4. Forced liquidation is a different risk from loss
Risk 4: Leverage
So far leverage has been arithmetic: it multiplies the movement, subtracts the interest, charges the drag. All of that is serious and all of it is measurable. This module is about the risk that is not a larger version of loss but something else altogether: the one that takes away your ability to stay in the game.
Holding on is an asset, and you only have it if you owe nothing
With no debt, a 50% fall is devastating: it takes 100% to get back to where you started. But as long as you do not sell, that loss is an accounting entry and the position is still yours. You can be wrong about the timing and still be right on the substance.
That ability to wait is a real asset that appears on no statement. Leverage is bought by selling it.
The maintenance margin, in numbers
An investor buys 100,000 euros of shares with 50,000 of his own and 50,000 borrowed: multiple 2. His broker requires a maintenance margin of 30%, that is, that his equity never falls below 30% of the market value of the position.
His equity, at any moment, is the value of the portfolio minus the 50,000 of debt. The condition holds as long as that equity is at least 30% of the value: 50,000 divided by 0.70 gives 71,429 euros. Below that level, he is in breach.
Translated: a fall of 28.6% triggers the margin call. At that point he has lost 28,571 euros on his 50,000, 57% of his capital, and that is when he is required to put up money within hours or has his position sold. The 50% fall that would wipe out his capital never even comes about: they take him out long before.
The general rule is this: with a multiple of *k* and a maintenance margin of *m*, the fall that triggers the call is one minus the ratio of (one minus one divided by *k*) to (one minus *m*). With a multiple of 3 and a margin of 25%, it comes to 11.1%. Eleven points. That is the entire margin for error a multiple of 3 buys you.
Why the requirement tightens exactly when it should not
Here comes the second blow, the one almost nobody sees coming. Margins are not constant: clearing houses and brokers calculate them from recent volatility, and when the market turns turbulent the requirement goes up.
Two things then happen at once and in the same direction: the value of your position falls and the minimum percentage demanded of you on it rises. An investor who has not lifted a finger goes from complying comfortably to being in breach. In March 2020 the clearing houses raised margins across the board within days, exactly while assets were falling.
Why liquidations cluster and amplify the move
The leveraged positions of thousands of investors share a structure: similar multiples, margins and assets. Their liquidation thresholds are therefore very close to one another. The sequence is mechanical:
- The price falls and a first group reaches its threshold.
- That group sells with no choice of price or moment, because the seller is not whoever wants to sell but whoever has to.
- The forced selling pushes the price lower, reaching the next group's threshold.
- Volatility rises, margins tighten and the threshold closes in on those who were still holding on.
The result is that falls with accumulated leverage are not orderly: they are cascades. The same piece of news produces a 5% decline in a market with no debt and a 20% collapse in one loaded with financed positions. Leverage does not only amplify your result: it amplifies the movement of the market you are trading in.
Being right and not being able to wait amounts to being wrong
Ana invests 40,000 euros with no debt. Bruno invests 40,000 of his own plus 40,000 borrowed on the same index. The index falls 35% and eight months later has fully recovered its previous level.
Ana ends up where she started, with a bad memory. Bruno, with the index at its low, had 80,000 of exposure turned into 52,000 and a debt of 40,000: 12,000 euros of capital, 70% lost. He was liquidated there. The subsequent recovery did happen, it was complete and it was not his.
The two did the same analysis and both were right on the substance. One collected and the other did not, and the difference lay not in the analysis but in who had the right to decide when the position was closed. The line usually attributed to Keynes, that the market can stay irrational longer than you can stay solvent, describes this: solvency is a deadline, and leverage shortens it.
That is why, in risk management, the right question to ask of a financed position is not how much can I lose, but how much can I lose without losing the right to decide.
Case study: Long-Term Capital Management, 1998
Long-Term Capital Management, founded by John Meriwether and with Myron Scholes and Robert Merton (awarded the Nobel Prize in Economics in 1997) on its board, did convergence trades above all: buy the relatively cheap asset, sell the expensive one and wait for the gap to narrow. Minuscule margins that only made money with a large multiple. At the start of 1998 its balance sheet was around 125 billion dollars on a capital of some 4.8 billion: about twenty-five times.
In August 1998 Russia defaulted on its domestic debt and devalued. The move was the opposite of the one the models expected and, above all, it was simultaneous across positions that were supposed to be barely related. The fund lost close to 90% of its capital in weeks. On September 23, a consortium of large banks coordinated by the Federal Reserve Bank of New York injected some 3.6 billion dollars to unwind the positions in an orderly way, for fear of the cascade described above.
The detail Roger Lowenstein highlights in his account of the episode closes this level: a good part of those trades converged afterwards, exactly as the analysis had predicted. The analysis did not fail. The deadline did. They were right and they could not wait, and in leveraged markets that is indistinguishable from being wrong.
None of this is a recommendation about what multiple to use. It is education: a mechanism worth understanding before somebody applies it to your portfolio in your name.