When Genius Failed Ch. 3: Why They Levered Anyway — Incentives, Not Ignorance

Investment Background

Chapter 2 explained the mechanism of the failure. This chapter answers a more uncomfortable question.

Scholes and Merton won the Nobel Prize for work on pricing risk. Meriwether had run bond arbitrage for twenty years, through 1987 and through the 1994 bond massacre.

They understood the dangers of leverage better than you do.

So why did they lever 25 to 1?

If the answer were "they were careless," this chapter would not be worth writing. The real answer is that a fully rational incentive structure pointed, at every single link, toward more leverage.

And every link in that structure has a counterpart in your own life.

The Wall Street Translation

Force One: The Edge Was Decaying

The convergence arbitrage described in Chapter 1 had a fatal commercial property: it was public.

The spread was observable. Any investment bank's proprietary desk, simply watching a screen, could spot the gap between on-the-run and off-the-run bonds. And three consecutive years of 40% returns announced to the world that there was money here.

So everyone came. Salomon, Goldman, Morgan, and a wave of new funds imitating LTCM all began putting on the same trades.

The spreads narrowed.

This is exactly what arbitrage is supposed to do — the market became more efficient. But for LTCM it meant one concrete thing:

The same strategy now made less money.

The drop to 17% in 1997 was the direct result. Seventeen percent is a fine year for any fund, but against the previous two years at 40% it read like decline.

Now: if you run a fund famous for 40% returns and your strategy will only produce 17%, you have a few options.

1. Accept the lower return and explain it to investors.
2. Find new strategies you are less expert in.
3. Apply more leverage to the same strategy.

LTCM chose the second and the third.

This is the most important mechanism in the chapter. Remember it:

When a strategy's edge decays, leverage is the easiest way to keep the return number intact. It is simultaneously the way that raises risk the most.

Force Two: The Fee Structure

LTCM charged 2% management and 25% performance — above the then-standard two and twenty.

Look carefully at the asymmetry.

If the fund makes $1 billion, the partners take $250 million.

If the fund loses $1 billion, the partners take zero — and owe nothing.

This is an option-shaped compensation structure. The partners own a quarter of the upside without bearing symmetric downside.

And a fundamental property of an option is: the higher the volatility, the more the option is worth.

Option Volatility and Pricing in this library covers this in detail. Scholes and Merton discovered the property.

Put plainly: the two people on earth who understood option pricing best sat inside a compensation structure that was itself an option. And raising leverage raises that option's volatility — which is to say, its value.

One thing must be said in fairness here: LTCM's partners had the overwhelming majority of their personal wealth in the fund. They were not gambling with other people's money while insulated — they lost very nearly everything. This is not a story of fraud.

Which makes the lesson harsher: even with interests tightly aligned, even with your own money inside, the incentive structure still pushes you toward more leverage.

Force Three: Cheap Money

The third force came from outside, and it is the easiest to overlook.

Banks competed to lend to them, on almost no collateral.

Normally, borrowing twenty-five times your capital to buy bonds means the lender demands a substantial margin buffer. But LTCM's reputation was such that initial margin on many trades approached zero.

Why would banks do this?

Because they wanted LTCM's other business. LTCM was one of Wall Street's largest trading clients, and the commissions were enormous. To keep the client, banks competed by loosening financing terms.

An anomalous loop appears here:

• Better reputation → looser borrowing terms
• Looser borrowing terms → higher achievable leverage
• Higher leverage → better return numbers
• Better return numbers → better reputation

Nowhere in that loop is there a link that automatically produces the signal "we should stop."

It has only one termination mode: external shock.

A Detail That Must Be Faced Honestly

In late 1997, LTCM did something that turned out to matter enormously:

They returned roughly $2.7 billion of capital to investors.

The reasoning was sound: good opportunities had thinned, and managing excess capital would drag returns down. On its face this is a responsible act.

But note the actual effect.

Returning capital shrank the denominator. Position sizes were not reduced proportionally.

Leverage went up.

On the eve of the collapse, LTCM was more levered than in its golden year of 1995.

There was no malice in this step, and it was not a blunder. Under the goal of "maintain high returns," it was a perfectly logical decision.

And it moved the fund to a more fragile position.

Executable Trading Rules

The rules here are about spotting the same forces operating on you.

1. When your strategy's returns fall, be suspicious of the urge to size up. This is the road LTCM walked. Falling returns usually mean the edge is decaying, and levering a decaying edge turns a small problem into a large one. The correct response is to shrink, not magnify.

2. Examine the compensation structure behind every financial product put in front of you. Ask one question: what does the person recommending this lose in the bad outcome? If the answer is "nothing," that product's risk is almost certainly understated systematically. This applies to hedge funds and equally to whoever is selling you an annuity.

3. Treat easy borrowing as a warning sign, not an opportunity. When your broker proactively raises your margin limit, when leveraged-product advertising multiplies, when financing costs are unusually cheap — the market is not rewarding you. These are signs the loop is reinforcing itself.

4. Do not treat "my own money is in it" as a safety guarantee. LTCM's partners had nearly everything at stake and it prevented nothing. Aligned interests solve honesty problems. They do not solve judgment problems.

5. Set yourself a non-negotiable leverage ceiling, and set it while calm. You cannot make this decision during a crisis, because by then you are already inside the loop. For the overwhelming majority of retirement investors, that ceiling should be zero.

Relevance to a Retirement Portfolio

This chapter appears to be about hedge funds, but the mechanism runs at the individual level daily.

The most common form runs like this: someone's savings are behind schedule. Behind schedule means "normal returns are no longer sufficient." So they start hunting higher returns — leveraged ETFs, selling options for premium, concentrating in a few hot stocks.

That is the identical response to LTCM's move after 40% became 17%.

And the result is usually identical too: the largest loss arrives at the moment you can least afford one.

This must be stated with total clarity:

If your retirement savings are insufficient, leverage is not the solution. The real solutions are raising your savings rate, delaying retirement, lowering spending expectations, or some combination. These answers are boring. They also cannot take you to zero.

Retirement Decumulation Mechanics on this site, and several of our calculators, exist so that you can solve this problem with numbers instead of leverage.

LTCM's deepest lesson is not "don't use leverage." It is "don't use leverage, under pressure, to close a gap that leverage did not create."

Chapter 4 covers the most dramatic part of the story: what the market does to you when everyone knows you have to sell.