Friday, August 2, 2013

Banking Example #4.1: The Two Kinds of Quantitative Easing

This post is a variation on Example #4. I've eliminated some parts and added others in an attempt to simplify, clarify and explain the two kinds of quantitative easing (QE) and the difference between them.

Setup: one Treasury Dept. (Tsy), one central bank (CB), one commercial bank (Bank A), and one person x. No reserve requirements. We start off with the Tsy having sold a $100 Tsy bond and spent the proceeds into the private economy on goods and services. This results in a net liability of $100 at the Tsy and a net asset of $100 (the Tsy bond) in the private sector, but not necessarily any bank deposits in the private sector. See Example #8 to see how this can be.

In Case 1, the Tsy bond is owned by Bank A. In Case 2, it's owned by Person x. In both cases the bond owner sells it to the CB during the QE operation. The Tsy balance sheet looks like this both prior too and after the QE operation in both cases:

Tsy
Assets Liabilities
$0 $100 Tsy bond
Negative Equity Equity
$100 ------------------


Case 1: First kind of QE: The bank sells its Tsy bond to the CB. 

Initial balance sheets prior to the QE operation (Person x not involved: assume his balance sheet is empty):

CB
Assets Liabilities
$0 $0

Bank A
Assets Liabilities
$100 Tsy bond $0
Negative Equity Equity
--------------------- $100


Balance sheets after the QE operation:

CB
Assets Liabilities
$100 Tsy bond $100 deposit for Bank A (reserves)

Bank A
Assets Liabilities
$100 reserves $0
Negative Equity Equity
--------------------- $100


Case 2: Second kind of QE: The non-bank sells its Tsy bond to the CB.

Initial balance sheets prior to the QE operation (both Bank A and Person x are involved):

CB
Assets Liabilities
$0 $0

Bank A
Assets Liabilities
$0 $0

Person x
Assets Liabilities
$100 Tsy bond $0
Negative Equity Equity
--------------------- $100


Balance sheets after the QE operation:

CB
Assets Liabilities
$100 Tsy bond $100 deposit for Bank A (reserves)

Bank A
Assets Liabilities
$100 reserves $100 deposit for x

Person x
Assets Liabilities
$100 deposit at A $0
Negative Equity Equity
--------------------- $100


Notice that nobody's equity changed due to QE in either case. For a fuller discussion of this fact see the discussion at the end of Example 4. Also notice that in both cases the CB's balance sheet changed in exactly the same way.

Final observation: I think the second kind of QE (case 2) is more prevalent. Also, I'm not showing some details in this case (the exact details of how the bank acts as an agent of the CB to purchase bonds on the open market), but these details are relatively unimportant.

14 comments:

  1. Hi Tom, I am just coming over from your comment to my post. Looking through your accounting here I think there is one important problem that may also explain part of your response to me. If you assume there is no money to begin with you first need to create some where you need the CB for:
    1. Treasury receives deposit from CB for Tsy:
    CB: A: Tsy L: $100 and Tsy: A: $100 L: Tsy
    2. Treasury spends money to person x who works with Bank A:
    CB: same Tsy: A: ($100)(NE) L: Tsy Bank A: A: $100 L: $100 X: A: $100 L: $100(E)
    3. Bank A converts $100 into interest-bearing Tsy
    CB: A:$100 L:$100 Tsy: same Bank A: A: Tsy L:$100 X: same

    Now Bank A would be illiquid. Hence, QE will serve to provide liquidity again by taking financial assets from the bank's balance sheet. However, the very important distinction to your post is that the $100 or the Tsy are never the equity of the bank. They are the collateral to their customer's deposit and hence cannot be spend on any non-financial asset by the bank. In essence, QE will only serve to swap financial assets around but will not lead to an increase in spending. Odie

    ReplyDelete
    Replies
    1. Hi Odie

      You write:

      "3. Bank A converts $100 into interest-bearing Tsy
      CB: A:$100 L:$100 Tsy: same Bank A: A: Tsy L:$100 X: same"

      quibble: CB's balance sheet should be A: $0 L: $0

      also.. in general I'd follow by Ex #8 to get to where this Ex starts off (i.e. the CB doesn't buy debt direct from Tsy)... but that doesn't change your point.

      "QE will only serve to swap financial assets around but will not lead to an increase in spending."

      I agree that QE is just an asset swap.

      I view "equity" as just an abstraction (dollar value of assets in excess of liabilities), so it could never be composed of Tsy debt or $ or any other asset. But again, another quibble... I agree with you that the bank's equity remains 0 here. In fact QE does not change anybody's equity. I think that is an excellent counterargument! I was just trying to point out Sumner's argument as I understood it. But I have the same reservations! Now at some point QE becomes a fiscal operation: say the CB buys bags of air... essentially they just pump $ into the private sector like deficit spending by Tsy. In the case of normal QE the CB marginally increases equity in the pvt sector by paying marginally more for the assets they purchase.

      Delete
    2. Hi Tom,

      Equity is the most important position on a bank's balance sheet; just ask its shareholders. ;-) That is where the earnings are going and of which any losses need to be paid off. From that balance sheet position the bank pays it dividends. It is also the place where changes in asset value will have an impact. With the drop off rates the value of old financial assets (with higher interest rates) rose (value of a bond is inverse to its interest rate). That additional value in assets by keeping liabilities the same led to an increase in the total assets of a bank. Hence, the net equity of banks rose when QE lowered rates. That is the reason they have now "higher capital". Increasing rates will have the opposite effect.

      Btw. I am also dismayed by the fact that everyone seems to care about the money but no one about why we do that actually because we all forget the equity position. In your/my example the treasury bought something with that money like a bridge. Hence the treasury does not have a negative equity but a bridge worth $100 on its balance sheet. X carries the liability, the labor it used to build that bridge. At the end, that is all we should care about since all financial assets and liabilities add up to zero. Sorry for the long post, but I wrote this one a while ago for some other site which may help you to illustrate my point:



      Delete
    3. "You may ask: What is the point of money summoned out of thin air and vanishing again once someone pays back the loan? Maybe I can show at a little exemplary economy how our monetary system benefits us:

      Let’s meet Mr. Grocer. As his name implies, he wants to open a grocery store and needs some shelves and tables to do so. He goes to his bank and asks for a small business loan of $1000. The bank opens its books and writes under assets:” Mr. Grocer owes us $1000.” On the other side they write under liabilities:” We owe Mr. Grocer the $1000 which are in his checking account.” Mr. Grocer then goes to Mr. Carpenter and offers him $1000 when he installs the furniture in his store. Once Mr. Carpenter is finished, he receives a check of $1000 which he deposits at his bank. (Could be the same or a different one, does not matter.) Since Mr. Carpenter wants to marry soon he goes to Mr. Taylor and orders a wedding gown for his wife and a suit for himself. Mr. Taylor finishes those, receives $1000 for it, which he then uses to buy groceries for him and his family for the next months. Mr. Grocer accumulates that income in his checking account and finally pays back his loan. The bank will then wipe off the loan from its books as well as the deposit that was in Mr. Grocers account.

      When we look at the monetary balances we will see that in the beginning there was no money and at the end there is no money either. So what changed? When we look at the non-monetary assets the following happened:

      Mr. Grocer: + store furniture
      Mr. Carpenter: + wedding dresses
      Mr. Taylor: + groceries for several month

      For all three participants in our economy the balance of their non-monetary assets increased due to the debt-money created by Mr. Grocer. Would any of that have happened when the bank would have denied the loan? Most likely not. Mr. Grocer could have gone to Mr. Carpenter and asked to put the furniture in for future groceries as in a moneyless barter economy. That may work with groceries but what if he wanted to sell bikes and Mr. Carpenter had no interest in working for getting a bike later?

      The simplified mechanism I described here is the way how debt based money creates value for our society. It essentially works the same way with thousands of consumers/businesses doing transactions. Money is used to create demand and to enable the growth of the material and immaterial asset base of our society. Just as an example: How many houses would we have built if we did not have access to mortgages? However, if consumers and businesses are worried about their own personal balance sheet and refuse to take on more debt and instead pay back their old ones that economic activity stops. Then the government should step in by taking on more debt to make sure we keep our society moving forward because the non-monetary assets are the things we really need to live our lives. Nevertheless, I hope it also shows that given the fleeting nature of money the result we really should look for are the material and non-material assets we acquire. Do the things that we buy with that money really benefit us in the long run? Something worth thinking about."

      Odie

      Delete
    4. Odie, haven't read it all yet... but it sounds like it's similar to this from Cullen Roche:

      http://pragcap.com/yes-government-deficits-equal-private-surpluses

      BTW, regarding an explanation of QE, etc. I would recommend JP Koning on this. He writes about it in terms of the "convenience yield" and so far sounds the most sensible of anyone in the "the central bank CAN do it all" crowd (although I'm not sure he actually believes that). Also try Dave Beckworth (who I'm not as familiar with, but I know Cullen has a lot of respect for him). Anyway, I would suggest going to JP Koning's site... I think he's clearer than Sumner, Rowe or Glasner (although they all have their moments). I'm not as familiar with Christensen, Nunes, or Woolsey.

      Koning I think has the advantage of not really being in the group... so he's free to point out how similar Krugman's position is! ... and also poke some holes in MM myths. Of course Cullen does that too, but Cullen is looking at things from the PKE/MR perspective... which is great, but I'm ALWAYS on the lookout for overlap. I'm not saying Koning and Cullen have the same message... I would favor the MR crowd... but for a different view of things, so far Koning is my favorite. But I'm pretty fickle... my "favorite" changes on a daily basis. In that light Beckworth might be best for an alternative view... just going on Cullen's recommendation. Sometimes Sumner leaves me very cold. (I don't agree completely w/ Rowe BTW, but he does have some very interesting and creative explanations and I often refer to them... just don't necessarily trust him on Reflux... he's very much in the minority there, even among the MMists).

      But I can tell you that seeing things from JP's "convenience yield" and "non-pecuniary yield" (same thing) view is superior (for me) than either Rowe or Sumner's stories. And here's where I maybe part ways with Koning a bit: it's clear he sees the paradox inherent in the CB only view: credibly suppress rates in the future or permanently increase the base (same thing)... but do it too far and you get in a trap! Look at Sumner's case 5b (in Sumner's HPE Explained post)... and Jared's question about it and Sumner's response! Both JP and Sumner think we're closer to 5c than 5b, but I'm not so sure! (at least both acknowledge it's a continuum).

      Look at the comments there from Mike Sproul too... very interesting! I'd say JP's last five or six articles make a nice story.. the current one being the Hot Potato, hot, cold, hot one... with the flaming potato album cover! Basically he tells you precisely why the potato's no longer hot. His "Leverage" post tells four ways to reheat it.... not all of which seem that credible/feasible to me. What he leaves off the list is sensible fiscal combined w/ monetary... (although I think the "Sproulian Lever" is precisely such a thing... minus the "sensible" part). I think Cullen's name for that is "The Fed buys bags of dirt." Lol!

      Delete
  2. Hi Tom:

    Following up on behalf of a post from the gyroscopicinvesting site...

    A question often asked (I know *I've* asked it) is "where does the CB gets the funds to buy the bond?" And the answer is that they conjure it up out of thin air. From that perspective, would it be reasonable to simply say that the CB loans themselves the funds? That is, in the imaginary instant just before the CB purchases the bond, they get a loan from themselves, giving them $100 reserves as an asset, along with $100 deposit at the CB as a liability. They then use the $100 reserves to purchase the bond from Bank A, leaving the balance sheets as you show in example 1 above.

    JGF / TPG

    ReplyDelete
    Replies
    1. JGF / TPG,

      Offhand, I don't see anything wrong with thinking of it that way. But it seems to have extra complication to me. I prefer to think of it like this:

      Every financial asset has two sides: two ways to look at it: 1.) from the perspective of the creditor and 2.) from the perspective of the debtor. This applies to loans, cash, Fed deposits, bonds, etc. My Example 1.2 colors the balance sheets to show these two views for each financial entity.

      Also, all financial entities are essentially IOUs: the debtor writes the IOU and gives it away, and the creditor receives the IOU. The debtor is stuck being the debtor until his IOU is returned to him, but the creditor can pass the IOU on to a third party, thus transferring creditor status to that party.

      So when the Fed credits an entity with a Fed deposit or cash, they've essentially become a debtor to that entity. The Fed deposit (or cash) is the transferable IOUs capable of being passed onto other parties.

      Imagine that the Fed could buy bonds directly from Tsy (they can't legally, but imagine they could). Then if they did this, it would essentially be a case of two entities (the Fed and the Tsy) swapping IOUs with each other: the Fed deposit is the Fed's IOU and the Tsy bond is the Tsy's IOU. There are two financial assets (the Fed deposit and the T-bond) and thus there are four ways of looking at them altogether. The Fed & Tsy simultaneously become creditor and debtor to each other... until the bond matures, at which point they return each other's IOUs and the financial entities (deposit & bond) cease to exist, along with the creditor and debtor statuses of the Fed & Tsy.

      The same thing happens when an individual takes out a loan from a bank: The bank credits the borrower with a bank deposit (the bank hands them an IOU), and the individual signs a loan document (the individual hands the bank their IOU). Again, each of these two financial entities (loan & deposit) has a creditor and a debtor associated with it, and thus there are a total of four views of these two items on the balance sheets.

      I think that's a more clear and consistent way of looking at it. It also explains why all financial instruments (deposits, cash, bonds, etc) add up to zero equity if all balance sheets are consolidated together: the single consolidated balance sheet will have assets exactly cancelling liabilities.

      The one exception is coins (in the US). Coins don't appear as liabilities on anyone's balance sheet. But they are still an off balance sheet obligation of the Tsy (i.e. the Tsy is committed to accepting coins back again as payment).

      Delete
    2. Re: coins: thus if all balance sheets were consolidated together, then the total equity would match the total dollar value of all the coins: every other financial asset would have an exactly offsetting financial liability.

      Of course I'm not counting land, gold, cars, etc as financial assets here. Those do not have offsetting liabilities.

      Delete
  3. one commercial bank (Bank A), and one person x. No reserve requirements. We start off with the Tsy having sold a $100 Tsy bond and spent the proceeds into the private economy on goods and services. GetSomeDosh

    ReplyDelete
  4. This results in a net liability of $100 at the Tsy and a net asset of $100 (the Tsy bond) in the private sector, online loan from direct lenders

    ReplyDelete
  5. Mentioned above...most QE purchases are from private citizen x, rather than from Bank A. Banks don't have billions-trillions of extra treasuries on their books at any given random point in time. At the margin, they clear trades for customers (and/or their brokers). My question...when the $100 asset called "Reserves" ends up on the bank balance sheet...is the bank, at that point, free to lend it out (ie, convert "reserves" to "loans")? Can they pull that 'money' out of their Fed account and lend it to a customer? Or, does the $100 become the base so that they can now take in $1000 of deposits, and lend out *that* money?

    Where I get hung up is the notion that bonds aren't bought straight off of bank balance sheets...the bank had to spend money (credit a deposit account) to get the bond. Fed pays in 'reserves' -- bank pays by deposit account.

    ReplyDelete
    Replies
    1. Hmmm, I'm not sure how to answer that. Say you went in a bank and borrowed $10 cash. They could go in their vault and bring you the $10. That a case of reserves (vault cash in this case) being directly lent out. They replace one asset (the $10 cash) which pays 0% with another asset (your loan agreement) which pays some interest rate. As long as you're a good credit risk they're happy to do so.

      Far more common is for them to simply credit you a deposit account. Then when you take money out of the ATM you're diminishing their reserve levels, but not before.

      If banks lend reserves, this is far more likely to happen electronically from one Fed deposit holder (them) to another Fed deposit holder (another bank). Why? Perhaps the other bank needs reserves to meet it's reserve requirements.

      Delete
  6. The Tsy balance sheet looks like this both prior too and after the QE operation in both cases. chicago accounting services

    ReplyDelete