“I
have found out what economics is; it is the science of confusing stocks with
flows”
Michael Kalecki, quoted by Joan Robinson
in Shedding Darkness
An excellent example of this confusion
is how some economists think about inflation – inflation is a flow
characteristic (since prices are expressed via transactions) yet many somehow think
that changes in money stocks automatically produce inflation (e.g. the quantity
theory of money). It is true that changes in stocks of money can influence
money flows but it is incorrect to ASSUME an independent causal relationship. Instead,
the relationship between money stocks and money flows is dependent on numerous
other factors (e.g. money stock distribution, interest rates, animal spirits, etc.)
which would need to be explicitly identified and controlled for if one wanted
to quantify the relationship (this would be a very difficult, perhaps
impossible, econometric task). You cannot simply assume the “velocity” of money
is constant. Inflation can occur from any level of money stocks (think about
what would happen if an asteroid was headed for earth). This leads to a central
point I’ve banged the table about for years, and especially in my Applied
Economic graduate studies – assumptions are of critical importance but which
often do not receive an appropriate amount of attention. I believe that the
root source of most mistakes in many fields of analysis stem from incorrect
assumptions. I attribute my sensitivity to assumptions from my background in
the natural sciences where assumptions are explicitly addressed even though in
many cases they are far less pervasive than in economics (a social science).
This last point is why I initially turned away from economics after my first
few months of college (in the early 90’s), and why I choose to get a second
master’s in Applied Economics with a focus on microeconomics (although
similarly flawed I think mainstream microeconomic assumptions are less
problematic).
Keeping with the assumptions theme, and
putting on hold my economic policy recommendations (which the birth of my third
child and a recent house purchase/move have prevented me from developing), I
will now discuss a key assumption I made in a prior post as well as a CBO
document that is full of bad assumptions.
“I will assume that all financial assets not
backed by a bank or the government cannot function as a money flow. I will
ignore such items.”
I made this assumption because it
simplified the discussion and did not influence the main points. However, this
statement is not 100% correct – lending can occur between non-bank entities which
results in a new financial asset being created that can be used as a money flow
(e.g. accounts receivable (an asset to the owner) in exchange for a real flow).
In addition to this type of transaction there is also “on-lending”, which
entails the non-bank lending of existing money stocks to others (e.g. a deposit
asset in exchange for a loan asset). While different at one level (“on-lent”
transactions use bank/government money while A/R type transactions use money
–like instruments) these processes are similar in that they both require
someone to accept a non-bank IOU (asset) in exchange for a good, service or
asset. The end result of such lending is that debt stocks now exceed money
stocks and that all debt stocks can only be fully extinguished via multiple
money flows through the economic cycle. A debt created from such lending is a
stress on the system in the same way that interest expense is in that no money
is created for its payment but instead payment must occur from existing money
stocks. However on the other hand one can think of non-bank lending as being
similar to lending via regular banks as it has the same short term macroeconomic
consequence – it can result in additional money/financial asset flows that can
increase real flows (economic output), assuming it is spent on good/service
production.
Non-bank lending transactions can be
performed by anyone, but due to trust issues and challenges in finding a
suitable counterparty(s), it is usually facilitated by a non-bank financial
institution (e.g. pension fund buying MBSs from a hedge fund). These non-bank
financial intermediaries are part of the “shadow banking” system – which has
been estimated to account for nearly 25-30% of the entire financial system1.
I am still trying to fully understand the shadow banking system (by reading
articles such as this: http://www.newyorkfed.org/research/epr/2013/0713adri.html
) but can briefly summarize such institutions as follows:
- Shadow
banks consist of – hedge funds, money market funds, SIVs, private equity funds,
finance companies, etc.
o Money
market funds are different than other shadow bank entities – I will ignore them
for now
- They
(shadow banks) do not have an account at the Federal reserve, they do not
have access to Federal Reserve lending facilities, they do not issue
FDIC insured deposits and they are not closely regulated like banks. The
implications of these facts include:
o The
liabilities shadow banks issue are often explicitly tied to assets on their
balance sheets (e.g. “secured” borrowing or borrowing via repurchase-type
contracts)
o An
obligation issued from a shadow bank is not automatically accepted by regular
banks because the lack of an account at the Fed means that the destination bank
cannot receive reserves (an asset to the destination bank) when the transfer is
made, meaning that the shadow bank must have some other asset to give to the
destination bank in order for them to create the liability on their books. This
factor contributes to shadow bank asset withdrawal timing restrictions and
makes shadow bank liabilities (assets to the counterparty) generally unsuitable
for use as “money”.
o Shadow
banks are risker than regular banks since they are much more likely to
experience liquidity-driven problems (as actually occurred during the financial
crisis until the Federal Reserve provided them with a lending facility)
- The
differences between banks and shadow banks are blurred because shadow banks are
often owned by banks, undertake many activities that banks perform and require
banks to facilitate their functioning (e.g. for interbank payments, etc.).
Shadow banking is a complicated subject
because of firm and activity diversity, but in my opinion the reason for the
growth of the shadow banking system can be attributed to banking regulations –
the more we try to limit banking risk the more activities end up being pushed
out of the system into the “shadows”. This is a troubling development,
especially when one realizes that the recent housing boom and subsequent
financial crisis would not have been possible without the shadow banks. This is
a big topic for which I will try to write further. In the meantime, I would
like to point out that the shadow banking system ends up funding housing debt
with deposits (the original accounting entries on the bank’s balance sheet when
the house loan/deposit was created) by a complicated chain (forming a circle
when connected to the source bank) of repo transactions that breaks up the risk
into pieces (e.g. liquidity risk, credit risk, interest rate risk, etc.) and
results in:
1) The
removal of both the asset (loan) and liability (deposit) from the bank’s balance
sheet [This is key because the liquidity, credit and interest rate risk
associated with the accounting of the initial loan (a long term loan paired
with a short term deposit) would require a significant amount of bank capital
to be set aside (lowering bank leverage/profitability). Although this can also
be mitigated by the bank converting the deposit to a different liability (e.g.
term deposit, bank issued bond, etc.)2 ]
2) An
asset swap on the shadow bank system balance sheet (deposit asset (to regular bank)
for loan asset (from regular bank))
3) Multiple
(up to 7 or so) loans between shadow banks (each balance sheet would have an
asset in the form of a loan due to them and a liability in the form of a loan
due to another shadow bank, with the underlying real asset (housing/land)
securing each loan).
An example of bad assumptions
undermining an analysis is the CBO’s “The
2013 Long-term Budget Outlook”. Below I analyze three particularly
problematic passages of this report – it saddens me to think of the policy
mistakes that follow from this incredibly flawed report. Wake up CBO!!!
Page
3:
“Harmful Effects of Large and Growing Debt
How long the nation could sustain such growth
in federal
debt is impossible to predict
with any confidence. At
some point, investors would begin
to doubt the government’s
willingness or ability to pay
U.S. debt obligations,
making it more difficult or more
expensive for the government
to borrow money.
Moreover, even before that
point was reached, the high and
rising amount of debt
that CBO projects under the
extended baseline would
have significant negative
consequences for both the
economy and the federal budget:
- Increased borrowing by the
federal government would
eventually reduce private investment in productive
capital, because the portion of total savings
used to
buy government securities would not be
available to
finance private investment. The
result would be a
smaller stock of capital and lower output and
income
in the long run than would otherwise be the
case.
Despite those reductions, however, the
continued
growth of productivity would make real
(inflation adjusted)
output and income per person higher in
the future than they are now.
- Federal spending on interest
payments would rise,
thus requiring larger changes in tax and
spending
policies to achieve any chosen targets for
budget
deficits and debt.
- The government would have
less flexibility to use
tax and spending policies to respond to
unexpected
challenges, such as economic downturns or
wars.
- The risk of a fiscal
crisis—in which investors
demanded very high interest rates to finance
the
government’s borrowing needs—would increase.”
Page
77:
“How Increased Federal Borrowing Would Affect
the
Economy
Increased borrowing by the
federal government generally
draws money away from (that is, crowds out)
private
investment in productive capital because the
portion of
people’s savings used to buy government
securities is not
available to finance private investment. The
result is a
smaller stock of capital and lower output in
the long run
than would otherwise be the case (all else
held equal).3
Two factors offset part of that crowding-out
effect. One
is that additional federal borrowing tends to
lead
to greater private saving, which increases the
total funds
available to purchase federal debt and finance
private
investment. That response occurs for several
reasons:
- Additional federal borrowing
tends to raise interest
rates, which boosts the return on saving;
- Some people anticipate that
policymakers will raise
taxes or cut spending in the future to cover
the cost of
paying interest on the additional accumulated
debt, so
those people increase their own saving to
prepare for
paying higher taxes or receiving less in
benefits; and
- The policies that give rise to
deficits (such as tax cuts
or increases in government transfer payments)
put
more money in private hands, some of which is
saved.”
Page
81:
“With those two offsets taken together,
when the deficit
goes up by $1, private saving rises by 43
cents (so
national saving falls by 57 cents), and
foreign capital
inflows rise by 24 cents, ultimately leaving a
decline of
33 cents in investment,
according to CBO’s central estimates.
To reflect the wide range of estimates in the
economics
literature on how government borrowing affects
national saving and domestic investment, CBO
produced
estimates of the economic effects of the
budget scenarios
using three assumptions about those effects.
Those
assumptions imply that for each dollar that
deficits rise,
national saving is reduced by 39 cents, 57
cents, or
71 cents, and domestic investment is reduced
by
15 cents, 33 cents, or 50 cents.”
Starting from the top – the first
underlined section discusses investors being unwilling to “finance” the federal
debt. This reminds me of the people who feared a credit downgrade by the rating
agencies as it would increase US borrowing costs. The fact is US borrowing
costs decreased after the credit downgrade because US borrowing costs are under
the control of the Federal Reserve - not “bond market vigilantes”. My prior
posts should make this point abundantly clear. In the worst case scenario the
Fed would buy all the Treasury debt and leave the private sector with reserve
balances (which have zero duration and a yield set by the Federal Reserve),
unless of course you want to argue that the Fed will ignore their mandate and
let the Treasury default.
The second underlined point regarding the
“crowding out” of private investment refers to one of the most problematic
assumptions in economics – specifically the idea that there is a fixed amount
of money and therefore an opportunity cost of using money to “fund” federal
deficits. This is another point which I have discussed at length in prior posts
– the “loanable funds” assumption is totally invalidated when one understands
that “loans create deposits” and “investment creates savings”. There is not a
fixed pool of money that is divided up between competing groups – in the real
world money/loans expand to meet the needs of the economy. This flawed
assumption is central to the IS/LM model, who folks like Paul Krugman use even
though the model’s creator (John Hicks) ended up rejecting it.
The third point that larger interest
payments makes required deficit limiting tax/spending changes larger is not
accurate because the result depends on other factors. One of the main themes
throughout my posts is that the fundamental building block of market-based
economies is transactions, and that transactions REQUIRE at least two parties.
In this case as Federal interest payments increase the other transacting party
(the private sector) see an equivalent increase in interest revenue. It is this
second automatic consequence of increased interest expense that makes the
statement problematic – we need to make assumptions about who receives these
payments, what people do with the interest revenue payments and what
tax/spending laws exist at the time before we can ascertain what the
ramifications will be regarding the overall fiscal stance of the Federal
Government. Higher interest expense can reduce the budget deficit via the
impact of the resulting monetary flows on the existing automatic stabilizers in
place at the time.
The fourth point, about less flexibility
in the event of an adverse economic shock, is just another implication from the
assumption that investors may be unwilling to finance the federal debt. If this
worry is not correct (as I have argued) then there is no reason why high debt
would limit flexibility in regards to tax and spending decisions.
The fifth point, about increased risks
of a financial crisis, is yet another way to state the incorrect assumption
that investors may become unwilling to “finance” the federal debt. It’s really
sad that after a private debt induced financial crisis people still end up
focusing on public debt. “You can’t solve a debt problem with more debt” is the
common meme, but again this idea fails because the analogy represents an
incorrect assumption – private entities (without a money printing press) can
fail to pay debts (and go bankrupt) while governments with a fiat currency and
no foreign currency denominated debt can always pay their debts if they so
choose. Before leaving this point I need to highlight that foreign holdings of
public debt (or US$ financial assets in general) can be problematic in that
quick movements of these funds out of the country can create exchange rate issues
and problems in the banking system (since banks facilitate foreign currency
conversion and can get caught with significant foreign exchange losses – this
sort of scenario could be resolved via capital controls, or other regulations).
However, this is a trade deficit issue, not a federal debt issue (sadly these
issues are often confused).
The sixth point, about federal borrowing
increasing interest rates, is yet another consequence of the incorrect
“loanable funds” assumption. In the past empirical support for this idea was
only possible via omitted variable biased regressions (i.e. by not controlling
for central bank reaction functions and different monetary regimes). Now even
these biased regressions do not tell this story – public debt as a % of GDP has
expanded almost everywhere, yet interest rates have gone down.
The seventh point is one which I have
explicitly discussed in a prior post – it’s the fairytale about Ricardian
equivalence, which apparently is too closely aligned with “inter-temporal optimization”
to discard.
The eighth and final point is a quote
that is actually consistent with the sectoral balance identity ((I - S) + (G -
T) + (X - M) = 0). The problem with the quote is 1) the wide range of
postulated effects from government borrowings (admission that they have no
clue), 2) they did not mention that in their scenario the private sector ends
up with an increase in net financial assets equal to $0.76 (and the foreign
sector with a balance that is $0.24 higher than before) – either because they
ignore balance sheets or because this realization would cast doubt onto their
(incorrect) assumption regarding the effect of deficit spending on interest
rates (i.e. that deficits increase them), 3) they did not mention the deficit’s
impact on short term economic output but instead focused on investment which is
moving in a negative direction due to bad assumptions about loanable funds and
4) they mix in negative foreign sector leakage effects without discussing the
assumption explicitly (and in fact discuss a trade deficit as if it is a good
thing)
I
could go on and on – in addition to a total lack of understanding of banking
and money the report is also full of the standard neoclassical assumptions
(marginalist optimization, etc.) many of which are fundamentally flawed. Why
can’t the CBO at least include a scenario/addendum that is derived from a
different economic school of thought, especially in light of the fact that the
economic school of thought they use has failed so miserably?