← How an Economy Grows and Why It Crashes
第 2 章 · 22 分钟
Credit: transfer of existing saving, or creation from nothing
The load-bearing wall of every conclusion in the book.
概念地图
Capital structure
Production stages ranked by distance from consumption, their proportions coordinated by the interest rate.
深色为本章概念，浅色为其他章
陪练模式
不想只是读？让它带你走一遍。
导师把这一章拆成小步，每一步先问你，再讲；你用自己的话答，它先认下对的部分再纠偏。用你自己的 AI，对话只存在这台设备上。
Locus01
延伸阅读
Wicksell · turn of the century · the cumulative process
Argued that a bank rate below the natural rate keeps pushing prices up, against the older view of interest as merely the outcome of bargaining. The cost: the natural rate is unobservable, and every argument that has invoked it since inherits that hole.
Foundational
Mises · 1912 · The Theory of Money and Credit
Split credit into the kind backed by real saving and the kind that is not, with only the second producing cycles. This gave the book its key distinction and planted this chapter's dispute: the two kinds are indistinguishable on a bank's books.
Turn
The endogenous-money line · second half of the twentieth century
Holds that lending comes first and reserves follow: loans create deposits. It targets exactly the intermediary picture in which banks pass existing saving along, and central bank explanatory material in recent years has adopted the same formulation.
Counterexample
Schiff · 2010 · chapters on the fish-storage bank
Uses a warehouse of stored fish as the bank, so credit is written entirely as a physical transfer. Clear, at the cost of leaving the point above out of the book altogether.
Subject of this course
机制02
Moving the stock
Someone has stored fish and no wish to build; someone wants to build and has no stored fish. Hand the first person's fish to the second and both are better off. That is the book's loanable funds market. Note what it moves: an existing stock. The island's total fish does not rise because of the trade. Here the interest rate is the price of the move, set jointly by how much fish is stored and how eager people are to build.
机制Real saving, through the price called the interest rate, is moved to whoever will produce roundaboutly.
可迁移性测试
Move it to a lab: an idle sequencer lent to another group. Total machine hours are unchanged, only the user is; the fee stands in for the interest rate. The isomorphism breaks in that machine hours cannot serve two groups at once, whereas a deposit is spendable by its holder while lent out.
机制03
Credit with nothing behind it
If nobody actually ate less when the loan was made, the extra purchasing power has to come from somewhere. Mises called this credit unbacked by real saving, and the book takes it as the origin of crises: the loan starts a longer net, and the island does not hold enough stored fish to feed that stretch. What the mechanism lands on is not inflation but a capital structure stretched past what the physical stock can carry. Everything in chapter 6 follows from this one line.
机制Credit without real saving behind it, through a lowered rate signal, stretches the roundaboutness of production.
可迁移性测试
Move it to a company: management rolls short-term paper to fund a five-year retooling. The rate looks low, the work starts faster than cash flow can carry, and when the paper market tightens, a half-built line is a pure loss. The isomorphism breaks in that a company simply goes bankrupt, while an economy's clearing gets interrupted by policy.
Derivation04
From 'loans create deposits' to 'who bore the waiting'
What is being derived: even granting that lending comes first and deposits follow, the physical constraint the book cares about has not gone anywhere.
A bank can lend with no matching deposit in place
Chosen, and also an accurate description of the current monetary system. The book assumes the opposite; this course grants it.
Real output at a point in time is given
Chosen, and the load-bearing one here. Relax it — let idle capacity be drawn in — and this passage collapses; chapter 3 handles that.
The borrower spends the loan on physical inputs that exist now
Chosen: it assumes the loan does not sit on the books. It fails when the loan buys financial assets instead.
推导 · 0 / 4
Worked derivation05
Two rates, two detour lengths
Real saving on the island covers a 100-working-day build. If the rate is pushed down to reflect only the waiting cost of 200 days, how long are the projects that get started?
Compute the length the signal supports, then the length the physical stock supports; the gap is the difference.
裂缝06
Cracks in this framework
争议地形07
本书主张
A bank is essentially an intermediary for saving. Credit without real saving behind it is not created wealth but a transfer away from existing holders, paid for with a false signal about time.
另一种看法
The endogenous-money line holds that banks lend first, deposits are created by the loan, and reserves are supplied afterwards by the central bank. On that account the picture of lending as prior saving fails at the institutional level.
分歧扎在
The disagreement is rooted in whether the physical constraint is the same thing as the monetary constraint. The book treats them as one; the other side separates them: money can expand endogenously, physical output cannot. Note that both accept a short-run ceiling on real output; the dispute is whether the banking system imposes it automatically.
什么证据能裁决
What would settle it is the path of physical input prices after a credit expansion: if new credit goes mainly into new projects, engineering labour and materials should rise in relative price within 2–6 quarters; if it goes mainly into existing assets, asset prices should rise while input prices do not. Split lending by use and the two groups should point in different directions.
The evidence points to 'it depends on where the credit goes' — data broken out by use show both patterns. What is missing is a classification of loan purpose both sides accept: existing classifications follow the borrower's industry, and the same industry borrows both to buy land and to build plant.
Falsification08
Credit with nothing behind it, as a preregistrable test
The mechanism under test: credit without real saving behind it, through a lowered rate signal, stretches the roundaboutness of production.
Data source: the credit-to-GDP gap from macroprudential databases, with sectoral fixed investment and construction cost indices
Public and regularly revised; whether initial or final vintages decide the test must be declared in advance.
Sample: economies with twenty unbroken years of sectoral investment data, full sample, crisis countries included
Excluding crisis countries turns the result into a description of calm periods; sealed off in advance.
Window: quarters 2 to 8 after the credit-to-GDP gap turns up
Under two quarters and no project has started; over eight and the next policy round has covered it.
Threshold: investment growth in sectors far from final consumption (mining, heavy equipment) minus growth in consumer-goods sectors is significantly positive
A difference rather than a level, to net out the common cycle.
Failure condition: the difference is insignificant, wrong-signed, or significant only after splitting by whether a crisis later occurred
The third counts as failure: the result would rest on hindsight grouping, not on the mechanism.
推导 · 0 / 3
接口09
Which slot it hangs on
挂在哪个槽位The conclusion you already hold is probably 'borrow more when rates are low' — a judgement about financing cost. This chapter goes at its premise, not its direction.
Does this chapter (a) replace 'borrow more when rates are low', or (b) constrain it by asking first whether that rate reflects real saving or policy?
慢变量Register three slow variables: credit growth in your industry, the price of far-end inputs (equipment, engineering labour), and the average payback period of projects. Look at the third one first in a few days — its lengthening tracks the detour being stretched.
小结10
本章小结
01'Loans create deposits' is a bookkeeping identity, not the substantive split between the camps.
02The split is over what the loan buys once it exists: physical inputs or existing assets.
03Physical inputs give crowding out and a tilt toward far-end sectors; existing assets give asset price rises with the structure unchanged.
04Whether a given loan has real saving behind it cannot be read off the books; that distinction is unfalsifiable.
05A preregistrable test can verify the structural tilt, not whether the tilt is misallocation.
提取练习 · 合上书，先自己答一遍。
?Is 'bank assets and liabilities rise together' (a) an identity or (b) an empirical finding about bank behaviour?
?If asset prices rise during a credit boom while construction costs do not, does that weaken (a) the second assumption or (b) the third?
Forced choice11
B · Identify the tag
'Making a loan records an equal deposit' — what kind of step is that?
二选一
Forced choice12
D · Judge the interface
Which one hangs on a slot in your existing structure?
二选一
读到这里，把它记为已读。