2. When looking at monetary expansion during Greenspan years, I think it important to also consider the foreign Central Banks, esp. the carry-trade countries, such as Japan, who should also shoulder much of the blame.

On the question of exactly how much the banks and other such multiplyed the money that Alan Greenspan put into the system (in his various bailouts over the years):

It is hard to know the exact number – because Alan Greenspan (as was long noted by Ron Paul) ordered that M3 numbers no longer be published.

It is hard not to conclude that this was part of a policy of deception – of increasing the money supply to a vast extent whilst pretending to be a moderate (much like Central Bankers in the United States and Britain in the late 1920′s – who paid lip service to the “gold standard” whilst playing every trick to increase credit money).

The late Ayn Rand was right to put that dinner plate in Alan Greenspan’s face.

For your illustrative bank, the effective required reserve is only 0.588 / 130 = 0.45%. Clearly, the money multiplier may quickly exceed 10 if this were typical of all banks. Indeed, plugging into the geometric sum formula I provided previously, we find that the multiple does indeed equal slightly less than 10 after only 10 rounds of loans at this lower reserve. However, it would require quite a few rounds to reach the theoretical maximum of 221. After 100 rounds, the multiple is still only about 80.5. After 1000 rounds, it is 218.7.

This is purely a mathematical exercise and is not intended to detract from your points.

My response was directed at the earlier question offered by Grant. (“How often is the multiplier actually 10?”) My point was that even the theoretical “10″ we throw around is based on very false assumptions, and that the official calculations do not make reasonable adjustments based on the fact that these assumptions are false.

100 + 100 * 0.9 + 100 * 0.9 ^ 2 + … + 100 * 0.9 ^ n

Expressing this as a geometric series reduces the amount to

100 * (1 – 0.9 ^ (n + 1)) / 0.1

Here, n equals the number of additional loans made with the initial $100. For ten loans, a total of $686 has been created – nowhere near the maximum muliplier of 10.

Clearly, as n approaches infinity, the multiplier approaches 1 / 0.1 = 10. For a 10% reserve, the multiple approaches 9 with 20 additional loans and is virtually to 10 after 100.

Legally, the Fed has different reserve requirements for different deposits.

“Transactions” deposits (which are mostly checking accounts) have a graduated system, not unlike the income tax structure. As of January next year,

The first $10.3 million has a reserve requirement of 0%. The next $34.1 million has a reserve requirement of 3%, and everything above the first $44.4 million has a 10% reserve requirement.

Nontransaction deposits – which includes all time deposits (think “Certificates of Deposit”) and most savings accounts (as long as the number of withdrawals is limited to some number each month… I think that’s 6 or less, with no more than 3 by check, but I’m not sure… it’s considered a “savings account”). These have a reserve requirement of $0.

So, say that you have a small bank with $100 million in savings accounts and $30 million in checking accounts. We can calculate the reserve requirements as:

$10.4m x 0% + $19.6m x 3% + $100m x 0% = .588m

So, on its $130 million of deposits, the bank is only required to hold $588 thousand dollars, giving a “true money multiplier” (deposits/reserves) of 221 for that particular bank, if they only hold required reserves.

Hope the example helps clarify things a bit.

Also, I tend to agree with you that we should “get the government out of banking and have a 100% reserve system”. The trick is “How?” I think that Rothbard has an answer: if we make banking legally equivalent to warehousing, then holding fractional reserves is fraud, and prosecutable. Currently, that’s not the legal status of banking. This switch can be done on a free market.

Also, I’d suggest that free banking would result in banks holding more reserves than they currently do. The current system has the FDIC, Fed, and high probability of bailouts all artificially reducing the cost of holding few reserves. So, even if we did have the government get out of banking and free banking resulted, we’d probably see more reserves than we do now. Naturally, though, this statement is somewhat speculative.

Christopher Richard Wade Dettling

Pete, you cannot get the government out of banking and simultaneously end fractional reserves. That sort of banking has existed absent government intervention; to end it you’d probably need to outlaw it. I believe the era of free banking in Scotland, banks tended to have reserves of around 2-3%, but don’t quote me on that.

Assuming a 10% required reserve ratio the money multiplier cannot go above 10. I don’t understand why you say reserves can be 0% and 3% at times. Please explain in further detail about non transaction deposits.

Greg,

you said “What I really want to hear from someone from this school of thought, what is the solution to get us out of this present condition? And a realistic forecast of the effects of any solution put on the table.”

First we can start by getting the Government out of the mix, and end fractional reserve banking. This stops the manipulation of credit. When we first start this it will cause secondary effects that seem negative, but that is the market adjusting to real terms. Once we get back to a real economy we will see true prosperity.

i think all this agonizing over the reserve multiplier is wasted energy. sweeps have made this calculation futile. see hatch article:
http://mises.org/journals/scholar/hatch.pdf

as for grant’s question as to why a particular sector becomes “hot”, my view is that initially plain old fundamentals kick in, and money flows to whichever sector was neglected in the previous bubble. austrian insights only predict the creation of the bubble, not where the bubble forms. that’s up to the market.

Some points about the multiplier:

First, “the multiplier is 10″ is a theoretical maximum based on obviously false assumptions. First, we assume that no currency is held by the public (so that the monetary base is only reserves). Second, we assume that all money that can be lent is being lent. Third (and this is rarely mentioned), we’re assuming that all deposits actually have a reserve ratio of 10%. These are all false.

In reality, a lot of the monetary base is held in currency nowadays. actually, as of December 2007, something like 95% was currency. Today, about 60% is. The larger this number is, the lower the money multiplier will look. (Though we can correct for this by just removing currency from both the money supply and monetary base, that would give us deposits/reserves again.) Say that there is $1 of monetary base, $0.50 in currency, $0.50 in reserves with maximum legal lending (10x reserves). Then we’d have a calculated multiplier of 5.5 instead of 10.

Also, in reality, as you note, not all money that can be lent legally is actually being lent. Banks generally hold some excess reserves… but not much, most of the time. Just looking over the past 60 years really quickly, actual reserves tend to be about 5-10% higher than required reserves. So, though the “maximum multiplier” may be 10, the “actual” one (if we only think about this impact) is closer to 9.

Both of these tend to push the money multiplier down below the theoretical level. They also allow the multiplier to fluctuate because of the behavior of banks and the public.

However, I would argue that the more important (false) assumption is the last one. In reality, 10% is the HIGHEST reserve requirement in a graduated, compartmentalized system. The first $X of deposits have a reserve requirement of 0%. The next $Y of deposits have a reserve requirement of 3%. It’s only dollars after $X+Y that have a 10% reserve requirement. Now, since banks tend to be very big, the 10% reserve requirement isn’t too inaccurate… for transactions deposits. But, there’s a very large set of “nontransactions deposits” which have a different reserve requirement… specifically, 0%. Right now, there are about $700 Billion in transactions deposits and over $6.5 Trillion in nontransactions deposits. The nontransactions deposits can be lent out, in their entirety, ad infinitim. Naturally, this would tend to increase the money multiplier above 10… perhaps far above 10, if it is appropriate to include these nontransactions deposits. Personally, I’d argue (as Rothbard did) that savings accounts should be included but time deposits should not. The savings account portion is about $4.4 Trillion.

As an exercise, I calculated the following: (True [AKA "Austrian"] Money Supply – Currency)/(Monetary Base – Currency). So, all the problems caused by having so much currency floating around will be cancelled out, and all that remains are the impacts of holding excess reserves and the fact that not all accounts have the same reserve requirements. For August, the number I got for this “true” money multiplier was 105. In fact, it had been over 100 since September of 2007. From 1981 to 1995, this number was right around 30-45, so 100 was HUGE. Right now? It’s just under 8. The Fed has injected reserves like mad, and the money isn’t getting loaned out (yet).

The lesson I’d pull from this: 10 isn’t too bad of a number for how much an addition $1 of reserves can create. If anything, it seems to be a conservative estimate.

Now, my interpretation of the data is somewhat different than Dr. Murphy’s… mostly because when I do my calculations, I see that reserves increased at the same time as the “true money multiplier” was increasing. So, part of the bubble was caused by Fed injections, but part a very large part was caused by public choices between transaction/nontransation accounts. (Transactions accounts barely moved which nontransactions accounts exploded.) This choice allowed for money to multiply much more than before. The big mystery in my mind is what changed in 1995 that caused the true money multiplier to start increasing then, as it increased very steadily from about 33 to 105 over the course of 10 years. That’s something I’m thinking about, but have no good answers for yet.

Another alternative calculation if you don’t like including savings accounts: To find the “actual reserve multiplier” calculate (M1-Currency)/Total Reserves. Making that calculation, the result is between 13 and 16 from 1981 until the past few months. So, 10 is still a conservative number for how much an additional dollar of reserves will multiply into. Really, it seems that the official multiplier is very heavily impacted by the fact that a lot of money is being held as currency rather than deposits. From my point of view, that’s good, as it minimizes credit expansion. On the other hand, with so many deposits in accounts with a reserve requirement of 0%, that “minimization” does almost nothing, as the multiplier on these deposits is theoretically infinite.

Anyway, just another perspective on that problem.

And, just because something sounds simple, doesn’t mean it’s wrong. Answer this question, where did all the money come from to go into the stock or housing bubbles. It either has to come from somewhere else (i.e. take from there), or more money is created.

If it came from somewhere else, there should have been areas that were slowing down as much as the bubbles were growing. Since there were not also great busts during the booms, this indicates that it was new money pumping things up.

After all, each time the prices increase (e.g. stocks, or houses) even more money is required to keep it going at the higher prices. This is why all it takes is a slight lowering of the speed of the money injections to cause a bust. You don’t need to stop inflating altogether, just slow it down some and those at the margins (the ones who couldn’t have stayed in business w/o the easy money) begin to fail. AKA, the bust.

We could debate the causes of the boom, which really was a boom in a few limited markets, and never hit on all the causes. What I really want to hear from someone from this school of thought, what is the solution to get us out of this present condition? And a realistic forecast of the effects of any solution put on the table.

My solution is simple. Every employer pays a matching FICA tax for every employee. When an employee buys a house, they submit the purchase and loan information to their employer and the employer sends their matching FICA tax to the bank instead of the government and this payment continues for 3 years.

While the government is going to collect less from employers, they will see their overall tax collection increase as the employee has less interest deductions, the people that make money on the real estate transaction and in the case of new construction, the amount of money that is made by all those that build the house. All those will pay income taxes as well as FICA taxes which will more than replace the money the government lost.

Hayek shows in “Monetary Theory and the Trade Cycle” (available in pdf on Mises.org) that banks will temporarily dip below the 10% requirement during a boom, or at the start, in order to keep from having to raise interest rates to attract new deposits. In general, banks stay as close to the 10% minimum as possible, otherwise extra reserves are sitting in the vault not earning a return. During the downturn they may have more reserves because no one is asking for loans.

Grant: “I really don’t think any Austrians claim that ABCT is the only source of business cycles? ”

No. They mention war, drought, natural disaster and other things that people have understood as causing financial cycles for millenia. The ABCT seeks to explain the cycles that those phenomena don’t explain. In “Monetary Theory” Hayek includes a tech breakthrough as one reason that banks will begin to expand credit.

Here’s why credit expansion is necessary for a business cycle, but not necessarily a bubble. Assume a fixed supply of money. If prices start to rise in one commodity, people will have to reduce their consumption of something else in order to have the money to pay for the more expensive item. So you could have a bubble in a particular product, such as in the tulip mania in the Dutch Republic, without having a business cycle. But to have a general rise in business activity as we see in a typical business cycle, the money supply must increase.

In the current cycle, we saw a bubble in housing. If the money supply had remained fixed, then some other sector would have had a recession that mirrored the housing bubble. That didn’t happen. The lack of a recession in another sector indicates that the money supply was increasing enough to fund the housing bubble as well as keep all other sectors firing at the same time, such as the energy sector where prices increased as well.

Hayek goes through all of the explanations of business cycles and shoots them down one by one in “Monetary Theory and Trade Cycles”. There are two chapters that look at other theories, “Non-Monetary Theories of the Trade Cycle” and “Monetary Theories of the Trade Cycle.” As old as the book is, I have seen anyone come up with a theory not included in the book, which shows Hayek’s brilliance. He responded to economists who didn’t even write until after his death.

http://www.bloomberg.com/apps/news?pid=20601039&refer=home&sid=a.YJmSfnHD9o

The bloomberg article discusses a $3T bailout for homeowners, via fixed 4.5% 30 yr notes, which looks like it would effectively socialize the entire mortgage industry (more so than it currently is if that’s possible).

I’m curious how it would impact some of the conclusions of this article.

Regards,
gfh

I really don’t think any Austrians claim that ABCT is the only source of business cycles? I’m not sure I’ve heard that rigorously defended anywhere? Logically speak, proving that A causes B does not disprove a theory that states X, Y or Z can also cause B.

No. It only has to go to ten loans to reach the max increase in money.

I think you’re incorrect here. I can’t draw out formulas here, but look at the following wikipedia page:
http://en.wikipedia.org/wiki/Fractional-reserve_banking#Money_multiplier

The theoretical maximum multiplier of 10 is an asymptote. I should also note that not all banks operate at the limit of the reserve requirement; 10% is the minimum.

wren, no it doesn’t discredit the theory to me. It might lead me to believe other causes are at work besides credit expansion, though I’d really like to read that QJAE paper when Mises.org releases the latest volume.