# Central Banks, Large Scale Asset Purchases and Inflation

First, the overview:

With the fractional reserve system, money/credit is primarily created through loans on deposits but the bank is required to maintain their deposits up to a certain amount (roughly 10% I think). The other primary avenue of money creation is by the Federal Reserve printing money, but there is severe reservations on printing money given past episodes of hyper inflation, though recent events with quantitative easing and large scale asset purchases, the Fed's balance sheet has reached over 4 trillion dollars. The Fed has stated that any profits accumulated from these assets will be reinvested into new securities when they mature and any additional profit will be deposited into the treasury.

Here is my question:

If the federal reserve were to regulate the banking industry so that fractional requirements were raised to something like 50% on all future deposits, that would have an appreciating effect on the dollar since the rate of increase in the supply of dollars needs to keep up with the growth of the economy every year. If the Fed offsets this appreciation with additional asset purchases by printing money, the Fed's revenue/balance sheet would increase... and over time that excess profit could start to make up the overwhelming portion of the fiscal budget (which could take decades or centuries)... Historical issues with the central bank funding the budget have always resulted in hyper inflation, but with this work around, where the funding of the budget was not coming from the direct printing of money but from the profits from security purchases as they matured, hyper inflation would be avoided, right?

Please explain why this is a bad idea... In theory the banking system wouldn't have a noticeable impact since, as additional money is created to offset the increased fractional requirements, that money will be deposited into the banking system (where else would it go) increasing holdings and increasing the money available to be lent out...

This is in essence what the Japanese have been doing with paying off Government debt in order to prevent their currency from appreciating naturally (instead of raising reserve requirements), right?

Response (wasn't allowed to respond at length so I'm making an edit to my original post):
Lending would have "some" reductions, but it wouldn't be as severely impacted as it would if it weren't offset by more asset purchases, purchases that are then deposited into banks, expanding the monetary base which then expands the amount to be lent out... There's also the argument that, by replacing tax's with the profits from maturing assets, there would be a stimulating effect on GDP, which would cause the dollar to appreciate, which would then require more printing or lending to maintain stable prices... though of course it would take a significant amount of time for profits on maturing assets to become a sizable portion of the fiscal budget, but of course the FED isn't purchasing securities with the idea of maximizing revenue so I'd imagine their rate of return could get a little better if they really tried and this where to become a primary mandate...

I'm definitely not saying raising the supply of dollars increases their value, the opposite actually... My argument says raising reserve requirements would reduce the supply of dollars, but printing would offset this reduction, maintaining the equilibrium. I suppose an implicit assumption I didn't spell out was that the amount of dollars printed would not equal the number of dollars no longer being lent out since fractional banking has the multiplier effect, i.e. today one dollar created by the FED is equivalent to 9~10 dollars lent out in the banking system.

Here is my question:

If the federal reserve were to regulate the banking industry so that fractional requirements were raised to something like 50% on all future deposits, that would have an appreciating effect on the dollar since the rate of increase in the supply of dollars needs to keep up with the growth of the economy every year. If the Fed offsets this appreciation with additional asset purchases by printing money, the Fed's revenue/balance sheet would increase...

Reserves are essentially a deposit at the Federal Reserve by banks. Banks have a minimum required level of reserves; anything beyond that are excess reserves. As a result of "Quantitative Easing", American banks already have large amounts of reserves. Raising the reserve requirement would just change the status of the reserves from excess to required, and could easily have little effect on the economy. The precise effects of Quantitative Easing on the economy are debated, but they are certainly not very obvious.

In other words, a policy very similar to raising required reserves was already tried, and had little observable effect.

One thing to keep in mind that reserve creation is distinct from bank money creation. When a bank extends a loan, it creates an asset (the loan) and a liability (the deposit) simultaneously. This step is done independently of existing reserves, or central bank activity.

However, this deposit creation creates a need for extra reserve holdings (during the next reserve accounting period). If excess reserves exist in the banking system, the bank can just go out and borrow the reserves from other banks. That is, lend first, get reserves later. If excess reserves do not exist, the central bank has to create some - if it wants the inter-bank rate to remain under control. Otherwise, some banks will fall short of regulatory requirements, and are effectively illiquid, and will be willing to borrow at extremely high rates to meet regulatory requirements. In other words, reserve creation follows deposit (bank money) creation.

(In case it is not obvious, once the Federal Reserve creates reserves, about the only way they can be destroyed is by currency (notes and coin) withdrawals, or by net tax payments being transmitted to the Treasury. An individual bank can transfer/lend them to another bank, but this does not change the amount outstanding. Therefore, the Fed was able to force an increase of reserves without raising requirements.)

Textbook descriptions of fractional reserve lending tend to be oversimplified. Only some accounts are covered by reserve requirements, and banks have been working techniques to minimise their costs for decades. A lot of "money" is in other money market instruments ("shadow banking") which are not covered by reserve requirements. Raising reserve requirements has the effect of making regulated banks less competitive versus their shadow banking competitors.

Setting a reserve requirement of 50% (for example) would mean that a bank might have to issue a \$50 bond in order to be able to make a loan of \$100 safely. (It has to finance the massive amount of reserves that are imposed on the asset size of its balance sheet.) It would be borrowing in the bond market in order to leave money on deposit at the central bank at the overnight rate.

Balance sheet assets:

• \$100 loan; and • \$50 reserves.

Liabilities:

• \$100 deposit; and • \$ 50 bond.

The spread on the \$100 loan would have to be increased to make up for the negative spread on the \$50 in reserves (versus the bond). The banking systems ability to lend would be largely crippled, and firms would be forced to switch to alternative sources of financing. If firms cannot get financing, they cannot invest or hire more people, and that makes capitalists and politicians unhappy.

The main economic effect of required reserves is that it is a means of forcing banks to lend to the central bank at a low rate. Until the Fed was given the right to pay interest on reserves in 2008, banks received no interest, and so reserves were an interest-free loan to the government. Even if reserves pay interest, it is at an administered rate set by the central bank (which effectively becomes the short-term risk-free rate). Banks have no choice but to accept that rate of interest, whereas Treasury bill/bond rates are set in auction. (In practice, Treasury bill rates trade close to the policy rate.)

This policy of forcing banks to lend to the government (via the central bank) is often called "financial repression." This matters more for countries with poorly developed bond markets, but the U.S. Treasury can lock in low rates for a very long time without much difficulty.

and over time that excess profit could start to make up the overwhelming portion of the fiscal budget (which could take decades or centuries)... Historical issues with the central bank funding the budget have always resulted in hyper inflation, but with this work around, where the funding of the budget was not coming from the direct printing of money but from the profits from security purchases as they matured, hyper inflation would be avoided, right?

What percentage of bonds are purchased by the central bank is not an issue. The Bank of Japan has been holding large amounts of Japanese Government Bonds for years, and inflation has not budged. The key is that the deficit is very large (working from memory, some cases were 30+% of GDP), and they were often associated with losses of productive capacity. In other words, the big deficits were the problem behind hyperinflation; a central bank buying existing debt is just a reshuffling of government liabilities.

Attempting to run "too large" deficits and trying to use central bank purchases of bonds reserve requirements to control inflation will not work in the long run. The velocity of money would just increase as people move towards shadow banking. If done for a long enough time, people can just switch over to using a foreign currency, which is usually a critical step towards hyperinflation.

In summary, running a deficit of 1% of GDP and having the central bank buy all the debt (monetising it) is not going to have much of an impact on inflation. (The recent experiments in quantitative easing provide a test case.) Running a deficit of 30% of GDP is probably going to cause serious inflation (unless the country switches over to rationing, such as in World War II), and what the central bank does with its balance sheet is not going to make a big difference.

• Stack exchange doesn't give you enough space for in depth responses so I've copied everything into this google document, your words are in red and quoted with >> << docs.google.com/document/d/… – Jacob May 5 '17 at 19:20
• One possibility is that you break out some of your questions into new questions. The preferred format here is to have small, focussed questions that have tightly-defined answers. This way, the questions/answers can be useful to new users. (You can link to this question.) I will try to add some clarifying comments here shortly, but some are opening up new territory. – Brian Romanchuk May 5 '17 at 20:29

It seems your premise is that by increasing the reserve requirement ratio and the monetary base in tandem we could essentially leave lending as it is and that this would not devalue the dollar, all the while the Fed could massively expand its balance sheet and start making a ton of money for the federal government. Where I see this breaking down is that it seems to ignore the quantitative theory of money, i.e. you're confusing raising the supply of dollars (or the monetary base) with making the dollar worth more, when this would almost surely make the dollar worth less. If MV=PQ and M increases fivefold (to be consistent with 10% RRR to 50%) then unless your plan gives any reason for velocity falling or GDP rising inflation ought to skyrocket (assuming the money supply rises fivefold).

On your second to last paragraph, if you merely offset the RRR with new money then you've held all the ratios constant and I don't see a rationale for expanded lending.

Also, your question is contradictory, as you say you avoid printing money to expand the monetary base but you initially purchase trillions of dollars in assets by printing money. Perhaps you could clarify that point.

• Sorry, mistakenly edited your post and wasn't allowed to retract the edit (just added the ...... to the end), I've added to my original post with a response to you since stack exchange wouldn't allow such a long comment – Jacob Mar 27 '17 at 22:58
• I think my main objection is that such a huge increase in Mb and Ms would have such inflationary tendencies as to drastically outweigh the impact of anything else you've posited. Real money supply is roughly constant from what I understand, and so the changes you estimate I doubt would happen in real terms. Lowering taxes may stimulate demand, but if the government is making the same real income from investments the Fed holds then this would just bankrupt the federal government. – doubletrouble Mar 28 '17 at 2:48
• It's my understanding that the monetary base was never constant or meant to be constant, increases here and there weren't necessarily reflected in the FED's balance sheet... like operating budgets not funded from the treasury, but also, increasing the monetary base would be data dependent (near comprehensive with modern IT/banking) and maintaining stable prices would be the first concern of course... I argued this process would be gradual (decades/centuries) and deliberate... the 50% figure was just a placeholder, I'm sure there would be a more technocratically justifiable figure than 50%... – Jacob Mar 28 '17 at 20:37
• Ah, according to his chart, there does seem to be a bit of a take off there in 2008 fred.stlouisfed.org/series/BASE lol though the chart does show it gradually increasing over time and it still seems that what I am asking is possible... or at least I am having a hard time understanding why it wouldn't be possible or at least a good idea, especially now with modern computers and the data dependency of these sorts of things – Jacob Mar 28 '17 at 20:48