Thursday, February 21, 2013

World GDP versus Total Value of Financial Asset Market Exchanges

Here are two important statistics:
(1) World GDP (2007):
US $65.61 trillion

(2) Global annual value of major financial asset market transactions (2007):
US $900 trillion*

* This includes foreign exchange turnover and stock market trading (excluding bonds and other over-the-counter transactions).
What is significant about this?

The significance is that global GDP is dwarfed by the value of major financial asset market transactions. A vast amount of spending that occurs every year is on secondary financial markets, including markets where foreign exchange, stocks, and shares are sold.

I assume that some of this $900 trillion involves foreign exchange transactions for international trade and so on (that is, for purchases of goods and services), but it is estimated that over 90% of foreign exchange transactions are speculative. At any rate, the total value of world trade (merchandise exports plus commercial services) was about $16.9 trillion in 2007, but not all of this required foreign exchange transactions (e.g., trade between nations in the Eurozone involves countries using the same currency, the Euro). Moreover, given that over-the-counter transactions and bond trading is excluded from the estimate above, it is obviously an underestimate of the real global aggregate value of such trading.

Any economic theory that ignores this type of spending and its sector of the economy (i.e., the secondary financial asset markets) is deeply flawed and liable to be missing something fundamental about modern market economies. In any one year, money can get sucked into this world of financial asset market transactions and essentially trapped there for a significant period of time as it is used to buy and sell assets over and over again.

Spending on the secondary financial asset markets is essentially a type of transaction that does not induce employment in the way that spending on final goods and services does. Most financial assets are non-reproducible, in the sense that businesses do not hire a significant number of workers or factor inputs when demand for these assets rises (Davidson 2002: 44), for they already exist in vast quantities in many different countries.

This type of spending is also a fundamental reason why Say’s law is one of the most ridiculous ideas ever formulated by economists.

The late Frank H. Hahn pointed out why:
“there are ... resting places for saving other than reproducible assets [i.e., final goods and services – LK]. In our model this is money. But land, as Keynes to his credit understood, would have just the same consequences and so would Old Masters. It is therefore not money which is required to do away with a Say’s Law-like proposition that the supply of labour is the demand for goods produced by labour. Any non-reproducible asset will do. When Say’s law is correctly formulated for an economy with non-reproducible goods it does not yield the conclusions to be found in textbooks. As I have already noted Keynes was fully aware of this and that is why he devoted so much space to the theory of choice amongst alternative stores of value.” (Hahn 1977: 31).
As individuals become richer and richer, the less likely it is that their income will be spent on final goods and services. That is to say, the wealthy have a lower marginal propensity to consume than poor classes of people.

So what do the rich and ultra-rich mostly spend their money on? The answer is: mostly assets on secondary financial markets (either directly or, more likely, indirectly via financial institutions). This is why there is no necessary reason for Say’s law to hold true in any modern capitalist economy.

The stupidity of the assumptions behind Say’s law goes right back to Jean Baptiste Say (1767–1832) himself:
Every producer [= capitalist] asks for money in exchange for his products, only for the purpose of employing that money again immediately in the purchase of another product; for we do not consume money, and it is not sought after in ordinary cases to conceal it: thus, when a producer desires to exchange his product for money, he may be considered as already asking for the merchandise which he proposes to buy with this money. It is thus that the producers, though they have all of them the air of demanding money for their goods, do in reality demand merchandise for their merchandise.” (Say 1816: 103–105).
When you believe (like Say) that capitalists only ever spend their money on good and services (whether consumption goods or factor inputs for further production), it is a recipe for disastrous economic theory.

UPDATE
It occurs to me that I should have mentioned another source of non-employment inducing demand: spending on real assets with low or relatively low elasticities of production (e.g., gold).

As one moves to real assets with moderate to high elasticities of production, the story is different, of course.

Furthermore, one might reply with the question: why is it that during asset bubbles, economies tend to have booms? But the booms associated with asset bubbles tend to be driven by the modern financial system and its creation of credit to fuel asset bubbles, consumption and investment, with a positive wealth effect amongst owners of the assets whose price is rising (either more consumption out of income or credit). Capitalists are caught up in the boom via their optimistic expectations.

If in a recession demand to hold money increases and demand to hold non-reproducible financial assets also rises (by using money to buy them), we have in the latter case exactly what can be called non-employment inducing demand.

BIBLIOGRAPHY

Davidson, P. 2002. Financial Markets, Money, and the Real World. Edward Elgar, Cheltenham.

Hahn, F. H. 1977. “Keynesian Economics and General Equilibrium Theory: Reflections on Some Current Debates,” in G. C. Harcourt (ed.), The Microeconomic Foundations of Macroeconomics Macmillan, London. 25–40.

Say, J. B. 1816. Catechism of Political Economy, or, Familiar conversations on the manner in which wealth is produced, distributed, and consumed in society (trans. J. Richter). Sherwood, Neely, and Jones, London.

37 comments:

  1. Hahn is wrong (to be honest I've never thought much of the guy).

    1. The asset to occupy the position of placeholder for wealth -- or wealth hoard -- need not be non-reproducible. Housing works, as does gold. The important point, as Kalecki knew, is that it should be of limited supply.

    2. Money is NOT non-reproducible. In fact it has potentially almost infinite capacity for increasing supply. So why money? Well, that's why Keynes is more interesting than Hahn. Hahn was good at mathematics, but he was far from a profound thinker.

    ReplyDelete
    Replies
    1. (1) Hahn seems to have been a curious fellow! His relations with Post Keynesians were not exactly friendly:

      http://robertvienneau.blogspot.com/2006/12/hahn-and-harcourt-amusing-crowds.html

      (2) yes, that is certainly true. Obviously real assets can be "resting places" for savings too.

      (3) but surely under the old gold standard days, you would have to say that the elasticity of production of the monetary base (gold) was mostly low? Credit money (though you can create it) has limited value.

      I think we might have a problem with proper definitions here.

      I mean "non-reproducible" in the sense that "businesses do not hire a significant number of workers or factor inputs when demand for the asset rises".

      When demand for money rises, businesses simply do not employ the unemployed workers to "manufacture" money.

      Certainly the central banks can create all the base money needed in these days of fiat money, but that does not reduce unemployment per se.

      Delete
    2. Just to further clarify what I mean:

      "A zero elasticity of production means that there is no change in the number of workers in the money producing industry required to provide any given percentage change in the quantity of money supplied.
      An endogenous (credit) money supply, on the other hand, usually implies an infinitely elastic quantity of money supply curve – the horizontalist argument of Moore (1988) – even though the credit money's elasticity of production is approximately zero. In essence, there will be little or no change in employment in the banking (credit supplying) industry to produce any change in the quantity of credit money supplied."


      Davidson, P. 1996. "In Defence of Post Keynesian Economics: A Response to Mongiovi," in S. Pressman (ed.), Interactions in Political Economy: Malvern After Ten Years. Routledge, London and New York. 120-132 at p. 128.

      Delete
    3. " but surely under the old gold standard days, you would have to say that the elasticity of production of the monetary base (gold) was mostly low?"

      Yeah, supply is limited... until it isn't. Gold standards never actually work because when the gold reserves deplete the government unpegs. Yes, they can try austerity first but this is just superstition and you see the same thing under fiat money systems. Gold standard thinking really is barbarous. The intellectual equivalent to tribal fetishism.

      Delete
  2. While the kind of asset-trading that you describe may (or may not) be indicative of speculative bubbles and other sign of economic ill-health I do not see how they are directly relevant to spending on final goods or are relevent to say's law.

    For every asset traded there is both a buyer and a seller so overall the money supply is not affected. It is true that the velocity of these kinds of transactions may be increasing during a particular time period and this may lead to a decrease in the velocity of spending on final-goods. Such changes in velocity (aka the demand for money) have the potential to disrupt the real economy and cause (in your terms) Says law to "fail". However such "monetary disequilibrium" conditions are well understood by economists. Even a simple market-monetarists model has a solution - keep spending on final goods constant by adjusting the money supply and the effects of changes in velocity are minimized.

    Such AD stabilizing measures would not prevent the "asset bubbles" that may have accompanied the decrease in velocity - but controlling such bubbles is outside the scope of monetary policy.

    ReplyDelete
    Replies
    1. "Such changes in velocity (aka the demand for money) have the potential to disrupt the real economy and cause (in your terms) Says law to "fail". However such "monetary disequilibrium" conditions are well understood by economists. Even a simple market-monetarists model has a solution - keep spending on final goods constant by adjusting the money supply and the effects of changes in velocity are minimized."

      No, it is unlikely this monetarist and "monetary disequilibrium" Austrian solution will in fact stabilize aggregate demand.

      Why? Because it is all dependent on the feeble, blunt instrument of monetary policy.

      Increases bank reserves will NOT increase capital goods investments or consumption, if business expectations are shocked and households and businesses are already struggling with excessive debt and debt deflationary dynamics. If monetary policy could reliably stabilize aggregate demand, then why has round after round of QE failed to restore high employment?

      The only effective solution is fiscal policy: Keynesianism.

      Delete
    2. Well said LK.There is no shortcut,either of "Market"- or- "New -Monetarist or MMT or whatever monetary solution in itself that creates a solution to a recession.The only effective way to go,is a active Keynsian fiscal policy!

      Delete
  3. I am only aware of QE being used in a context of inflation targeting - where it has succeeded. I am not aware of it being used to hit an AD (NDGP) target but there is no reason why it would not that I can see.

    Fiscal policy can obviously also be used to increase AS, though it does have the problem that it can be negated by off-setting monetary policy. If used wisely (for example on a general subsidy on investment spending ) then perhaps it could be used usefully during deep recessions where QE is politically unacceptable. Fiscal policy however if used badly (for example for direct govt employment of recourses) will likely further reduce private investment and further delay a real recovery.

    ReplyDelete
    Replies
    1. I am only aware of QE being used in a context of inflation targeting

      This is obviously not correct.

      You're saying that the Fed has not been attempting to stimulate the economy by means of 3 rounds of QE?

      Why, then, is that exactly what Bernanke et al. have been saying all this time?

      http://money.cnn.com/2012/09/13/news/economy/federal-reserve-qe3/index.html

      Delete
  4. Correction: Fiscal policy can obviously also be used to increase "AD" (not "AS")

    ReplyDelete
  5. The fed targets inflation not NGDP so of course QE in USA has been done in a context of IT not NGDPT. Note that inflation has stayed on target since QE has been implemented.

    ReplyDelete
    Replies
    1. Let me rephrase the question: what monetary policy, in your view, would constitute NGDP targeting?

      And if you say: having the Fed print money (whether in cash or electronically) and give to to people to spend, no, that ceases to be monetary policy; that is a de facto fiscal policy.

      Delete
    2. If you advocate the basic market monetarist form of NGDPT, the I assume you want this:

      - A McCallum Rule for the NGDP level. Such a rule would increase the growth rate of base money if NGDP is above trend and reduce the growth rate if NGDP is below trend.
      - Target the forecast. A central bank could look at past NGDP, current asset prices, TIPS spreads, and forecasts of economists to guide their policy. This type of policy suffers from the circularity problem.
      - Target the market forecast. Create an NGDP futures market and use it as a guide.
      - Make the currency convertible to NGDP futures.

      http://www.ngdp.info/

      But that is actually the same error committed by any number of monetarists and non-Keynesians since the 1930s really: it is pushing on a string.

      Expanding the base money will not drive more investment or consumption in this economy. QE 1, 2, and 3 and HAVE increased base money dramatically.

      Have they driven the economy to back to full employment?

      In general, see here:

      http://unlearningeconomics.wordpress.com/2012/05/08/on-the-lousy-reasoning-behind-ngdp-targeting/

      Delete
    3. Also:

      http://www.winterspeak.com/2011/10/latest-stupidity-ngdp.html

      Delete
    4. "Further, it's not clear how a nominal GDP target would be different from what the Fed is already doing, namely acting as a lender of last resort, and keeping interest rates (short and long, the latter through QE) low."

      http://nakedkeynesianism.blogspot.com/2011/10/nominal-output-targeting.html

      Delete
    5. Seems like you are saying 2 related things

      1. It is sometimes impossible to hit the target
      2. Even it it was possible it would be a bad idea

      On 1) The only way it would NOT be possible to hit the target is if beyond a certain point increasing the money supply did not lead to any increased spending. Is that what you are claiming is the case ?

      on 2) I can also see some scenarios where targeting NGDP is not optimal compared to the free-market alternative. Still thinking that through.

      Delete
    6. Base money (M0) is distinct from broad money (M1, M2, and discontinued M3), right?

      I have already pointed you to the problem: the only direct transmission mechanism from M0 to broad money (and hence private investment and consumption spending) is via bank credit.

      Increases in bank reserves will not increase capital goods investments or consumption, if business expectations are shocked and households and businesses are already struggling with excessive debt and debt deflationary dynamics, and do not want to take on the necessary new debt.

      If monetary policy could reliably stabilize aggregate demand, we would have seen significant US output and employment growth after QE 1, QE 2, and QE 3.

      And without a direct transmission mechanism, then all you really saying is: if the Fed increases the monetary base by x%, then nominal GDP should growth by x%.

      Well, the monetary base in the United States soared by 140% from about 2008-2009!

      Did that lead to high NGDP growth that created full employment in the US? Even nominal GDP growth was nowhere near that.

      Delete
    7. LK,

      What is your opinion on overt money financed deficits (helicopter money)?

      http://www.fsa.gov.uk/library/communication/speeches/2013/0206-at

      Turner refers to Koo's work about Japan. Fiscal deficits have kept Japan's economy float, but this meant a transfer of the debt burden from the private to the public sector, which could also be unsustainable. OMF would have avoided this argues Turner (However, Koo argues that fiscal policy was not consistently applied in Japan and this damaged the economy and increased public debt. So maybe if the Japanese government would have been more wise, fiscal policy might have done more to increase growth.)

      Delete
    8. (1) Koo is right. Japan turned to austerity from 1997-98 and the deficit soared

      http://socialdemocracy21stcentury.blogspot.com/2013/01/japanese-real-gdp-growth-19252001.html

      (2) Yes, if Japan had

      (a) fixed its banking sector (clearing banks of bad debts and non-performing assets)

      (b) reduced private debt to GDP levels by writing off and restructuring debt, and

      (c) engaged in sustained and strong stimulus with functional finance (some direct central bank buying of bonds)

      yes, they could have avoided very high levels of government debt and ended the lost decade earlier.

      Delete
    9. "(c) engaged in sustained and strong stimulus with functional finance (some direct central bank buying of bonds)"

      So OMF can be a useful tool.

      Delete
    10. On page 174 of The Holy Grail of Macroeconomics (updated edition, 2009) Koo writes that fiscal policy is only effective when the private sector is paying down debt. He also writes that Keynesian policies in the fifties and sixties were a mistake, as the private sector was not paying down debt then. I assume you don´t agree with that?

      Delete
    11. Jan,

      (1) "On page 174 of The Holy Grail of Macroeconomics (updated edition, 2009) Koo writes that fiscal policy is only effective when the private sector is paying down debt. "

      Only "when the private sector is paying down debt"?

      That sounds awfully strange to me.

      Shouldn't it be only really/very effective "when the private sector is not *significantly* paying down debt"?

      (2) If Koo really thinks fiscal policy was a mistake in the 50s and 60s, then I do disagree.

      Delete
    12. Yes, he even put ´only´ in italics. That´s one problem I have with Koo. His ideas are very clear but often his thinking is quite rigid. It is black or it is white. Oh, and he believes in the money multiplier.

      Delete
    13. To eliminate all doubt:

      "In a normal or yang phase, in which businesses have healthy balance sheets and are maximizing profits, private-sector loan demand is robust and responsive to changes in interest rates. In this world, monetary policy should be the main tool for reducing fluctuations in economic activity. Fiscal stimulus should be avoided, because it leads to crowding out, inflation and rising interest rates, and can interfere with the optimal allocation of resources. In the yang phase, upon which the theories of the neoclassicals, monetarists, and New Keynesians are all based, smaller government is better." (p. 147)

      "Although I was 100 percent immersed in conventional economics in the late 1970s...I support Reagan because I believed that America's economic problems could not be solved by conventional macroeconomic policy , and instead required a substantial expansion of the nation's ability to supply goods and services. I still believe the decision I made at that time was correct." (p. 5)

      Delete
    14. So is he endorsing the idea that "fiscal stimulus should be avoided, because it leads to crowding out, inflation and rising interest rates, and can interfere with the optimal allocation of resources" even when the private debt level low?

      If so, Koo - for all his good sense on Japan's lost decade - is much more conservative/neoclassical than I thought he was.

      Delete
    15. He's very much the asian Paul Krugman I'm afraid.

      He needs to read more Minsky.

      Delete
    16. Well, he doesn't refer to the level of private debt, the only thing that seems to be important for him is the balance sheet of the private sector (negative equity or not). As long as the private sector is willing to borrow, fiscal policy is not necessary or even harmful. He is very good at describing the consequences of what happens if a debt-fuelled bubble bursts, but he doesn't give much attention to the build-up of such a bubble through debt en credit dynamics (as I said, his view of money is quite classical, he accepts the money multiplier), at least not in the The Holy Grail.

      Delete
  6. Monetary policy = policies that change the supply of money
    Fiscal policy = policies that change the distributions of income or wealth

    Some policies would have element of both. The one you mention "having the Fed print money (whether in cash or electronically) and give to to people to spend, no, that ceases to be monetary policy; that is a de facto fiscal policy" would both increase the money supply (monetary) and change the distribution (the people who get the new money are better off).

    QE-type policies are as close as one get to pure monetary policy. The CB swaps assets for money. The people who sell assets are marginally better off because they got a higher price than they otherwise would but this is minor compared with just printing money and giving it to people.

    The monetary effects should be the same in both cases though: Some of the new money will be spent and some saved. If you engage in enough monetary policy (QE or just giving money away) then you can hit any level of spending you want and so hit any inflation or NGDP target you set.

    You could probably come up with different definitions of fiscal v monetary (I know that Randall Wray thinks that QE is fiscal policy). But that a minor issue. The important thing is to understand what "increasing the money supply" means and what its effects will be - no matter how you categorize it.

    Keynesian counter-cyclical fiscal policy works by leaving the money supply unchanged and redistributing it between those who save and those who spend. In a recession govt borrowing is used to tap unused saving and divert it to the govt who will spend. The problem for me with this is that it thwarts rather than helps with the recovery.

    ReplyDelete
    Replies
    1. " If you engage in enough monetary policy (QE or just giving money away) then you can hit any level of spending you want and so hit any inflation or NGDP target you set."

      But just look at what advocates of NGDPT want: it is not "just giving money away" at all.

      Their only substantive policy actions are (basically) just inflating the base money by x% and announcing beforehand that they want the nominal GDP to grow by x%.

      That won't necessarily even increase real GDP, and the whole purpose of any stimulus is to increase employment and output.

      No, Rob Rawlings, NGDPT is just more monetary nonsense. If QE 1, QE 2, and QE 3 could not reduce unemployment and stimulate the US economy by massive inflation of the base money supply, then NGDP certainly won't.

      And I assume you object to fiscal policy because you subscribe to the ABCT?

      ABCT is simply wrong:

      http://socialdemocracy21stcentury.blogspot.com/2013/01/debunking-austrian-economics-101-updated.html

      (See links at the end)

      Delete
  7. Its theoretical possible that increasing NGDP within the context of a target will have no effect on RGDP but doesn't that seem unlikely to you - as a Keynesian to whom the role of AD is central ?

    I'm not against fiscal policy if redistribution is what you want to achieve (for example to address low wage issues). I just think it sub-optimal as a policy to stabilize the economy. BTW: ABCT would likely make one opposed to monetary policy rather than fiscal policy so that claim makes no sense.

    ReplyDelete
    Replies
    1. (1) increasing NGDP by fiscal policy is totally different from attempting to do so by monetary policy.

      Monetary NGDPT would be just as useless as QE: indeed it is difficult to see any substantive difference between the 2 ideas.

      (2) "BTW: ABCT would likely make one opposed to monetary policy rather than fiscal policy so that claim makes no sense."

      Yes, you are right there.

      But you would be surprised how many internet Austrians do not believe that. I was just pointing to those links for the sake of showing the general problems with ABCT.

      Delete
  8. Any wagers on how soon (if not already) we will hear the first use of the phrase "that's quadrillion(s) with a Q" from a member of the press?

    ReplyDelete
  9. Asset exchange is the perfect vehicle to tax to distribute wealth to those how actually make things and supply useful services.

    ReplyDelete
  10. Hi, LK. Good post.

    Say: "...for we do not consume money, and it is not sought after in ordinary cases to conceal it"

    "In ordinary cases". Things were different then -- in Say's time, and Adam Smith's -- than things are now. Their time was "the greatest age of the inducement to invest" as Maynard said. I see the rise and fall of civilization as one gigantic business cycle, and Maynard's "greatest age" a happy moment in the cycle.

    It's not really so much the "stupidity of the assumptions behind Say’s law" as it is that the economy is different now.

    ReplyDelete
    Replies
    1. The Arthurian@February 23, 2013 at 12:54 AM

      But secondary financial and real asset markets existed in Adam Smith's day and in J.B. Say's day.

      They still had business cycles in those days, perhaps a combination of agricultural cycles and the modern business cycle (whether industrial, ones set off by credit cycles, or collapse of asset bubbles).

      Delete
  11. Rob Rawlings said:

    "Keynesian counter-cyclical fiscal policy works by leaving the money supply unchanged and redistributing it between those who save and those who spend. In a recession govt borrowing is used to tap unused saving and divert it to the govt who will spend."

    Your comment is Factually Incorrect. Government deficit spending adds to savings.

    ReplyDelete
  12. Thank you, thuglife.

    Great post Lord Keynes...

    ReplyDelete