Monday, September 22, 2014

A few things the Fed has done right -- the oped

Now that 30 days have passed, I can post the whole oped from the Wall Street Journal. See previous post for additional commentary

A Few Things the Fed Has Done Right

The Fed's plan to maintain a large balance sheet and pay interest on bank reserves is good for financial stability.

As Federal Reserve officials lay the groundwork for raising interest rates, they are doing a few things right. They need a little cheering, and a bit more courage of their convictions.

The Fed now has a huge balance sheet. It owns about $4 trillion of Treasury bonds and mortgage-backed securities. It owes about $2.7 trillion of reserves (accounts banks have at the Fed), and $1.3 trillion of currency. When it is time to raise interest rates, the Fed will simply raise the interest it pays on reserves. It does not need to soak up those trillions of dollars of reserves by selling trillions of dollars of assets.

The Fed's plan to maintain a large balance sheet and pay interest on bank reserves, begun under former Chairman Ben Bernanke and continued under current Chair Janet Yellen, is highly desirable for a number of reasons—the most important of which is financial stability. Short version: Banks holding lots of reserves don't go under.

This policy is new and controversial. However, many arguments against it are based on fallacies.

People forget that when the Fed creates a dollar of reserves, it buys a dollar of Treasurys or government-guaranteed mortgage-backed securities. A bank gives the Fed a $1 Treasury, the Fed flips a switch and increases the bank's reserve account by $1. From this simple fact, it follows that:

• Reserves that pay market interest are not inflationary. Period. Now that banks have trillions more reserves than they need to satisfy regulations or service their deposits, banks don't care if they hold another dollar of interest-paying reserves or another dollar of Treasurys. They are perfect substitutes at the margin. Exchanging red M&Ms for green M&Ms does not help your diet. Commenters have seen the astonishing rise in reserves—from $50 billion in 2007 to $2.7 trillion today—and warned of hyperinflation to come. This is simply wrong as long as reserves pay market interest.

• Large reserves also aren't deflationary. Reserves are not "soaking up money that could be lent." The Fed is not "paying banks not to lend out the money" and therefore "starving the economy of investment." Every dollar invested in reserves is a dollar that used to be invested in a Treasury bill. A large Fed balance sheet has no effect on funds available for investment.

• The Fed is not "subsidizing banks" by paying interest on reserves. The interest that the Fed will pay on reserves will come from the interest it receives on its Treasury securities. If the Fed sold its government securities to banks, those banks would be getting the same interest directly from the Treasury.

The Fed has started a "reverse repurchase" program that will allow nonbank financial institutions effectively to have interest-paying reserves. This program was instituted to allow higher interest rates to spread more quickly through the economy.

Again, I see a larger benefit in financial stability. The demand for safe, interest-paying money expressed so far in overnight repurchase agreements, short-term commercial paper, auction-rate securities and other vehicles that exploded in the financial crisis can all be met by interest-paying reserves. Encouraging this switch is the keystone to avoiding another crisis. The Treasury should also offer fixed-value floating-rate electronically transferable debt.

This Fed reverse-repo program spawns many unfounded fears, even at the Fed. The July minutes of the Federal Open Market Committee revealed participants worried that "in times of financial stress, the facility's counterparties could shift investments toward the facility and away from financial and nonfinancial corporations."

This fear forgets basic accounting. The Fed controls the quantity of reserves. Reserves can only expand if the Fed chooses to buy assets—which is exactly what the Fed does in financial crises.

Furthermore, this fear forgets that investors who want the safety of Treasurys can buy them directly. Or they can put money in banks that in turn can hold reserves. The existence of the Fed's program has minuscule effects on investors' options in a crisis. Interest-paying reserves are just a money-market fund 100% invested in Treasurys with a great electronic payment mechanism. That's exactly what we should encourage for financial stability.

The Open Market Committee minutes also said that, "Participants noted that a relatively large [repurchase] facility had the potential to expand the Federal Reserve's role in financial intermediation and reshape the financial industry." Yes, and that's a feature not a bug. The financial industry failed and the Fed is reshaping it under the 2010 Dodd-Frank financial-reform law. Allowing money previously invested in run-prone shadow banking to be invested in 100%-safe reserves is the best thing the Fed could do to reshape the industry.

Temptations remain. For example, with trillions of reserves in excess of regulatory reserve requirements, the Fed loses what was left of its control over bank lending and deposit creation. The Fed will be tempted to use direct regulation and capital ratios to try to micromanage lending. It should not. The big balance sheet is a temptation for the Fed to buy all sorts of assets other than short-term Treasurys, and to meddle in many markets, as it is already supporting the mortgage market. It should not.

The Fed is making no promises about the stability of these arrangements—a large balance sheet and market interest on reserves available to non-banks. It should. In particular, it should clarify whether it will allow its balance sheet to shrink as long-term assets run off, or reinvest the proceeds as I would prefer.

Most of the financial stability benefits only occur if these arrangements are permanent and market participants know it. We can debate whether interest rate policy should follow rules or discretion, be predictable or adapt to each day's Fed desire. But the basic structures and institutions of monetary policy should be firm rules.

The remaining short-term question is when to raise rates. Ms. Yellen has already made an important decision: The Fed will not, for now, use interest-rate policy for "macroprudential" tinkering. This too is wise. We learned in the last crisis that the Fed is only composed of smart human beings. They are not clairvoyant and cannot tell a "bubble" from a boom in real time any better than the banks and hedge funds betting their own money on the difference. Manipulating interest rates to stabilize inflation is hard enough. Stabilizing inflation and unemployment is harder still. Additionally chasing will-o-wisp "bubbles," "imbalances" and "crowded trades" will only lead to greater macroeconomic and financial instability.

Here too a firm commitment would help. Otherwise market participants will be constantly looking over their shoulders for the Fed to start meddling in home and asset prices.

Plenty of uncertainties, challenges and temptations remain. Tomorrow, we can go back to investigating, arguing and complaining. Today let's cheer a few big things done right.

Mr. Cochrane is a professor of finance at the University of Chicago Booth School of Business, a senior fellow at the Hoover Institution, and an adjunct scholar at the Cato Institute.


  1. John,

    This statement needs some explaining:

    "The Fed is not subsidizing banks by paying interest on reserves. The interest that the Fed will pay on reserves will come from the interest it receives on its Treasury securities. If the Fed sold its government securities to banks, those banks would be getting the same interest directly from the Treasury."

    It sounds like the central bank is in essence "forcing" banks to hold more Treasuries. Central bank sets reserve requirements and then uses reserves to buy Treasuries. Bank with reserves receives interest payments on Treasuries through the central bank.

    How does the central bank set the interest paid on reserves independently of the auction interest rate on government bonds? Let's say the federal government auctions 5 year bonds at 3%. Two years from now, the central bank wants to set the interest on reserves rate to 4% - how do they do it if the interest payments on the five year bonds they are receiving from the Treasury are 3% annual?

    1. It's a bit dicer than I let on in the oped, for lack of space. It really goes the other way, if the Fed offers 4% on treasuries, it's counting that nobody will buy new treasuries at 4%, since they can put money in bank accounts that competition has (?) driven to 4%, or under the reverse repo program can put in reserves directly. We'll see if even the Fed is big enough to drive the whole treasury market -- or if it can drive the whole treasury market without really expanding the balance sheet.

    2. John,

      "It really goes the other way, if the Fed offers 4% on treasuries..."

      What does that mean? For the Fed to "offer 4%" on treasuries, it may have to sell it's Treasuries at a discount to par. But I thought that the Fed was going to hold Treasuries to maturity to fund interest payments on reserves.

      If the federal government auctions 3% 5 year bonds, where does the extra 1% come from to make 4% interest on reserve payments? The only way I could see to do it would be for the Fed to buy Treasuries at a premium and sell them at a discount.

      This would not be whole lot different than just having the central bank print the extra 1% interest and giving it to banks.

  2. John,

    "The Treasury should also offer fixed-value floating-rate electronically transferable debt."

    What are you describing here?

    1. Are you talking about zero coupon bonds sold at a discount par and traded publicly?
    2. Are you talking about floating interest rate securities (similar to ARM's)?

    In the first case, the Treasury (I think) already does this. In the second case, the Treasury department cannot do this without the ability to either raise tax revenue or cut other expenditures.

  3. "Reserves that pay market interest are not inflationary. Period."

    As long as it really is the MARKET RATE of interest. But if the market rate is 3%, and the Fed pays 4% on the reserve balances that private banks hold at the Fed, then the Fed's assets will fall relative to its liabilities, the Fed will eventually not have enough assets to buy back all the currency and reserves it has issued, and inflation will result.

    1. I tried to leave fiscal theory out, for space reasons. Yes, if the Fed loses a lot, it eventually needs to be recapitalized from the Treasury. If the ECB loses a lot on its Greek bonds, it similarly calls up the German taxpayer.

    2. The "period" needs to be removed. If the Fed finances the carrying cost of debt by printing money, it will be inflationary. So, depends on how interest cost is financed.

    3. Good point. I implicitly assumed the Fed holds only short term treasuries, that the treasury rate equals the interest rate on reserves, and that the Fed has treasuries equal to outstanding reserves. So the Fed can finance IOR from interest on treasuries. If not, reserves still do not have to be inflationary if the Treasury gives the Fed enough to pay the interest. Back to fiscal theory...

  4. You mention "Large reserves also aren't deflationary.....[not] "starving the economy of investment." ... no effect on funds available for investment"

    Isn't it true that reserves are a non-circulating currency (owned by banks at the Fed)? How are banks able to monetize "reserves" at all? they can't spend them like money, nor loan them like money. they can only benefit via unencumbered loan growth (w/o having to reserve against it in future) but that too if and only if they have unprecedented deposit growth to lend in the first place. How is that not deflationary?

  5. One of the Fed's great unsung success stories was offsetting the fiscal contraction of the sequester. It actually acknowledged in its September 18, 2013 statement that it was "taking into account the extent of federal fiscal retrenchment" in evaluating QE3 and whether or not to taper.

    Remember, at the beginning of 2013 many folks (including a a very popular economist-pundit were estimating that the sequester would cost 700,000 jobs, yet growth and unemployment improved at basically the exact same pace (slightly quicker, actually) in 2013 as 2012.

    Time to start beating the drum about the failure of the promised austerity crisis to appear.

    1. The Fed crossed a line when it moved past ZIRP to the MBS purchasing program. The elected politicians decided, in their collective wisdom, to have a contractionary fiscal policy. The Fed sought to frustrate the natural consequences of that policy by running what is in reality an off-setting fiscal policy of their own. On philosophical grounds and on practical political grounds it is wrong for the Fed to do that.

      From my point of view there is an irony in the Fed, which is so widely reviled in Republican circles, saving the Republicans' political skin.

    2. It's called in political circles "plausible deniability".

    3. Absalon,

      I beg to differ with your claim that on “philosophical grounds . . . it is wrong for the Fed to” counteract a “contractionary fiscal policy” implemented by “elected politicians”.

      Of course the latter Fed policy is arguably undemocratic. On the other hand, should politicians who know less about economics than the average snail be taking decisions on how much stimulus is suitable? You might as well have politicians design NASA’s next rocket or nuclear power stations.

      In contrast to that, politicians and the electorate certainly SHOULD take strictly political decisions, like what proportion of GDP is allocated to public spending and how that is split between education, roads, defence, etc.

      And disentangling the technical decision as to how much stimulus is suitable from the above strictly political decisions is easy. In fact that disentanglement is already achieved by the central bank committees that decide on interest rate changes and QE. I.e. when those committees decide on interest rate changes and QE, they do not at the same time tell politicians what proportion of GDP should be allocated to public spending, etc.

    4. Ralph

      When politicians run on a platform of cutting spending and get elected on that platform then it does not matter if you or I or a technocratically minded bureaucrat thinks that policy is bad economics. There is no doubt that the Republicans have a mandate from their voters to cut spending and shrink government. It is wrong for the Fed to seek to frustrate that agenda, no matter how wrong they think the policies are.

      There is also a moral hazard aspect here. If the Fed makes it their policy to cover for the consequences of bad policy choices by governments then that will only encourage voters and politicians to make more bad, self indulgent, policy choices.

      I do see a line that was crossed when the Fed moved from ZIRP to MBS purchases.

    5. Absalon,

      You haven't got the point I made above. Perhaps I didn't make it clear enough.

      I'm very definitely not saying that if the electorate wants to cut public spending as a proportion of GDP, the technocrats should interfere with that decision. The decision as to proportion of GDP should be allocated to public spending and how that spending is split between education, defence etc are POLITICAL decisions and it would be totally out of order for technocrats to interfere with those decisions.

      In contrast, what I AM SAYING is that technocrats should decide on the level of stimulus that is suitable. E.g. technocrats should decide by how much government spending should exceed taxes in the next year. Having made that decision, it is entirely up to the electorate and politicians to effect that extra net spending via more public spending or less tax.

      In fact technocrats already pretty much decide on stimulus (while not interfering with political decisions) in that central bank committees decide on interest rate changes, QE, etc.

    6. Ralph: We focused on different aspects. Your position is that "technocrats should decide by how much government spending should exceed taxes in the next year"

      I disagree. Strongly. I disagree in principle. No unelected group should be given that power. The Fed has not been given that power, either explicitly or implicitly. The Fed has the power to set the Federal funds rate. When they started to buy MBS they moved beyond their function.

  6. the most important of which is financial stability. Short version: Banks holding lots of reserves don't go under.

    It does not seem that holding bank reserves should be any better than holding treasuries. The Fed is just acting as a particularly large money market fund intermediating between bank investments by way of reserves on deposit with the Fed and a diversified portfolio of Treasuries and MBS.

  7. "People forget that when the Fed creates a dollar of reserves, it buys a dollar of Treasurys or government-guaranteed mortgage-backed securities. A bank gives the Fed a $1 Treasury, the Fed flips a switch and increases the bank's reserve account by $1."

    When the Fed "flips a switch" to increase the banks reserve account, how is that not the same thing as printing $1 and giving it to the bank? Doesn't the money supply increase by 1 extra dollar that did not exist before? Presumably, that dollar will come right back to the Fed at maturity, so assuming that the bank maintains its reserves, the Bank still has its dollar and now the Fed, at maturity, has another, which you prefer it use to buy more treasuries.

    From that fact, isn't that inflation????

  8. Fine by me. But, seems like the Fed can avoid inflation even paying well-below market interest. Banks are sitting on reserves earning 0.25%.
    Should the Fed eliminate the entire national debt?
    Also, in the past, John Cochran has stated the Fed could pay interest simply by printing new money. Now we seem be talking about transferring money from the Treasury to the Fed to pay interest on reserves.
    Which is it?

  9. Prof Cochrane,

    You say “The demand for safe, interest-paying money . . . can all met by interest-paying reserves.”

    Why is there any obligation on taxpayers to pay interest to people who want to hoard base money / reserves? The latter comes to the same thing as hoarding $100 bills under one’s mattress. If people want to save in the “mattress” way, that’s fine by me. But there’s no reason to reward anyone for doing so.

    In contrast, assuming NO INTEREST is paid on reserves, there is a very good reason to supply the private sector with the stock of reserves / base money it wants: if that supply is not made available, the private sector will try to save so as to acquire to the stock of base money it wants, and we’d get paradox of thrift unemployment.


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.