Money & Finance › Banks & credit › How banks create money: lending and fractional reserves
Leverage: amplifying gains and losses with debt
· Concept
Fact, law or dataA rule, statute, definition or measured number.
Leverage means using borrowed money to control more assets than the owner's capital alone could buy, magnifying both gains and losses.
See it on the Money & Finance map →How it works
Buy a $1 million building with $250,000 down and a $750,000 loan. A 10% rise in value is a 40% gain on the equity; a 10% fall is a 40% loss, and a 25% fall wipes it out. Banks run about $10 of assets per $1 of equity. Before 2008, big investment banks ran about 30 to 1, so a drop of roughly 3% could erase them. Stock buyers can borrow up to half the price on margin.
Why it matters
It drives returns in real estate and private equity, and turns small price drops into failures, as at LTCM in 1998 and across Wall Street in 2008.
Think of it like
Loan-to-value is leverage: a 75% LTV loan means $4 of property for every $1 of equity.
Facts, rules and data
Lehman's bankruptcy examiner found it ran about 31-to-1 leverage at end-2007. The Fed's Regulation T sets the 50% initial margin; bank call reports show capital ratios.
Debates and risks
How much leverage the whole financial system carries is unclear, since hedge funds and private funds disclose little.
Connections (9)
▸Basel capital rules (Basel I, II, III) limit Leverage: amplifying gains and losses with debtFact, law or data
The connection
Risk weights let banks hold little capital against assets labeled safe, like AAA mortgage bonds before 2008. Basel III added a blunt backstop: Tier 1 capital of at least 3% of total exposures whatever their risk, capping leverage at about 33 to 1. The largest US banks face a stricter 5% standard, which regulators moved to ease in 2025.
When and why
Agreed in 2010 and a binding international minimum from 2018; US banks have had a leverage ratio of their own since the 1980s.
How we know
The Basel III texts; Lehman's roughly 31-to-1 leverage in 2007 showed what risk-weighted rules alone missed.
Debates and risks
Banks argue the leverage ratio discourages low-risk Treasury market-making; whether easing it trades safety for smoother Treasury markets is debated.
▸Leverage: amplifying gains and losses with debt magnified losses at Long-Term Capital Management collapse (1998)Fact, law or data
The connection
LTCM's trades earned tiny margins, so it borrowed heavily to make them pay: about $125 billion of assets on under $5 billion of capital, plus over $1 trillion in derivatives. When spreads widened after Russia's default, a move of a few percent against it erased most of its equity.
When and why
Greenwich, Connecticut; losses of about $4.6 billion over roughly four months in 1998.
How we know
The President's Working Group report (1999) documented its leverage, and lenders' exposures came out during the rescue.
Debates and risks
Why lenders competing for its business asked so little collateral, and how much leverage hedge funds run today, remain open questions.
▸Leverage: amplifying gains and losses with debt is central to returns in Real estate as an investmentFact, law or data
The connection
Most property is bought with debt, often 50–75% of value, so equity returns hinge on the gap between the cap rate and the loan rate. A 6% cap with debt at 5% boosts the cash yield on equity; debt at 7% cuts it ('negative leverage'). With 65% debt, a 10% value drop erases about 29% of the equity.
When and why
In every property market; vivid in the 1980s Texas bust, the 2008 crash and the post-2022 rate rise that squeezed highly levered office and apartment owners.
How we know
Lenders' loan-to-value and DSCR data, and default studies after 2008 and 2023, show the most levered owners failing first.
Debates and risks
How much of real estate's historical returns come from leverage rather than the property itself, and what level of debt is prudent when rates jump.
▸Leverage: amplifying gains and losses with debt powers buyouts in Private equity and leveraged buyoutsFact, law or data
The connection
In a buyout the PE fund puts in perhaps 40–50% equity and borrows the rest, owed by the company bought. Example: buy a $100 million company with $40 million equity and $60 million debt; sell for $130 million and the equity becomes $70 million, a 75% gain on a 30% rise. A 30% fall would cut the equity 75%.
When and why
Pioneered by Jerome Kohlberg, Henry Kravis and George Roberts at Bear Stearns in the 1960s–70s; in the 1980s debt often covered 80–90% of the price.
How we know
Loan and bond documents disclose deal financing; studies such as Kaplan and Strömberg (2009) document leverage levels and their link to returns.
Debates and risks
How much PE returns come from better management versus leverage and rising markets is debated; some research finds buyout targets go bankrupt more often than similar firms.
▸The capital stack: who gets paid first in a property deal shows how much a deal uses Leverage: amplifying gains and losses with debtFact, law or data
The connection
The stack measures leverage directly: debt layers as a share of value. With a 65% loan, a 20% fall in value wipes out about 57% of the equity; with 80% combined debt, the same fall wipes out all of it. More debt raises returns on equity in good times and risk in bad.
When and why
In every financed property; leverage tends to peak before downturns, as in 2005–07, when some deals stacked debt to 85–90% of value.
How we know
The equity-loss math is arithmetic; CMBS loss data from 2009–12 show losses concentrated in the most highly leveraged deals.
Debates and risks
How much leverage is safe for each property type is judgment; lenders loosen standards in booms and tighten them after busts.
▸Debt service coverage ratio (DSCR) measures how safely a building carries Leverage: amplifying gains and losses with debtFact, law or data
The connection
With NOI fixed, more debt means bigger payments and a lower DSCR. Example: $1 million of NOI; a $7 million loan at 6.5% on 25-year amortization costs about $567,000 a year, 1.76x; a $10 million loan costs about $810,000, 1.23x. The second leaves only a 19% income cushion before payments go uncovered.
When and why
In every income-property loan; lenders set a minimum DSCR as a cap on leverage.
How we know
Standard amortization math; loan agreements and credit policies tie loan size to DSCR.
Debates and risks
Whether DSCR or LTV better guards against default depends on the cycle; after 2022 both failed some lenders.
▸Leverage: amplifying gains and losses with debt fueled the 2015 boom and bust in Shanghai Composite index and China's 2015 stock crashFact, law or data
The connection
In 2014–15 brokers' margin loans reached about 2.27 trillion yuan, and unofficial lenders offered far more leverage, sometimes several times an investor's cash. Rising prices drew more borrowing. When regulators cracked down on grey-market margin in June 2015, margin calls forced sales, which pushed prices down and set off more margin calls.
When and why
Shanghai and Shenzhen, mid-2014 to mid-2015; the CSRC, brokers, trust companies and online lending platforms.
How we know
Official margin-debt statistics from the exchanges, CSRC announcements, and later IMF and academic studies of the margin spiral.
Debates and risks
How much unofficial borrowing there was; estimates vary widely because much of it was off the books.
▸The 2022 UK gilt crisis: the mini-budget and pension LDI funds was amplified by pension funds' Leverage: amplifying gains and losses with debtFact, law or data
The connection
LDI funds held gilts and interest-rate swaps worth several times their cash, often using repo borrowing, so pensions could hedge long liabilities while keeping money in growth assets. When long gilt yields jumped more than a point in days, lenders demanded collateral; funds sold gilts to raise it, pushing yields higher in a self-reinforcing spiral.
When and why
September 23–28, 2022, in UK pooled and single-client LDI funds run by large managers such as Legal & General, BlackRock and Insight Investment.
How we know
Bank of England letters reporting several LDI funds near negative net asset value on September 27, when the 30-year yield rose 67 basis points from its morning low.
Debates and risks
How much leverage was too much is debated; since 2023 UK regulators expect LDI funds to withstand about a 2.5-point jump in yields.
▸China's property crisis and Evergrande's collapse (2021–) began with 2020 'three red lines' curbs on Leverage: amplifying gains and losses with debtFact, law or data
The connection
In August 2020 Beijing set the 'three red lines': liabilities under 70% of assets (excluding pre-sales), net debt under 100% of equity, and cash at least equal to short-term debt. Developers breaching them could not add debt, cutting off the credit that kept them afloat.
When and why
Beijing, August 2020; the People's Bank of China and the housing ministry.
How we know
Official announcements and developers' financial filings.
Debates and risks
Whether the curbs prevented a bigger crash later or triggered this one.