Insights into Federal Reserve Policy
Revisions:
- Revised on June 9, 2023
- Revised on June 10, 2023
I reviewed several of John Hussman's monthly newsletters to draw my own conclusions about Federal Reserve policy.
In a Twitter thread from April 10, 2023, I draw the conclusion that what's forefront on the Fed's mind is to signal, through its actions, that it is serious about defending the value of the US dollar. More so than whether recession or unemployment might ensue. The significance of this lies in convincing the world that the US dollar merits remaining the global reserve currency, something which is probably more valuable to the Fed than averting recession or unemployment.
In another more recent Twitter thread from May 24, 2023, I draw the conclusion that it's unlikely that the Fed will ever reduce its balance sheet to pre-2009 levels, meaning that the approximately $8 trillion of liquidity injection since 2008 is going to remain a permanent fixture.
I also draw the following conclusion. If true inflation reduction requires reducing the money supply (and not just raising interest rates), which is how Paul Volcker fought inflation in the 1980's, then the Fed is unlikely to succeed at combatting inflation.
This isn't actually that bad because if nominal GDP rises faster because of inflation, then the national debt-to-GDP ratio is going to fall faster, which is itself desirable.
The "liquidity preference curve" which has been constructed from observable data (copied below and sourced from Hussman's April 2023 newsletter) shows that when Fed liabilities are extremely high relative to GDP, it is impossible to maintain a short-term interest rate that's any higher than the interest paid on reserve balances. This means the following.
First, so long as the Fed is willing to pay interest on reserve balances, the short-term interest rate will track the interest rate that the Fed is paying on reserve balances. It won't be any lower because then no one would be willing to buy short-term Treasury bills. It won't be any higher because the Treasury doesn't want to pay any more interest than it has to.
But then if the Fed keeps paying interest on reserve balances, it could drive itself into bankruptcy! This will happen unless the Fed's assets produce more income than these interest payments. But its assets are nothing other than US Treasuries (probably with medium to long term maturities), meaning that their interest income has to exceed the interest being paid by the Fed. This portends higher rates on medium and long-term Treasuries. But then if these rates are higher, it puts a bigger burden on the fiscal budget ... (Tax hikes coming??? Spending cuts coming???) This story doesn't seem to have a good ending.
Second, if the Fed raises the Fed funds rate further, it will have to pay a higher interest rate on reserve balances. This again puts additional stress on its balance sheet. So, would this mean that the Fed has lower propensity to raise rates further? I wonder.
Third, to the extent that the Fed stops paying interest on reserve balances, the short-term interest rate is likely to fall back toward zero from its current level (of 5%).
In a nutshell, the problem doesn't seem to be the level of short-term interest rates. The problem is the Fed's bloated balance sheet (i.e. residing at the far right end on the x-axis in the above chart). And, so long as the Fed's balance sheet remains bloated, the general tendency of the economy will be to move toward lower interest rates. Money is abundant, so its price (the interest rate) is low. Its price cannot be kept elevated unless it is artificially elevated through paying interest on reserve balances.
And then, if the Fed chooses never to reduce its balance sheet to pre-2009 levels (as I alluded to above), then the only hope is for nominal GDP to grow big enough so that the ratio of Fed liabilities to nominal GDP falls to a more reasonable level. (This corresponds to moving to the left along the x-axis in the above chart, at which time the economy could naturally sustain higher short-term interest rates). More productivity helps in this regard (because it raises real GDP). So does inflation (because it raises nominal GDP). In other words, inflation is the Fed's friend.
Here's another issue. Nominal GDP would have to increase by a factor of something close to 3 in order for short-term interest rates to rise above zero, assuming no change in the Fed's liabilities. (See chart; then, imagine moving from 0.33 on the x-axis -- "you are here" -- to something like 0.1, which is a factor of essentially 3.) If this were to happen over 20 years, it would equate to an annual nominal GDP growth rate of 5.6%. (Proof: (1 + 5.6%)^20 = 3) This seems plausible, but 20 years is a very long time. Over 10 years, the growth rate would have to be 11.6%, which is unrealistic. (Proof: (1+ 11.6%)^10 = 3))
To get there in 10 years, one scenario would be for nominal GDP to grow at 5.6% per annum and for Fed liabilities to shrink by 5.3% per annum. (Proof: (1 - 5.3%)^10 / (1 + 5.6%)^10 = 1/3.) Fed liabilities are currently about $8.4 trillion (see here). A 5.3% annual reduction would require shrinkage by $445B annually or $37B monthly. We can monitor the Fed's actions against this benchmark rate (of reducing its balance sheet). For example, over the past 12 months ending June 7th, the Fed has reduced its balance sheet by ... $528B (see previous source). This is good.
Update (June 10, 2023): I think that the "liquidity preference curve" can be thought of as the price of money as a function of the quantity of money. Toward the left end of the x-axis, when the supply of money is tight, its price (i.e. interest rate) can and will be somewhere materially above zero (as determined by market forces of supply and demand). Toward, the right end of the x-axis, when the supply of money is plentiful, its price cannot and won't be anything other than zero (as determined by the market forces of supply and demand).
Second, all the talk about friend-shoring and trade embargoes between US-China can be explained by a desire to move from the right end of the x-axis toward the left.
Third, the 10-year hypothetical plan mentioned above is likely to be derailed if the US were to be hit by another financial crisis requiring yet another multi trillion dollar liquidity injection.
Fourth, I ask myself, why is it preferable for the economy to be residing toward the left end of the x-axis as opposed to the extreme right end, which is where we are today and where we have never been before?
Author is also on Twitter
