Quote:

Here's an interesting excerpt from the book Investing Public Funds by Girard Miller about borrowed reserves.

"Another useful indicator of the Federal Reserve's relative monetary policies can be found weekly in the Federal Reserve data. A key statistic is the net free reserves or net borrowed reserves line item. This statistic measures the degree to which depository institutions have found it necessary to obtain funds in the Fed Funds market and through the Fed discount window in order to obtain required reserves.

During periods of central bank credit-tightening operations, the depository sector might find it necessary to borrow funds to meet reserve requirements. This practice results in net borrowed reserves, which shows as a negative number. Conversely, if ample funds are available through the banking system to meet reserve requirements, banks can become net lenders of reserves through the Fed Funds markets"

Given that the Fed is not in a credit tightening mode, we must look for a better explanation. Here it is: Banks in aggregate have now burnt through all of their capital and are forced to borrow reserves from the Fed in order to keep lending.

Total Reserves for two weeks ending January 16th are $39.988 billion. Inquiring minds are no doubt wondering where $40 Billion came from. It's a good question. The answer is the Term Auction Facility. You can see that figure in Table 1 of the H3 release (not shown).

Were it not for the Term Auction Facility, banks would have had to raise $40 billion in capital by selling assets or some other means. We will look at "other means" in just a moment.

For now, the Fed is not disclosing who is borrowing under the Term Auction Facility, probably out of fear that people just might find out what banks are capital impaired and by how much.

Quote:

The reserves have been spent because banks have expenses. They haven't had any significant income from other banks since August. CP is dead (banks lending to each other) so the banks turn to TAF to avoid closing their doors, or telling customers 'We're out of cash this month.'

As the Fed sends in emergency money, the value of the dollar drops.....

Unless their infusions are (a) revolving, so they don't create any net increase in money, or (b) the dollar losses from write-downs more-or-less match the money created by the Fed.

If the Fed is playing this game just right, they will increase the money supply to match the write-down loses that are made public every Friday.

If they blow it, we have either hyper inflation or a deflation spiral.

Eoghan on Calculated Risk 2/04/08 wrote..

I’ve spent quite a bit of time looking at this and trying to figure out what’s going on, so maybe I’ll be able to shed some light. I don’t think the situation is as dramatic as one might think on first glance – but there does seem to be something slightly odd going on.

The first thing to focus on is the definition of reserves. As you rightly say, the reserve requirement is the proportion of what you deposit that is ‘held back’ in reserve – ie held in cash by the bank in a vault somewhere. The other thing that counts as a reserve is deposits held by a (commercial) bank at the Federal Reserve. As you rightly point out – if you accept the concept of factional banking at all, then you should be comfortable with this.

In % terms – I believe the reserve requirement is 10%, but there are a number of different exceptions for different types of account, and of course physical dollar cash has no reserve associated with it, so it’s not surprising that reserves as % of the monetary base is less than 10%.

The other thing about reserves is that they are not the same thing as capital. Reserves are held to provide liquidity – they don’t guarantee solvency. They’re held to ensure that, in the normal course of business, the bank has enough immediately liquid assets (cash) to pay out any depositors who show up looking for their money back. Of course, given that the reserve held is a small % of deposits, there is no way the bank would be able to pay back everyone at once. A bank can have adequate liquidity but be insolvent (the Japanese banks in the 1990s). It is also possible for a bank to have inadequate liquidity but be solvent (Northern Rock in the UK, last year). So reserves are not a good measure of how strongly capitalized a bank is. A very strongly capitalized (safe) bank and a very weak (likely to go bust) bank will hold the same % of deposits as reserves.

With all this in mind, let’s take a look at the Federal Reserve data you linked to (http://www.federalreserve.gov/releases/h.... Current/). Table 1 shows non-borrowed reserves declining to -$8.751 billion on Jan 30th as you mentioned. However, look at the definition of ‘non-borrowed reserves given in the footnote. This is – “total reserves less…. Borrowings of depository institutions from the Federal Reserve”. Banks don’t tend to hold more reserves than they have to. Hence, as you’ll see in the table, total reserves and required reserves track each other closely – in other words excess reserves (column 4) are small. So non-borrowed reserves is roughly = required reserves + about $1.5 billion of excess – “Borrowings of depository institutions from the Federal Reserve”. It’s clear from the table that Requred Reserves have not changed much recently – ie US banks have generally the same amount of deposits as they had previously and are hence required to hold the same amount of reserves against them. What has changed is ‘Borrowings of depository institutions from the Federal Reserve”. This has increased by $50 billion due to the introduction of the TAF, where banks are allowed to borrow term money from the Fed. The TAF has increased from 0 to $50 billion in the space of a month. This has, arithmetically, driven ‘non-borrowed reserves’, defined as above, below zero.

Does this matter? No. Firstly, reserves are not capital. So this number tells us nothing about the capital adequacy of the US banking sector. By the way, if “Oliver Q.” (upthread) has any evidence whatsoever for his assertion that “The reserves have been spent because banks have expenses” I for one would love to see it. Secondly the definition of non-borrowed reserves is a bit weird – IMHO it’s comparing apples with pears – reserves being an ‘administrative’ quantity rather than a ‘real money’ quantity, whereas as the amount of money lent out by the Fed is a real number. So one minus the other doesn’t make a lot of sense.

The final reason why this doesn’t matter is – look at Table 2 – on the right hand side – “Vault cash (5)”. This column shows that the amount of cold hard cash held by US banks in their vaults has not decreased at all – they still hold $54 billion of cash and this number has actually increased over the last few weeks. This shows that the non-borrowed reserves numbers really is a bit of a red herring.

What I still think is a bit strange is that the TAF is now $50 billion - hence one would expect reserve balances with the Federal Reserve banks to be above $50 billion – the commercial banks submit their long-term paper into the TAF and are given cash balances in the Federal Reserve system. But the number is only $9.5 billion. While this is a bit strange the definition of column (4) is obscure enough that the $50 billion balancing item could be held somewhere else.

In short, while I think the numbers take a bit of getting your head around, they are not direct evidence of any fundamental problem in the US banking sector that we don’t already know about. Most importantly, reserves are not capital, and the difference between the two has to be held in mind.

Quote:

I’ve spent quite a bit of time looking at this and trying to figure out what’s going on, so maybe I’ll be able to shed some light. I don’t think the situation is as dramatic as one might think on first glance – but there does seem to be something slightly odd going on.

The first thing to focus on is the definition of reserves. As you rightly say, the reserve requirement is the proportion of what you deposit that is ‘held back’ in reserve – ie held in cash by the bank in a vault somewhere. The other thing that counts as a reserve is deposits held by a (commercial) bank at the Federal Reserve. As you rightly point out – if you accept the concept of factional banking at all, then you should be comfortable with this.

In % terms – I believe the reserve requirement is 10%, but there are a number of different exceptions for different types of account, and of course physical dollar cash has no reserve associated with it, so it’s not surprising that reserves as % of the monetary base is less than 10%.

The other thing about reserves is that they are not the same thing as capital. Reserves are held to provide liquidity – they don’t guarantee solvency. They’re held to ensure that, in the normal course of business, the bank has enough immediately liquid assets (cash) to pay out any depositors who show up looking for their money back. Of course, given that the reserve held is a small % of deposits, there is no way the bank would be able to pay back everyone at once. A bank can have adequate liquidity but be insolvent (the Japanese banks in the 1990s). It is also possible for a bank to have inadequate liquidity but be solvent (Northern Rock in the UK, last year). So reserves are not a good measure of how strongly capitalized a bank is. A very strongly capitalized (safe) bank and a very weak (likely to go bust) bank will hold the same % of deposits as reserves.

With all this in mind, let’s take a look at the Federal Reserve data you linked to (http://www.federalreserve.gov/releases/h.... Current/). Table 1 shows non-borrowed reserves declining to -$8.751 billion on Jan 30th as you mentioned. However, look at the definition of ‘non-borrowed reserves given in the footnote. This is – “total reserves less…. Borrowings of depository institutions from the Federal Reserve”. Banks don’t tend to hold more reserves than they have to. Hence, as you’ll see in the table, total reserves and required reserves track each other closely – in other words excess reserves (column 4) are small. So non-borrowed reserves is roughly = required reserves + about $1.5 billion of excess – “Borrowings of depository institutions from the Federal Reserve”. It’s clear from the table that Requred Reserves have not changed much recently – ie US banks have generally the same amount of deposits as they had previously and are hence required to hold the same amount of reserves against them. What has changed is ‘Borrowings of depository institutions from the Federal Reserve”. This has increased by $50 billion due to the introduction of the TAF, where banks are allowed to borrow term money from the Fed. The TAF has increased from 0 to $50 billion in the space of a month. This has, arithmetically, driven ‘non-borrowed reserves’, defined as above, below zero.

Does this matter? No. Firstly, reserves are not capital. So this number tells us nothing about the capital adequacy of the US banking sector. By the way, if “Oliver Q.” (upthread) has any evidence whatsoever for his assertion that “The reserves have been spent because banks have expenses” I for one would love to see it. Secondly the definition of non-borrowed reserves is a bit weird – IMHO it’s comparing apples with pears – reserves being an ‘administrative’ quantity rather than a ‘real money’ quantity, whereas as the amount of money lent out by the Fed is a real number. So one minus the other doesn’t make a lot of sense.

The final reason why this doesn’t matter is – look at Table 2 – on the right hand side – “Vault cash (5)”. This column shows that the amount of cold hard cash held by US banks in their vaults has not decreased at all – they still hold $54 billion of cash and this number has actually increased over the last few weeks. This shows that the non-borrowed reserves numbers really is a bit of a red herring.

What I still think is a bit strange is that the TAF is now $50 billion - hence one would expect reserve balances with the Federal Reserve banks to be above $50 billion – the commercial banks submit their long-term paper into the TAF and are given cash balances in the Federal Reserve system. But the number is only $9.5 billion. While this is a bit strange the definition of column (4) is obscure enough that the $50 billion balancing item could be held somewhere else.

In short, while I think the numbers take a bit of getting your head around, they are not direct evidence of any fundamental problem in the US banking sector that we don’t already know about. Most importantly, reserves are not capital, and the difference between the two has to be held in mind.

Last modified: 2008-02-04 10:38:12 by yal