naked capitalism

Anonymous says:

Shedlock’s analysis on borrowed reserves is wrong to the point of being embarrassing. The Baum article is quite correct. The Fed does control the level of system reserves. Use of the discount window and TAF affects the distribution of system reserves. The idea is to allow banks with eligible collateral to add to their reserves when they otherwise have difficulty in the interbank market. The use of the discount window and TAF adds to the gross level of system reserves, which the Fed can easily offset if desired at the aggregate level through its open market desk.

Anonymous says:

anon 11:40:

Shedlock doesn’t understand that reserves at the margin are supplied via system repos against billions in eligible, normal, treasury collateral. Reserves are not supplied only through the discount window or the TAF. In fact, these are the exceptions.

I tried to post the following on his site after he accused me of posting a ’stupid comment’. (He actually shut me out from posting further, after I responded in very mild kind to his outright insults of Baum and particularly Glassman, who in fact are quite correct on this subject. He doesn’t want any debate on his own very misguided views. You can read the earlier exchange on his site.):

“This is a stupid statement.”

There is an enormous amount of regular collateral (e.g. treasuries) available for normal system repos, which is the main channel for controlling reserve levels. The Fed doesn’t need to ‘control’ this collateral because it is not a binding constraint on the amount of total system reserves targeted and supplied by the Fed.

They can accept lower grades of collateral via the discount window or the TAF. If a borrower doesn’t have collateral that is acceptable under window rules, the borrower doesn’t get a discount loan and can’t qualify for TAF borrowing.

So collateral may be a constraint on the supply of reserves via either the window or TAF, but certainly not on regular system repos, which is the balancing mechanism for total reserves supplied.

The presence of negative non-borrowed reserves merely means they’ve satisfied more than the total system level of targeted reserves via the window and TAF, and therefore must withdraw some by reducing system repos. It doesn’t mean the banking system is short on reserves in total or that the Fed hasn’t controlled the level of total reserves it targets and supplies.

Really, this stuff is basic to understanding the mechanisms by which the Fed targets and supplies reserve levels. Not many people understand it, but Baum is certainly correct in her interpretation.

I’ll stand by my statement and Baum’s article.

Anonymous says:

anon 11:14

Yes – I’m referring to treasuries owned by the banks and used as collateral. No, the supply is not infinite.

Here’s my “101” explanation:

A bank’s balances sheet consists of assets, liabilities, and equity capital. Assets are on the left hand side; the other two are on the right.

The critical part of its reserves in question here is its own bank account with the Fed. It uses this account to clear transactions with other banks. This reserve account is an asset.

The bank can sell assets or raise new liabilities to increase its reserve account.

Some banks have had trouble doing this. So they can go the Fed discount window or the new TAF facility to get money from the Fed.

These types of transactions actually increase the level of total banking system reserves. The transacting bank has a new asset (increase in its reserve account) and a new liability (discount or TAF loan from the Fed).

A critical point that Shedlock doesn’t seem to understand is that the Fed controls total reserves in the system, or at least the bank account portion of these reserves. When it makes a window or TAF loan, aggregate reserves increase. While the Fed’s intention is to target increases at the individual bank level, it doesn’t necessarily want to do this at the system level. So what it provides to individual banks through the window or TAF it may well take away from the rest of the system via its day to day repo operations – e.g. renew fewer of these transactions, which reduces overall reserve levels. This has the effect of withdrawing reserves from liquid banks to offset what it provided to non-liquid banks through the window or TAF. This adjustment in turn shows up as a negative number under the “non-borrowed” reserve classification – negative meaning that it has withdrawn some of the reserves it provided via the window and TAF.

Bank reserves with the Fed, an asset, are completely distinct from equity capital, which is on the right hand side of the balance sheet.

The purpose of bank reserves with the Fed is to facilitate the daily interbank flow of funds.

The purpose of bank capital is to cover financial losses.

The two are linked to the degree that when depositors get worried about a bank’s capital position, they may withdraw funding support, thereby causing a liquidity ‘crunch’ for that bank, which may then have to go to the Fed for support. If the Fed loan is fully collateralized (which it is), this liquidity support, be it the window or TAF, is liquidity support only and does not constitute a capital ‘bail-out’. The Fed can provide this support so long as they view the bank as a going concern with enough collateral to back up the loan.

Again, a critical aspect is that the Fed controls the aggregate level of bank reserves in the form of banks’ accounts with the Fed. This is very important because it is the systemic mechanism by which the Fed controls the level of the fed funds rate. If the Fed ‘tightens’ reserves, banks scramble for funds generally and the funds rate will go up. If the Fed ‘eases’ reserves, banks have surplus funds in aggregate and the funds rate will drop. The Fed undertakes these adjustments all the time in order for the actual daily trading level of the funds rate to converge to the intended target level of the rate (currently 3 per cent).

Anonymous says:

anon 12:52:

You played it back more clearly than I explained it.

A few VERY round numbers:

The size of the Fed balance sheet approaches $ 1 trillion.

The normal size of bank reserve accounts at the Fed is probably less than $ 10 billion.

The size of US bank capital I think is in the area of $ 1.5 trillion.

The size of US bank balance sheets is many times that – probably in the area of $ 20 trillion.

(These are extremely round, because I don’t have the research time to look them up. But they’re close enough for illustration purposes.)

Discount window borrowing and TAF might be constrained at some point by the size of the Fed balance sheet, because the Fed needs to reverse some of its other activities in order to drain reserves as an offset (e.g. sell Treasuries outright or reverse dealer repos outright). The existing asset size of the Fed in effect becomes a limit of sorts because window loans and TAF become Fed assets. But the outstanding borrowed reserves are still small relative to the Fed’s balance sheet so this wouldn’t be a problem until these positions did get quite huge.

More importantly, the individual bank positions would become a constraint, because at some point they would indicate insolvency rather than liquidity problems. At that point, the Fed and others (e.g. treasury) would have to ensure that some sort of winding up of the bank was done, through merger, more permanent capital support, bankruptcy, etc. The Fed lines are only intended to be temporary. (Although TAF outstandings may well last a long time because the general situation is quite dire.) In any case, we’re a long way from borrowed reserves approaching the size of the Fed balance sheet.

Moreover, if window and TAF borrowings did increase, it would be likely that the Fed indeed would have to increase total system reserves in any event, in order to maintain the Fed funds rate where it targeted. This is because increasingly large TAF and window borrowings would reflect an increasingly distorted distribution of total system reserves. So the Fed might increase total reserves, which would allow it to expand its balance sheet.

As far as Shedlock is concerned, I don’t see any of this interpretation in what he wrote. Much of what he did wrong is downright incorrect. He was most disrespectful of Baum and Glassman. Glassman in particular was absolutely correct in what he wrote about the Fed’s ability to control total system reserves.