How the world works – Fractional reserve banking

In this first post, I will try to explain maybe the least understood and most important part of our financial system, the fractional reserve banking system, and its limitations. It will most likely end-up as a quite long post, as I want to be as clear as possible.

Let’s imagine a really simple system, comprised of 1 company (C1), 1 bank (B1) and the central bank (CB) of this system.

at the start, the company is starting its business, and has therefore 0 asset, 0 liability (let’s assume the full company assets will be funded by the bank).

The balance sheet (BS) of the various actors, at the start, is as follows:

Fractional reserve1

Then C1 requests, and obtains, a loan from B1 to develop its business. B1 creates the money “ex-nihilo”, by adding on its balance sheet, a loan to C1 on the asset side, and crediting the current account of C1 at the bank.

On the same day, C1, not ready yet to use the money, put the money back at the bank as a deposit. In our system, there is only one bank, so B1 it is.

Fractional reserve2

Here, let’s stop a moment to make several important remarks:

  • First, notice that the principal for the loan has been created by B1, but not the interest that will be necessary for C1 to repay the loan to B1 in the future. C1 will have to source the interest from outside. This has huge implication for the financial system; we will get back to that later.
  • We assume that C1 puts back the money to B1 as a deposit. In reality, of course C1 would spend the money to develop its business, but this money would then end up in C2 or C3 balance sheet as “cash” and at some point this cash is bound to be deposited back to a bank. Of course the money may not end up back to B1, but maybe in another bank, such a bank would also emits new loans the same way B1 is doing, and some of that new “money as credit” would similarly end up in B1 balance sheet. If we assume C1 represents all the company in a complex system and B1 all the banks, then the money lend by B1 to C1 will end up back on B1 balance sheet as a deposit at some point.
  • An important point is therefore that once money is created through loan creation (let’s call that “money as credit”), this “money as credit” never leaves the banking system unless:
    1. The “money as credit” is kept as physical cash outside the banking system (limited option as there is only a fraction of the whole amount of “money as credit” moving in the system compared to the physical cash i.e.: bills and coins).
    2. The loan is paid back by C1 to B1, in which case the situation goes back to (1) above, except for the interest that C1 should have paid to B1.
    3. C1 defaults on its obligations to B1, in which case on B1 balance sheet, the loan to C1 disappears and on the liability the Capital of B1 disappears: a loss to B1.
  • As we have seen, by creating new loans, the size of our simple economy’s money supply has expanded. Making new loans is indeed “inflationary”. A contrario, paying back a loan makes the economy goes back from stage (2) to (1). It reduces the amounts of “money as credit” floating around (money supply) and reduces the size of the economy. This is therefore deflationary.

B1 can create almost unlimited loans that way. The only limitation is what is called the “reserve ratio”:

Fractional reserve3

In case (3) above, B1 has made a lot of loans to C1. Its ratio of loans (1000 in bold red above) to reserve (100 in bold black above) is now of 1000/100 = 10. This reserve ratio is mandated by the CB as a rule. It can change over time. If, as in above example, the maximum loan to reserve ratio allowed by the CB is 10, B1 cannot make any additional loans before first having its loan to reserve ratio go down.

Suffice it to say that in our current economy, banks are not close to their maximum loan emissions at all, I will explain why later in the post. Nevertheless, most of the CB “extraordinary” measures taken place since 2009 have been to increase the reserves of the Banks, hoping that with more reserves, the banks would make more loans and grow the economy. This is why the CB actions so far have been mostly useless for the physical/real economy (the economy of people and company) and have not resulted in growth. It has nevertheless had huge impact on the financial economy (bank balance sheets, stock market size etc.).

Before explaining why those CB actions had no impact on the real economy, let me show, in accounting term, the Quantitative Easing (QE) impacts. For that I will had an additional actor, the government, G1. QE so far has mostly (not only) taken place by the CB purchasing government bonds, which is why I add this actor. This actor is not mandatory though, as QE would work exactly the same way with the CB purchasing company bonds or stocks (it is actually the case in Japan QE and to a lesser extend in US and Europe QE).

Fractional reserve4

I have restated the BS of the various actors a bit to balance the whole thing and show several things:

  • The government issued some bonds to finance its activities (the usage of the money is not important here, so the assets are mentioned as XXX).
  • Such bonds have been acquired half by B1, half by C1 (instead of keeping its treasury as cash, the company elects to have some government paper, really common occurrence).
  • In the above situation, B1 has reached its loan to reserve ratio of 10 (500 loan for 50 reserve, the rest of the reserve having been used to purchase the government bonds)

Now what is the impact of QE?

Fractional reserve5

The CB purchases the bonds from B1 by replacing such bonds with newly created reserves. Those reserves appears ex-nihilo, they are not coming from the initial Capital of B1 (as in example (1) above).

It has several far-reaching impacts:

  • First, as when B1 is issuing a new loan, it expands the size of the money supply, it is therefore inflationary.
  • Second, now B1 has new reserves, and its loan to reserve ratio is back to 5, it can issue for 500 more of new loans to B1. Should it do so, it would be really inflationary.
  • Third, by adding a new buyer for government bond in the market (the CB), demand for such bonds increases. Since supply did not (in our above example between (4) and (5), there is no more government bonds issued), it means government bonds value increased -> Government bond interest rates decreased.

Impact of the decrease in interest rates on government bonds:

  • The value of the bonds on the Balance Sheer of C1 have increased, it could sell them back to B1 for a profit. B1 would then be in a position to sell that to CB through more QE for even more effects.
  • The government can now issue extra debt, or roll existing one for cheaper, giving it more money to spend on something else than interest. It is one of the main effects of QE, keeping the government rates low, as if rates were where they should be (higher than now with QE), most of the developed countries government would not be able to service the huge amount of debt they have.
  • The problem here is that high interest rates are a good limit on government profligacy; look at the government debt level of “crappy” country like India or Russia, which have high interest rate: they cannot borrow cheaply so they do not borrow. Western governments have not used the extra savings from cheaper rates to try to reduce their debt; instead they have piled on more debt, since it was cheaper (lower cost of debt -> increased demand…).
  • All rates are impacted by a decrease of the government bonds rate (housing loan rates, credit cards rates …), since all else equal, a buyer would find government bonds less attractive after QE since the rate is down and would consider buying another type of asset, increasing its price and decreasing its yield. Therefore all rates would go done as well somehow.

Impact of all rates going down:

  • Everything bought on credit would initially be “cheaper”, increasing demand for such things -> prices would increase to compensate (price inflation, coming from the money supply inflation). See this post regarding real estate to get an idea of the mechanism. It works for any asset class (stock etc.).
  • A derivative of that increased demand would be Companies putting in place extra supplies (new factories, new commodities extractions etc.), that would be done on credit (good, the Bank has now additional reserve it can create loans against…). Such supply increases take time to be put in place, so there will be inflation first. Of course, such an artificial increase in demand / supply (which is only due to an artificial rate decrease, not due to extra population or any real reason for which the rates could have decreased) would at some point reach a limit, after which we would have oversupply -> deflation, but this is still several years away.
  • Of course, should rates increase again at some point, such extra supply, which was not needed with higher rates, would be superfluous, we would again have oversupply and deflation.

This QE process can be expanded, the only limits are:

  • Bonds inventory left to be bought: in our example, after step (5), B1 has no bond left. It can nevertheless buy the bonds of C1, but once that is sold to the CB as well, the CB owns the totality of G1 debt and there cannot be more monetization of Government bonds. CB can still buy other things like stocks, Mortgages Backed Securities (MBS), corporate bonds… Of course the government can issue extra debt as well. History has shown that one can always count on a government to issue new debt. As such, this is not a real limitation.
  • Such increase in money supply, if left unchecked, will end up at some point in hyper-inflation (post on that coming at some points), as there is so much money in the system that people lose faith in its value. This is the end-game if the process continues.
  • Should the process of extra QE to fuel extra demand and extra supply stop at some point and should the oversupply work its way through the system there will be deflation and an economic depression. Indeed QE was put in place to avoid a depression: we already had reached a point of oversupply, over-indebtedness in the system before QE. QE only lowered the rate (allowing more indebtedness) and increased demand (allowing more oversupply). Should it stop, the crisis will need to resorb both the previous indebtedness/oversupply as well as the extra additional one we created through QE.

Exit from current situation comes either from a huge depression and deflation, or hyper-inflation (loss of trust in the currency) and… A huge depression….

Now let’s have a look at why it did not work as “well” as mentioned above to get increased economic activity and inflation (do not get me wrong, there is inflation, but not a lot of economic activity).

In our case (4) above, we saw that the bank had already reached its loan to reserve ratio limit (and was therefore reserve constrained, it was not able to make new loans, even if there were demand for it). QE in such situation would results in extra loans made to the economy (C1) shortly after the new reserves were released on B1 BS.

Now assume that the bank is not reserve constrained as C1 had no need for up to 500 of loan, but for example only needed 300 at that point, and did not anticipate extra needs in current economic situation.

Fractional reserve6

Doing the same QE in that condition would have the below results:

Fractional reserve7

B1 would have more excess reserves, but since it was not reserved constrained before, there will not be extra loans made to C1 (aside from the lower rates and their various impacts described above).

Now assume that C1 is already in a tough financial situation, because of low demand (hey, we were in a depression at the time of start of QE). Even if it wanted new money to expand its activity, B1 would refuse to lend due to too much risk of not being repaid.

Now you can also assume C1 is relatively fine financially, but C1 clients themselves (part of C1 in my mini economy) have too much debt already and cannot buy more C1 product, even if they were cheaper on credit (again, we were in a depression at the time, so prudent people would actually try to reduce their debt in such case, as they fear for their jobs etc.). Again, C1 will not expand in that condition, and not request extra loans.

In such case, the only impact of QE is to give additional “excess reserve” to B1.

What can B1 do with such reserves?

It cannot buy stock directly (I think, comments welcomed), but it can buy more government bonds. After that, it can exchange those government bonds with another Bank through a repo transaction (repurchase agreement):

B1 sell the bond to B2 (who can use its reserve to pay for it) for cash with the agreement that it would buy it back at the same price, minus interest, at a given date in the future). B1, with this new cash (which is not “reserve” anymore), can buy stocks. In short, B1 cannot force the real economy to use more loans to grow the economy, but it can buy other financials assets (bonds, stocks, MBS, etc…). That creates extra demand for such assets, which is exactly what we have seen since the start of QE:

  • No real growth during the past 7 years for most developped economies (to get real growth, you need to discount the nominal growth (the figure reported by the governments) with realistic inflations numbers).
  • But huge increases in asset prices:
    • Real estate prices are extremely high in all Europe, most of the US and most big cities all over the world.
    • Stocks markets are at all time high or decades high.
    • Rates on all types of bonds (high grade, high yield etc.) are at all-time low (meaning prices are at all time high).

The conclusions of QE is that people with assets before the start of QE have seen those asset increase in value, but on the contrary people without asset have not gotten any benefit from QE (and on the contrary, they got inflation, lackluster economy, no salary increases…).

One final point on the fact in current economic paradigm of fractional banking the money as credit needed to issue new loan is created ex-nihilo, but not the interest to service such a loan.

The money to pay interest has therefore to come for elsewhere: as long as there is more new Money as credit created in the system (more loan getting emitted than loan being repaid + the interest for such loans), one can always get money from elsewhere.

Now should the loan growth slows down below new loan emission + interest for existing loans (or even below just new loan creation, since interest rates are so low those day we can “almost” ignore them), it will get harder and harder to source the money to pay back the existing loans.

It will first come from the “real Capital” of all the economic actors (in our model, only the bank had Capital, so it means other actors will default on their loan, resulting in loss to the bank, losses that would deplete Capital). Such mechanism will make such Capital more and more valuable as time pass, a system where the money value increase with time is the same as a system where asset value decrease -> deflation.

Once all Capital has been depleted, the system will implode, as it would become impossible to sell anything even for 1 cent of currency, as there will be absolutely no “real money” in the system, only debt.

Of course, before this happen, one can count on the government to issue lot of new debt to pay back old one plus interest, for the “independent” CB to buy all that debt and for everybody else to understand that in such system the government and CB will never be able to stop, resulting in in total loss of faith in the currency of the system: hyper-inflation. At which point, after years of hardship (read “When Money Dies” from Adam Fergusson for a vivid account of that during the Weimar republic days in Germany), existing debt in the system will be totally worthless and the system will restart with a new currency.

At the end of the day, what mathematically cannot be repaid won’t be…

When thinking about this last part, a system that cannot survive unless it grows bigger and bigger all the time and draws in new entrants (new loans) to pay back old ones with interest looks really much like a Ponzi scheme. A loooooong lived one (one can say it started when the convertibility of “money as debt” to gold disappeared), but a Ponzi scheme nonetheless…

Such a systems work correctly (if not fairly) as long as it is possible for it to grow bigger and bigger. It indeed misses the limitations of the previous economic systems (capability of emitting new currency not always matching the economic needs for such new currency). I am not sure the technological progresses that were made in the last 50~ years would have been possible without such a system.

Now, on a quite finite Earth (we are not yet colonizing Mars or expanding across the galaxy, so for all intend and purpose, today’s system is finite), with close to 7 billion inhabitants and counting (how many more can we add sustainably), the system may be reaching its limits. There is no perfect system, previous one was not, next one will not be, since human is not perfect anyway. The only thing one can do it get out of the way when the current system crashes.

Please leave your comments, things you did not understand, things that need extra explanation, I am pretty much interested in your feedback.


One thought on “How the world works – Fractional reserve banking

Comments are closed.