What it is
Leverage is the use of borrowed capital to amplify the exposure of an investment beyond what the investor's own equity could support. A 2:1 leverage ratio means that for every dollar of equity, one dollar of debt has also been deployed — controlling $2 of assets with $1 of capital.
Leverage amplifies both gains and losses proportionally. A 10% gain on the underlying becomes a 20% gain on equity at 2:1 leverage. A 10% loss on the underlying becomes a 20% loss on equity. This asymmetry becomes critical at the wipeout threshold: the loss level at which all equity is extinguished.
The math of ruin
With 2:1 leverage, a 50% decline in the underlying asset eliminates all equity. With 4:1 leverage, a 25% decline is enough. With 10:1 leverage — common in some hedge fund strategies and embedded in many bank balance sheets — a 10% adverse move is enough to wipe out all equity.
The wipeout threshold is: 1 / leverage ratio. At 2× leverage, it is 50%. At 5× leverage, it is 20%. At 20× leverage (common for investment banks before 2008), it is 5%.
What makes leverage dangerous is not just the mathematics but the timing constraint. An unlevered investor who sees a 50% decline can wait for recovery — their loss is unrealized until they sell. A levered investor who has borrowed against the asset will face a margin call — a requirement to post additional collateral or reduce the position — before the asset recovers. They are forced to realize the loss at the worst moment.
Long-Term Capital Management 1998
LTCM was a hedge fund that employed leverage of approximately 25:1 at its peak. Its strategies involved convergence trades — pairs of securities that were mispriced relative to each other and should converge over time. The trades were nearly certain to be profitable eventually. But "eventually" is a long time when you are 25:1 leveraged.
In the crisis following the Russian debt default in August 1998, the securities LTCM held declined together rather than converging — correlations that were assumed to be independent became synchronized. A fund with $4.7 billion in equity controlling $125 billion in assets faced mounting losses. By September, the equity was nearly gone. The Federal Reserve organized a bailout by major banks because an uncontrolled unwind of LTCM's positions would have destabilized global markets.
The lesson: the mathematics of the strategy were correct — the trades were eventually profitable. But LTCM ran out of equity before "eventually" arrived.
The psychology of leverage
Leverage strategies feel low-risk when they work. A strategy that returns 20% per year with 5:1 leverage on a 4% gross return looks very different from a strategy that bets on volatile assets directly. The return per unit of gross risk can be small; the return per unit of equity is large. This creates an illusion that the strategy is conservative.
The Kelly criterion provides the framework for correct leverage. The optimal leverage is determined by the expected return and variance of the underlying strategy. Overleveraging beyond the Kelly fraction destroys geometric expected return — you make money on average but go bankrupt eventually with probability one. The Kelly criterion says: never bet so much that a run of bad outcomes destroys all capital.
One thing most people get wrong
Leverage means you cannot always wait out a bad period. Unlevered investors have the luxury of patience. "My thesis is correct, I just need to wait" is available to an investor with no borrowed capital. It is not available to an investor whose lender will issue a margin call at a specified loss threshold.
This is why leverage transforms a question of "am I right?" into a question of "am I right quickly enough?" A correct view that takes three years to play out may be entirely irrelevant to a levered investor who is forced to unwind at six months. The most dangerous leverage scenarios are exactly those where the trade is correct but the timing is uncertain. LTCM was right; it just ran out of time. The equity is gone regardless.