A few thoughts on risk, primary dealers and quantitative easing
This note comes from an American equity analyst. I have underlined the most important parts. I will give my interpretation below:
Bank Money held by those in the real economy is contracting. Fed Money is a different animal. Fed Money “outstanding” (“in-standing” would be a better phrase and I’ll explain why below) has been rising as a result of quantitative easing. Fed Money cannot be used by the non-financial sector, however. Here’s a quick overview of what happens when the Fed “quantitatively eases”:
- The New York Fed announces a tender for $X billion of certain issues of U.S. Treasury debt.
- The primary dealers (20 banks which are required to bid for U.S. Treasury debt at auction and to thereafter make markets in that debt – see the end of this note for a list of the PDs) submit offers to sell those certain issues to the New York Fed.
- The New York Fed purchases the Treasury debt from the primary dealers at the lowest possible price using newly created bank reserves.
- The newly created bank reserves show up as assets (“cash”) on the primary dealers’ balance sheets where previously the Treasury debt was listed as assets on the primary dealers’ balance sheets.
- The amount of bank reserves available for borrowing and lending in the Fed Funds market increases by the amount of the New York Fed’s purchase. [Edward here: this is irrelevant because banks lend to creditworthy borrowers irrespective of the amount of reserves in the system]
- Interest paid by the Treasury Department on the debt purchased/sold accrues to the Fed instead of the primary dealers (the Fed returns most of it to the Treasury at the end of the year).
- Interest rate risk on the Treasury debt (i.e. the risk that rising interest rates lower the present value of the purchased/sold Treasury debt, creating a loss in the event the holder wanted/needed to sell before maturity) is borne by the Fed instead of the primary dealers, although a recent rule change puts the Treasury Department on the hook for any losses. The Treasury Department would put future Fed profits toward returning the Fed’s negative liability position to zero. That means that the Treasury is pledging the future interest paid on its own debt as collateral for any principal (mark-to-market) losses on its own debt held in the Fed’s portfolio. The net effect is to weaken the barrier between the Treasury and the Fed which of course lessens the independence of the Fed. More importantly, it gives Federal Reserve Notes (what the Fed and everybody else erroneously calls “money”) debt-like properties and gives Treasury debt money-like properties. That means that the Treasury has moved one step closer to becoming the de facto monetary authority while the Fed has moved one step closer to becoming the de facto fiscal agent. (Dick Eide, a friend of mine who trades and writes for bond dealer Pierpont Securities, has written a hilarious piece on the recent Humphrey Hawkins testimony that I can send to you if you like. It addresses some of these issues.)
- Unlike the “cash” that is made available to a bank’s customer when they’ve taken out a loan (“Bank Money”), the “cash” that primary dealers receive in exchange for Treasuries in a quantitative easing transaction is those banks’ assets and not their liabilities. No Bank Money has been issued, and the Fed Money that has been issued merely swells the pot of Fed Funds that banks lend and borrow amongst themselves. The liability side of the banks’ balance sheets does not change and the banks’ capital levels do not change.
- The primary dealers (which remember are banks) will not want to hold the non-interest bearing cash on their balance sheets because there is a cost for funding those non-interest bearing assets no matter how trivial that cost may be (~20 basis points probably). Therefore, they are likely to turn around and buy some combination of (interest bearing) Treasury debt and other risky assets which promise a risk-adjusted return that justifies the cost of funding. Given that the cost of fundingis so low and given that the PDs need to net acquire more Treasuries for the next QE tender suggests that the bulk of the proceeds from a QE transaction are plowed back into Treasuries. Indeed, the below chart suggests that this is precisely what is happening. If primary dealers weren’t plowing the proceeds of QE transactions back into Treasuries, they would have to meet the New York Fed’s demand for them out of existing inventories and that would see net holdings decline – not flat-line as they have done since QE2 began in mid-November.
- The primary dealers are therefore merely facilitators – middle men – in this process. The “cash” from the QE transaction, which is functioning as a financial asset and not as “money” in the real economy sense of the word, passes to the net seller of Treasuries or to the issuer itself – to the Treasury.
PRIMARY DEALERS
PRIMARY DEALERS^
Barclays Capital Inc.BNP Paribas Securities Corp.
*Cantor Fitzgerald & Co.
*Citigroup Global Markets Inc.
Credit Suisse Securities (USA) LLC
Daiwa Capital Markets America Inc.
Deutsche Bank Securities Inc.
*Goldman, Sachs & Co.
HSBC Securities (USA) Inc.
*J.P. Morgan Securities LLC
*Jefferies & Company Inc.
*Merrill Lynch, Pierce, Fenner & Smith Incorporated
*MF Global
Mizuho Securities USA Inc.
*Morgan Stanley & Co. Incorporated
Nomura Securities International Inc.
RBC Capital Markets, LLC
RBS Securities Inc.
SG Americas Securities LLC.
UBS Securities LLC.
My interpretation
So here’s what’s happening in plain English.
- Congress and the Obama Administration, the U.S. fiscal authorities, must deficit spend if they want to prop up aggregate demand in the face of consumer deleveraging and increased saving (see How To Reduce Government Budget Deficits). The worry is that without this extra demand from the federal government, the U.S. economy would lapse back into recession and debt deflation would take over. Despite what government officials say officially, they know this would be disastrous for bank balance sheets and potential bankruptcies.
- The only way to engage in this kind of deficit spending without increasing inflation expectations in a way that also increases bond yields and tanks the economy is by getting the Federal Reserve to buy up Treasuries via quantitative easing. This forces U.S. government bond holders like Bill Gross to accept negative real yields. Some of these bondholders will reach for yield and take on more risk than they should.
- The Fed remits its profit to the Treasury at the end of the year, creating a circularity in fiscal spending, Treasury bond issuance, interest payment, and Fed profit remittance. "The primary dealers are therefore merely facilitators – middle men – in this process." As I have noted in the past, this should be seen as a non arms-length relationship that puts the Fed in a quasi-fiscal role, reducing its independence and making it a handmaiden of the Treasury. In my view, this is why Bernanke has spoken out against deficits; he fears the circularity of QE in monetising the Treasury’s debt will put it under attack in a world in which Ron Paul wields considerable power.
- Moreover, the Fed is taking on interest rate risk by buying the bonds. If yields were to rise too much, it would be faced with a choice: a. the central bank could engage in communications to adjust future interest rate expectations downward by signalling it wasn’t going to raise rates any time soon regardless of inflation. That’s what they have done so far. This has kept yields in check (but has also kept short rates negative in real terms); or b. The Fed could let inflation expectations and yields rise, potentially snuffing out the recovery and making the Fed subject to significant losses on its Treasury portfolio. Can the Fed go bankrupt? No, it does not mark to market or use GAAP. However, William Ford, a former Fed President, has indicated many are uneasy with the leverage and risk that the Fed’s QE represents.
- Finally,this is merely an asset swap of bonds for cash. The primary dealers who engage in these transactions with the Federal Reserve see no adjustment to the liability side of their balance sheets where demand deposits and debt obligations sit. Before the transaction, the dealers had interest-bearing bonds. Quantitative easing drains the economy of these interest-bearing assets. The Fed collects the Treasury bond interest payments the primary dealers once received. Now the banks have cash and are loath to sit on it. "The primary dealers (which remember are banks) will not want to hold the non-interest bearing cash on their balance sheets because there is a cost for funding those non-interest bearing assets no matter how trivial that cost may be." As Fed Vice Chairperson Janet Yellen has indicated, "It is conceivable that accommodative monetary policy could provide tinder for a buildup of leverage". But Fed officials driving the QE bus believe the benefits outweigh the risks.
The fact that the primary dealers are recycling most of their cash back into Treasuries tells you fears that U.S. interest rates would shoot up if China stopped buying Treasuries are entirely misplaced. That won’t happen. What could happen is that the dollar tumbles against floating rate currencies. The Chinese would be forced to either appreciate their currency or buy up other U.S. assets. On the other hand, the primary dealers are not recycling ALL of their cash back into Treasuries. And this is where the tinder for speculative excess and a buildup in leverage resides.
So many questions:
You say the primary dealers are required to bid for U.S. Treasury debt at auction – is that by law, and if so, which law? To what extent are they each required to bid? Do they have to bid on enough Treasuries to make sure the auctions are successful (all the offered Treasuries are sold)? If not, what is their minimum requirement?
If QE is not a factor in Treasury’s decision on how much debt to issue (how many bills, bonds, and notes to sell at auction), then doesn’t every dollar’s worth of QE force another investor to buy a dollar’s worth of some other security to hold instead of the Treasuries he or she would otherwise have bought?
If the dollar amount of QE is significant, wouldn’t this mechanism lead to price inflation in whichever assets these displaced investors choose as substitutes, followed by a chain reaction of price rises in successively riskier financial assets as investors displace each other?
If the Fed suddenly competes in the marketplace for investments with “new” QE dollars, is it likely that the marketplace will adjust via some entities’ issuing more debt or shares to soak up the added money, or that some investors will choose to consume instead, or that some investors will be forced to hold cash because all the other investment opportunities have been bought by others?
There is a circularity to the investing pattern of the primary dealers. As the original note quoted here points out, the vast majority of cash that the dealers receive for the Treasuries they relinquish in a QE transaction is recycled back into Treasuries. But, at the margin, you do have an increase in cash available for investment or speculation in other asset classes and that certainly bids prices up in those asset classes. Look at the IPOs of PE-backed companies as an example. These deals are coming to market because now is a unique opportunity to cash out while markets are flush with cash.
Thanks very much!
If the primary dealers have been recycling most of the QE money back into Treasuries (this means keeping them, right?), the effect of that would be to push other investors out of the Treasury-buying market, unless the U.S. Treasury increased the amount auctioned per month by enough to keep availability to non-Fed investors on roughly the same path it was before QE.
Your observations about IPOs etc. imply the former is true (other investors are pushed out of Treasury buying) to a significant extent.
As an outsider whose profession (engineering) requires precision of thought, it is very hard for me to read economics discussions. Far too often there are statements, like the one about primary dealers being “required” to bid in Treasury auctions, that need a whole lot more explanation to be meaningful. I never know whether the writer is just as clueless as I am about the precise meaning of their statements, or instead it’s a case of “everyone knows that, of course!”
Can you recommend some reading that would explain these details of how the banking system actually works, and what laws and regulations shape it?
So many questions:
You say the primary dealers are required to bid for U.S. Treasury debt at auction – is that by law, and if so, which law? To what extent are they each required to bid? Do they have to bid on enough Treasuries to make sure the auctions are successful (all the offered Treasuries are sold)? If not, what is their minimum requirement?
If QE is not a factor in Treasury’s decision on how much debt to issue (how many bills, bonds, and notes to sell at auction), then doesn’t every dollar’s worth of QE force another investor to buy a dollar’s worth of some other security to hold instead of the Treasuries he or she would otherwise have bought?
If the dollar amount of QE is significant, wouldn’t this mechanism lead to price inflation in whichever assets these displaced investors choose as substitutes, followed by a chain reaction of price rises in successively riskier financial assets as investors displace each other?
If the Fed suddenly competes in the marketplace for investments with “new” QE dollars, is it likely that the marketplace will adjust via some entities’ issuing more debt or shares to soak up the added money, or that some investors will choose to consume instead, or that some investors will be forced to hold cash because all the other investment opportunities have been bought by others?
There is a circularity to the investing pattern of the primary dealers. As the original note quoted here points out, the vast majority of cash that the dealers receive for the Treasuries they relinquish in a QE transaction is recycled back into Treasuries. But, at the margin, you do have an increase in cash available for investment or speculation in other asset classes and that certainly bids prices up in those asset classes. Look at the IPOs of PE-backed companies as an example. These deals are coming to market because now is a unique opportunity to cash out while markets are flush with cash.
Thanks very much!
If the primary dealers have been recycling most of the QE money back into Treasuries (this means keeping them, right?), the effect of that would be to push other investors out of the Treasury-buying market, unless the U.S. Treasury increased the amount auctioned per month by enough to keep availability to non-Fed investors on roughly the same path it was before QE.
Your observations about IPOs etc. imply the former is true (other investors are pushed out of Treasury buying) to a significant extent.
As an outsider whose profession (engineering) requires precision of thought, it is very hard for me to read economics discussions. Far too often there are statements, like the one about primary dealers being “required” to bid in Treasury auctions, that need a whole lot more explanation to be meaningful. I never know whether the writer is just as clueless as I am about the precise meaning of their statements, or instead it’s a case of “everyone knows that, of course!”
Can you recommend some reading that would explain these details of how the banking system actually works, and what laws and regulations shape it?
The money stock & its rate of utilization is unknown & unknowable. Even if you could categorize & list deposit types, by depository institution, the figures would still be prone to error. As Mises might say: there has been a confusion of money & credit (i.e., the FED’s research staff has decided that there is no difference between liquid assets & money).
But there is a technical difference between the supply of money & the supply of loan-funds. It has to do with the context in which lending occurs, whether lending & investing is performed by money creating depository institutions, or by the financial intermediaries (non-banks). I.e., the differentiating factor is the financial institution’s ability to create new money & credit (or simply transfer existing savings). The differentiating factor is not confined to the deposit type properties (the focus of the Austrian school).
This presents 2 questions. Which institututions create new money & how fast are these institutions COLLECTIVELY growing?
The DIDMCA gave the S&Ls, MSBs, & CUs the legal authority to become money creating depository institutions. Thus you cannot use the commercial bank credit (member bank loans & investments), figures in isolation (as CB credit is now also created by the thrifts). E.g., “bank credit proxy” (total bank liabilities),was used as a FOMC directive between 66 & 69. In fractional reserve banking, system-wide assets approximate system-wide liabilities – because as MMT’ers say “loans create deposits”.
I.e., changes in total bank credit are almost exclusively reflected in changes in deposit liabilities. This can be explained as follows: time/savings deposits (in money creating depository institutions), rather than being a source of loan funds, are the indirect consequence of prior bank credit creation. And the source of time/savings deposits is almost exclusively the transfer from other types of customer deposits (primarily transaction deposits).
The ability of the thrifts to create new money is why the BOG combines “(4) other checkable deposits (OCDs), consisting of negotiable order of withdrawal (NOW) and automatic transfer service (ATS) accounts at depository institutions, credit union share draft accounts, and demand deposits at thrift institutions” with CB system deposits.
So omitting large CDs, repurchase agreements, Euro-dollar borrowings (& the Treasury’s General Fund account), from the money stock measures, understates the growth rate of the aggregates (most retail & institutional MMMFs deposits are not traceable to deposit creation).
The last piece of the money puzzle is how fast is money being spent? The historical behavior of money suggests that the speed its moves – travels in just one direction – up. Even if it fell, the large shift in the composition of deposit types (from interest-bearing to transaction based accounts) would increase the current volume of transactions – and generate higher levels of gDp.
The money stock & its rate of utilization is unknown & unknowable. Even if you could categorize & list deposit types, by depository institution, the figures would still be prone to error. As Mises might say: there has been a confusion of money & credit (i.e., the FED’s research staff has decided that there is no difference between liquid assets & money).
But there is a technical difference between the supply of money & the supply of loan-funds. It has to do with the context in which lending occurs, whether lending & investing is performed by money creating depository institutions, or by the financial intermediaries (non-banks). I.e., the differentiating factor is the financial institution’s ability to create new money & credit (or simply transfer existing savings). The differentiating factor is not confined to the deposit type properties (the focus of the Austrian school).
This presents 2 questions. Which institututions create new money & how fast are these institutions COLLECTIVELY growing?
The DIDMCA gave the S&Ls, MSBs, & CUs the legal authority to become money creating depository institutions. Thus you cannot use the commercial bank credit (member bank loans & investments), figures in isolation (as CB credit is now also created by the thrifts). E.g., “bank credit proxy” (total bank liabilities),was used as a FOMC directive between 66 & 69. In fractional reserve banking, system-wide assets approximate system-wide liabilities – because as MMT’ers say “loans create deposits”.
I.e., changes in total bank credit are almost exclusively reflected in changes in deposit liabilities. This can be explained as follows: time/savings deposits (in money creating depository institutions), rather than being a source of loan funds, are the indirect consequence of prior bank credit creation. And the source of time/savings deposits is almost exclusively the transfer from other types of customer deposits (primarily transaction deposits).
The ability of the thrifts to create new money is why the BOG combines “(4) other checkable deposits (OCDs), consisting of negotiable order of withdrawal (NOW) and automatic transfer service (ATS) accounts at depository institutions, credit union share draft accounts, and demand deposits at thrift institutions” with CB system deposits.
So omitting large CDs, repurchase agreements, Euro-dollar borrowings (& the Treasury’s General Fund account), from the money stock measures, understates the growth rate of the aggregates (most retail & institutional MMMFs deposits are not traceable to deposit creation).
The last piece of the money puzzle is how fast is money being spent? The historical behavior of money suggests that the speed its moves – travels in just one direction – up. Even if it fell, the large shift in the composition of deposit types (from interest-bearing to transaction based accounts) would increase the current volume of transactions – and generate higher levels of gDp.