure used to index benefits to one that is structurally lower, CPI rather than RPI). Radadhyaksha was also nervous about the US government’s contingent liabilities – losses on mortgages held by the GSEs (e.g. Freddie and Fannie) could be in the realms of $300 bn+. But the biggest contingent liabilities are the entitlements due to the US populations – and predominantly Medicare costs. After 2020, for every $2 trillion of taxes raised, spending will be $3.5 trillion. How do you close that gap without triggering a popular revolt, especially in an economy where median household incomes are only at the same level that they were back in 1998/99? Senator Judd Gregg, who some expect will run for the Republican Vice Presidential nomination next time round and sits on the Senate Budget Committee, suggested that the answer was to slash entitlements and cut taxes – this combination will encourage entrepreneurial spirits and reduce the deficit. It’s one possible outcome I suppose. (Earlier Ken Rogoff suggested that the Federal tax take needs to go up by a massive 25% to put a dent in the deficit.)
A panel session with Adam Posen of the UK’s MPC, and Former Fed Governor Larry Meyer asked whether Central Banks’ independence is under threat from concepts like Quantitative Easing (buying government bonds as part of the monetary policy, but also incidently (?) keeping yields down at times of budgetary pressure – not unlike the trigger for the Weimar Germany inflation experience), and some increasing commentary about Central Bank inflation targets being too low (including from the IMF’s research director Olivier Blanchard who thinks that 4% would be more like it). Posen believed that as long as a government is unable to fire the Central Bank Governor, and that the Bank is not made to buy government bonds in the primary market then independence is safe (although I didn’t get why there should be a difference between the primary market and secondary market). Most importantly, independence is not about legislation, but about a “buy in” from society – for example, the Bank of England was able to be made independent in 1997 because it had gained anti-inflation credibility in the preceding years, rather than prices subsequently falling because it was made independent. Meyer did, however, worry that the US Federal Reserve was more vulnerable to political interference than in the past – there was currently extraordinary hostility to the Fed from Congress as the result of the Fed’s bailout of the banking sector, and its new lending powers. Further more as fiscal deficits become unsustainable, could the Fed really hike rates in a world where the US needs to rollover half its debt every three years without triggering a downgrade or default?
Now for a word on the inflation measures that we use. I’ve lost track of the times that people have told me that the RPI, CPI or some other measure systematically under-report inflation – or that these measures are useless for pensioners, who don’t buy iPads, Blue Ray discs and SuperDry T-Shirts (sub-editors – please check that these things exist). Dean Maki (Barclays Chief US economist) and John Greenlees of the US Bureau of Labor Statistics put paid to a few of these inflation myths, and in particular pointed to the famous Boskin Commission Report of 1996 which concluded that in fact the US CPI measure was actually overstating inflation by something like 1.1% to 1.3%. The reasons why inflation measures tend to overstate actual inflation include substitution bias (the basket of goods doesn’t change to reflect the fact that if the price of something rises, consumers will switch to a cheaper alternative), outlet substitution (not capturing the lower prices charged by a new Aldi store in the data for example), quality change (more reliable goods with higher specifications) and new product bias (price deflation is often seen in new technology for example, but it may take a while for that new technology to enter the inflation basket). Another big complaint people have about CPI measures is the treatment of housing, and especially the US concept of Owners’ Equivalent Rent (OER), which is supposed to reflect the implicit costs of owner occupancy (“if someone were to rent your home today, how much do you think it would rent for monthly, unfurnished and without utilities?”). Recently OER has depressed inflation, causing critics to claim that this is somehow fiddling people with incomes linked to CPI. However, over the past 20 years or so, OER has actually boosted the CPI in most periods. The US does have an experimental measure of inflation supposed to better reflect the basket of goods for a pensioner (CPI-E), but it is only very marginally higher than the ordinary CPI. In fact there are no serious studies to show that western governments have suppressed the inflation measure to save money on inflation linked outlays (Argentina is a very different story however!) – the widely used inflation measures usually overstate inflation, which means both that inflation-linked bonds are good hedges for experienced inflation, and that there is a bias towards pensioners and other recipients of inflation-linked incomes being overcompensated.
If the CPI measure is so robust why does the Federal Reserve like to use the PCE deflator (personal consumption expenditures price index) as its preferred measure of inflation? Firstly it is chain-weighted, so it’s more flexible in changing its composition weights to reflect cost-conscious goods substitutions, and secondly, unlike the CPI measure the PCE deflator can be revised historically along with the GDP numbers as fuller data is received. Because the CPI is used to calculate things like bond coupon payments, once released it never changes. The real cynic would additionally say that it is because the PCE deflator is usually lower than the CPI!
Finally, a senior sovereign analyst from one of the major ratings agencies was asked whether there has been any pressure on him from AAA sovereign issuers imploring him to leave their ratings unchanged. The terse reply was “There has been absolutely no pressure from the US or UK authorities”. I wonder who has been calling?