Hatzius on Sector Balances, NGDP Targeting, Fiscal Policy & More

Joe Weisenthal has a really fantastic interview with Goldman’s Chief Economist Jan Hatzius over at Business Insider.  They touch on a lot of moving parts, but I found the sections on the sector balances of particular interest.  Obviously, I am biased because this is very similar to the framework I’ve been working under for the last 5 years, but Hatzius really gets the big picture view and does a great job connecting all the dots.

Other things he touches on include the outlook for 2013, the fiscal cliff, the risk of recession, NGDP Targeting, Richard Koo and the efficacy of fiscal policy versus monetary policy.

Here’s a brief snippet:

“HATZIUS:There is an accounting identity which is issued, if you start with the global economy, to simplify it, that every dollar of government deficits has to be offset with private sector surpluses purely from an accounting standpoint, because one sector’s income is another sector’s spending, so it all has to add up to zero.That’s the starting point. It’s a truism, basically.

Where it goes from being a truism and an accounting identity to an economic relationship is once you recognize that cyclical impulses to the economy depend on desired changes in these sector financial balances.

If the business sector is basically trying to reduce its financial surplus at a more rapid surplus than the government is trying to reduce its deficit then you’re getting a net positive impulse to spending which then translates into stronger higher, more income, and ultimately feeds back into spending.

Conversely, if the business sector reduces its desired surplus by less than the government sector tries to reduce the budget deficit, then you end up with cyclical weakness. It’s a little heavy-going, to put this into words, and can be challenging.

There’s a reasonable amount of jargon involved (ex post, ex ante type of stuff), and I generally try to avoid having too much jargon and too much heavy going in the sort of things that we write, but in this particular case it is useful and worthwhile to wade through this a little bit. It’s an exception, but I think it is useful in this case.”

Read the whole thing here.

Cullen Roche

Mr. Roche is the Founder of Orcam Financial Group, LLC. Orcam is a financial services firm offering research, private advisory, institutional consulting and educational services.

More Posts - Website

Follow Me:


  1. Reading Hatziu’s comments almost sounds like reading Pragcap. Did you steal his framework or vice versa? :-)

  2. I have a hard time grasping this concept.
    If the Treasury wants to borrow $5 trillion next year, no doubt the primary dealers can provide that money, but does that necessarily mean the private sector is saving $5 trillion.
    How does one know the primary dealers aren’t buying bonds with inside money, or the public switches to buying Treasuries instead of some other debt instrument.

    • My understanding:
      If the public is funding the debt, by buying bonds, then then we have a net gain of a financial asset.
      My dollar goes to the government and is spent, PLUS I still have my dollar, except it’s in the form of a Treasury. Now you have said this is 99 percent money, but the system operates under the assumption that a Treasury (especially a 90-day treasury) is cash. It acts like cash, it is cash.
      Plus … and this is critical … the U.S. is sovereign in its currency so if you won’t take my Treasury bond as payment, Uncle Sam will. So that makes it 100 percent cash, imo, and I think that is what you mean, if you’re not quite prepared to say that.
      The inflation constraint comes later — do you believe that someday those money-like Treasuries will be spent into the economy in a flood? If they are — if the holders want to spend those Treasuries, presumably they will need younger savers. Or from printing. Our society is getting older and young people are earning less — will they be able to use their savings to fund their grandparents’ retirement?

      • I don’t agree with you that treasuries are cash. They are securities. Low risk, highly liquid securities. Saying that treasuries are cash equivalents is like saying that Berkshire Hathaway stock is cash. I don’t think that’s correct. They are securities that represent claims on underlying cash flows from output.

        T-bonds have a very high moneyness (and I have NEVER said they’re “99% money”), but that doesn’t mean they’re cash equivalents. I think you’re moving down a very slippery slope when you start defining securities as money. They are money-like. But not money.

        • Respectfully, this is the heart of the debate.
          If a Treasury is cash, then we have no worries, other than the inflation constraint, because we are sovereign in our currency.
          And even if it would cause inflation, the government is going to redeem those bonds at par. If not, there would be riots, imo.

          But if a Treasury is viewed as a debt instrument, then the public is right to be worried.

          • The public does have a right to be worried about the value of its savings (which is all the debt is – the govt’s debt is the private sector’s saving). If the US govt behaves in a way that is so irresponsible that it detracts from the quality of overall output then the govt is directly devaluing the private sector’s saving.

            For instance, let’s just be extreme here and say the govt starts paying the entire US population to stay at home and watch TV. What would happen? There would be no output, no goods and services being made. Money would literally become useless. There would, by definition, be too much money chasing too few goods because the goods would cease to be produced. So your savings would get devalued because of the govt’s incompetence and this sort of a scenario would likely result in a hyperinflation that would wreck the value of US Dollars and therefore wreck the value of Tsy bonds (which are securities giving one the right to claim cash flow from output).

            The fact that the US govt has this special arrangement with the Dealers doesn’t mean it is immune to problems.

            • As an investor and observer of the economy and politics, do you believe that if I buy T-bonds and reinvest the interest that I will be buying the same, better or worse standard of living when I sell in 20 years?

      • John that is the way I see it, but some will bog you in technical details such as the definition of money. Some believe Treasuries are not “money” and reserves are. But in the end NFA’s have increased and the reserves remain as the government spends into someone else’s checking account. Also you have the income from the treasury adding to the economy – QE is draining that income.

    • Easy-peasy without all the jargon and details: when one sector runs a deficit, then by definition this sector is spending more than it earns. So far so good? Then by the same definition the rest of the world must be earning more than it spends. There is simply no other way around it – if the money is not “here” then it must be “elsewhere”. Thus the rest of the world is running a surplus. The rest of the world is also known as “non-government sector”. So, when the government is running a deficit, the non-govt sector is running a surplus. This is a truism, as Hatzius says. All the rest – issuance of bonds etc – is technicalities.

    • Hi Cullen, could you clarify “In other words, there is no new money creation from a t-bond sale.”

      Treasury sells bond to PD
      PD borrows from bank, pays Tsy
      Treasury spends $$’s (deposits in private accounts)
      $$’s deposited in MMF
      PD borrows from MMF and repays bank
      PD sells bond, repays MMF
      private sector holds Tsys

      I thought most of the money is newly-created inside money, except for small portion for reserve requirements? (Ref: http://pragcap.com/understanding-modern-monetary-system: “public sector issues coins, paper cash and banking sector reserves (“outside money”)”)?

      Also, “funding the t-bond purchase with pre-existing money” … as you say, it’s a flow; should think of money as first-in, first-out?


      • The intra-day loan doesn’t stay on the books. So there’s no net new inside money created from the bond auction. The govt is essentially recycling or redistributing existing inside money. The govt is in the business of taking from Peter to pay Paul (for public purpose). The actual “money printing” is the creation of the new bond. The bond isn’t technically money though (at least in my opinion). So, since the loan doesn’t result in new money creation the best way to think about “money printing” is to think about the loan creation process as it generally pertains to the private sector.

        • I guess the trillion-dollar question is what happens to the bond.
          I guess three things could happen.
          1. Future taxpayer receipts allow Uncle Sam to retire the debt. Odds of this happening: Zero? What future taxpayer would agree to this. What economist would agree to destroying money.
          2. Bond holders demand their savings back for living expenses and there are more sellers than buyers. Then the Fed steps in and monetizes the debt.
          3. Public runs out of savings so bonds can’t be rolled over and Fed steps in to buy bonds directly. Again, monetizing.

          You know, I don’t even think monetizing is a bad idea, anymore. It seems inevitable. If you think we won’t have inflation from doing this, let’s go ahead. Just please let’s spend the new money in the best way.

        • Thanks Cullen. I even screwed up my own accounting. Good thing I’m not an accountant.

          • Ha, well neither am I. The easiest way to think about this process is to think about you buying a bond via Tsy direct. Rather than all this repo nonsense you just use your existing inside money to buy a bond (a little oversimplified) and the govt takes that inside money and pays someone else for doing work on behalf of the govt.

            The dealer process is the same (in the end), but a bit more complex in the inner workings.

  3. “Here’s how the NY Fed has described the borrowing process to me”

    Look at Cullen going into the belly of the beast :)

    • Yeah, I finally got a straight answer from the source. This is from someone at their Capital Markets Group so yes, deep in the belly of the beast. And it confirmed MR’s thinking precisely. :-)

      • Nice detective work. Regarding viewing Treasuries as securities, I agree but they are a VERY special type of security. They are the only type of security that is a liability of the US govt. Cash is also a liability of the US govt. This makes cash very similar to Treasuries. I’m not saying Treasuries ARE cash, but they are pretty close! After all, the govt could replace every Treasury security with cash (or reserves) tomorrow in one giant QE with very limited consequences in terms of NFA or inflation.

  4. Finally (at least for me) Mr. Hatzius has identified (in the full interview) where MMT theory meets reality and provides actionable information. He says:

    “Where it goes from being a truism and an accounting identity to an economic relationship is once you recognize that cyclical impulses to the economy depend on desired changes in these sector financial balances.”

    Although “desired changes” seems like an odd choice of words, as opposed to say “expected trends”.

      • So, Hatzius’s assertion that “cyclical impulses to the economy depend on desired changes in these sector financial balances.” is not consistent with MMT and MMR theory for all practical purposes?

        • Kaldor and Godley created the sectoral balances. If you’re using the sector balances you’re not using MR or (MMT). You’re using the sectoral balances as created by Kaldor and Godley. Just as Hatzius properly attributed it.

          • Yes, thanx. I understand the sector balances framework is technically different than MR or MMT. I was simply suggesting that perhaps the practical ramifications of either was very similar, with respect to monitoring private savings against government spending deficits as a way to gage upcoming economic expansion rates. No?