The long-term story on inflation

The claim:

While cyclical inflation pressures should ebb in the near-term, there are elevated risks of structurally higher inflation into 2022 and beyond, which ultimately poses a risk for the dollar exchange rate.

The rationale:

The forecasts which influence monetary policy still assume a significant normalization lower in inflation beyond 2021.

The intuition:

“Inflation is always and everywhere a monetary phenomenon -- in the sense that it is and can be produced only by a more rapid increase in the quantity of money than in output.”

- Milton Friedman, 1956

“I think the downturn in terms of our form of debt crisis won’t just be debts, it will also be pension obligations, healthcare obligations, unfunded obligations…it will be about us having to sell a lot of treasury bonds to the rest of the world and that will also be an issue about two years out… 

We have to sell a lot of treasury bonds. And we, as Americans will not be able to buy all of those treasury bonds. And if interest rates rise too much, the way it usually works is that constricts credit, we borrow less. And that creates a weakness in the economy. So instead, because we will sell to foreigners, from a foreign perspective when they look at it, they care not about inflation they care about currency depreciation, and they look at the interest rate. So if a currency goes down, the bonds become cheaper. I think the Federal Reserve at that point will have to print more money to make up for the deficit, will have to monetize more and that will cause a depreciation in the value of the dollar.”

- Ray Dalio, interview to Bloomberg News, Sept 2018

The evidence:

Over long periods of time, there is a robust relationship between growth in the broad₁ money supply and the rate of inflation (R-sq of 0.66 with a 2-year delay in CPI).

Source: St Louis Federal Reserve; Own Calculations. Broad money uses ‘all currency and deposits’ (prior to 1959) and M2 after 1959.

Part of the recent rise in the light-green line is driven by the Federal Reserve’s quantitative easing program, which increases the narrow₁ money supply. That rate of increase should slow when the Fed announces a tapering, probably in September or December. The chart below puts the level and rate of purchases in an historical context.

Source: Annual Reports of Federal Reserve; St Louis Fed; Own Calculations.

However, there is latent growth in the broad money supply. The burden of household indebtedness is now at 50-year lows relative to disposable income. If household borrowing were to revert to its long-run average (at constant interest rates), this would equate to a 25% increase in broad money (i.e. it would raise by ~2ppts to the 10 year growth rate in the first chart).

Source: NY and St Louis Federal Reserve; Own Calculations.

Another way of visualizing this is to look at the loan : deposit ratio of commercial banks over time. Jamie Dimon noted in his 2021 letter to stockholders that the current levels of liquidity could support an additional $3trn of lending by banks equivalent to about 15% of broad money.

Note: 1969 to 1973 data is interpolated.

At the same time, foreigners are not significantly increasing their exposure to USTs. Starting from 2014, foreigners have only bought 11% of the increase in new marketable treasury debt compared to their 48% share of the marketable UST stock at that time. While the reduction in foreign holdings between 1998 and 2002 was mirrored by a reduction in US budget deficits (into a surplus), the current period is the longest ‘flat’ period since the 1960s which preceded a sharp (~30%) fall in the dollar exchange rate.

Note: Logarithmic scale to show growth rate. Source: US Department of Treasury (TIC data)

So why are UST yields and the dollar at current levels? Despite $5.1trn in net treasury issuance over the past 18 months, over 50% of that has been absorbed by the Fed and another 40% has been absorbed by households and banks due to an unusually elevated savings rate.

Going forward according to the CBO, the U.S. will need to issue on c $1trn on average over the next 5 years when the run-rate of foreign demand is ~$0.2trn, the Fed says it will stop its asset purchases and forward projections of household and bank demand suggest limited take-up.

A rough sensitivity provided by CBO is that 10bps higher interest rate equates to c 3ppts higher debt as % of GDP after 10 years, with the blended interest rate at 1.4% currently. In an interview to MNI, former NY Federal Reserve President Bill Dudley lays out the risks of more dramatic increases in interest rates ahead resulting from the change in the Fed’s approach to inflation targeting:

The Federal Reserve may need to raise interest rates to at least 3.5% and perhaps even above 4% because its new framework will generate lags in responding to inflation that require more aggressive tightening later... "The chance of a soft landing in this regime is virtually nil," Dudley said. They're going to be late and that's by design. So when they start tightening, they're going to have to go relatively quickly." Dudley is surprised markets are only pricing in a rise in the federal funds rate to 2% because "that's lower than what the Fed says is neutral" [in which rates neither support nor constrain economic growth].

With a 130% current debt : GDP ratio, one can clearly see the risk of “fiscal dominance” over monetary policy and a loss of control over inflation expectations.