Genius Isn't Enough: Howard Marks on Leverage, Hubris, and Margin for Error
Howard Marks's 1998 Oaktree memo on Long-Term Capital Management argues that intelligence cannot control the future, and that leverage, overconfidence, and a missing margin for error turn small surprises into permanent losses.
The memo's real target was overconfidence
On October 9, 1998, Howard Marks wrote about Long-Term Capital Management less as a post-mortem than as a warning about what investors do when intelligence starts to feel like control. The memo opens with David Halberstam's description of a belief that "sheer intelligence and rationality could answer and solve everything." Marks says the same belief applied to Long-Term Capital Management, whose founders included respected former Salomon Brothers partners, a former Federal Reserve vice chairman, and two Nobel Prize winners. 1
Marks states his purpose plainly: "My purpose here is not to discuss the facts, although I'll do so briefly, but rather the lessons to be learned." That choice still gives the memo its force. Long-Term's failure was not a strange historical accident reserved for mathematical geniuses. It was a concentrated example of a recurring investment error: confusing a powerful process with a certain outcome.
"Brilliance, like pride, often goes before the fall. Not only is it insufficient to enable those possessing it to control the future, but awe of it can cause people to follow without asking the questions they should and without reserving enough for the rainy day that inevitably comes." 1
That is the excerpt worth carrying forward. It contains two separate warnings. The first is about the limits of the brilliant people making the decisions. The second is about the people around them, who stop asking questions because the reputation of the decision-maker feels like a substitute for understanding the risk.
A small move can become a terminal event
Long-Term pursued bond arbitrage. It bought bonds it considered undervalued and sold short bonds it considered overvalued, expecting related prices to converge. The strategy appeared to produce consistent returns, but the capital structure underneath it did the real work. Marks writes that about $4.6 billion of equity supported roughly $150 billion of investments, while the fund's long and short derivatives positions were believed to have an aggregate notional value of $1.25 trillion. 1
The arithmetic did not require a spectacular forecasting error. Marks says that when assets exceed 25 times equity, even a 4% price decline is enough to wipe out the equity. His compact formulation is the memo's most durable line:
"volatility + leverage = dynamite." 1
Leverage is not merely debt in Marks's explanation. It is the mechanism that magnifies a change in the top line by the time it reaches the bottom line. It increases the reward when the position works, but it also turns an ordinary deviation into a capital event. The danger is structural, not rhetorical.
Probability is not a promise
The second lesson is more subtle. Long-Term's models were built around relationships that were expected to converge. But a historical relationship is not a law of nature. Marks writes that when two assets have less than a 100% probability of moving together, the trade contains basis risk. The bonds' yields diverged when they were supposed to converge, and the historic relationship proved less dependable than expected. 1
Earlier in the memo, Marks quotes his friend Bruce Newberg: "there can be a big difference between probability and outcome." That sentence is a useful test for any investment thesis. A probability estimate can organize uncertainty, but it cannot remove it. The more capital a strategy commits on the assumption that a relationship will hold, the more important it becomes to ask what happens when it does not.
Marks's related warning is deliberately ordinary: "It's always something." Any plan that depends on everything going right is unsafe. The point is not that investors should become paralyzed. It is that a sound position must survive an event the model did not make central.
The philosophy at the end of the memo
The closing section turns the LTCM episode into Oaktree's broader discipline. Marks writes that investors do not need to see the future to invest intelligently. Knowledge of the past can take them much of the way. His instruction is blunt: "Forget forecasting" and remember the lessons of history. 1
The memo closes with four operating axioms:
- We cannot know everything about the future, especially when the question asks about the "bigger picture."
- We must expect something to go wrong and build in a margin for error.
- We should recognize risk when greed dominates, and take advantage of bargains when fear goes too far.
- We must keep reminding ourselves of our limitations, because hubris can make even a valid method and talented team fail. 1
This is why the title matters. "Genius isn't enough" does not mean skill is irrelevant. It means skill has to be paired with humility, position sizing, and room for error. A strong investor is not the person who can explain every moving part. It is the person who knows which parts remain unknowable, and refuses to bet the portfolio as if they were certain.
References
- 1Howard Marks, "Genius Isn't Enough" (October 9, 1998)
oaktreecapital.com
This story was produced automatically by a channel. One sentence is all it takes for Neodrop to keep producing for you.