Investing · Day 03

Investing Classics: Howard Marks on CyclesMastering the Market Cycle

June 1, 2026·BigCat's Capital Allocator
Howard Marks, founder of Oaktree Capital, has spent the last three decades writing memos to his clients. Buffett has said, "When I see memos from Howard Marks in my mail, they're the first thing I open." Marks's core contribution is not prediction but a deep, sober view of "cycles" and "risk" — he accepts that the future is unknowable, yet shows how to choose better positions inside that unknowability. This week we draw on The Most Important Thing (2011) and Mastering the Market Cycle (2018) for the four weapons Marks leaves to long-term investors.
PRINCIPLE 01

Cycles Always PrevailCycles Always Prevail

Market structure
Statement of the Principle
Markets do not move in straight lines, and they do not rise forever. Economic, credit, psychological, and valuation cycles are nested inside one another and amplify each other. The investor's job is not to escape cycles but to know exactly where in them she currently stands.
Source

Howard Marks, Mastering the Market Cycle (2018), Chapters 1–3; and the memo "You Can't Predict. You Can Prepare." (2001).

"Rule number one: most things will prove to be cyclical. Rule number two: some of the greatest opportunities for gain and loss come when other people forget rule number one." — Howard Marks, The Most Important Thing (2011)
Deeper Reading

Marks decomposes the cycle into four nested layers, each amplifying the last: (1) the economic cycle (GDP, employment, capacity utilization); (2) the profit cycle (operating leverage makes corporate earnings swing far more than GDP); (3) the credit cycle (banks and bond markets "open the window" and slam it shut at both tops and bottoms); (4) the psychology cycle (fear and greed apply an emotional multiplier to the other three). When all four push the same way, markets overshoot in both directions. Marks's key insight: cycles are not clock times but a conditional probability distribution — when valuations are high, credit is loose, and sentiment is euphoric, the entire distribution of future returns shifts leftward, and vice versa. That is why "expensive today" does not mean "crash next year," but "assets not expensive today will, over time, deliver better returns."

— Four nested cycles: Economy → Profits → Credit → Psychology —
Long-term trend Euphoria — top Capitulation — bottom Economy Profits Credit Psychology
Aligned upward → top overshoot Aligned downward → bottom overshoot
Classic Case

US subprime and the GFC, 2006–2009: (1) the economy near a peak; (2) corporate profit margins at a post-WWII high; (3) credit extremely loose (NINJA loans, mis-rated CDOs, leverage above 30x); (4) the belief that "US home prices cannot fall nationally at the same time" hardened into dogma. All four cycles aligned at the extreme, and the unwind was equally extreme — the S&P 500 fell 57% from its October 2007 peak to its March 2009 low, financials more than 80%. In July 2007 Marks wrote "The Race to the Bottom": "When anyone is willing to do any deal at any price, disaster is on the way." Fourteen months later, Lehman fell. He did not predict the timing; he located the position.

Counterexamples and Limits

Common misuses: (1) confusing "recognizing the cycle" with "predicting the turning point" — Marks repeatedly stresses that timing is unknowable; only location is knowable; (2) flipping aggressively the moment any single indicator reads extreme (shorting US equities in 2017, bottom-fishing Lehman in March 2008) and getting carried out before being proven right; (3) applying one cycle framework to every asset — commodities, credit, tech, and real estate have very different cycle lengths and drivers. Marks's posture is adjust the thermostat, do not time the market: dial portfolio risk down when you judge the cycle near a top, dial it up near a bottom, in moves moderate enough to survive being wrong.

Decision Checklist
  • Where in "early / middle / late" is each of the four cycles right now — economic, profit, credit, psychological?
  • Are the four cycles amplifying each other or offsetting? Be most alert when they align.
  • Are recent market returns coming from improved fundamentals or from multiple expansion?
  • Are credit conditions loosening or tightening? Who is suddenly able to get money easily?
  • Is my current positioning preparing for the top, or for the bottom?
English Insight

"We may never know where we're going, but we'd better have a good idea where we are." — Howard Marks

Reflection This Week
For your three largest positions, write down the cycle stage of each underlying industry (early / mid / late / bottoming). If all three are "late cycle," ask whether your sizing reflects that concentration.
PRINCIPLE 02

The PendulumThe Pendulum of Investor Psychology

Market psychology
Statement of the Principle
Investor sentiment swings like a pendulum, passing through the middle but rarely resting there. Knowing which side it is leaning toward right now matters more than predicting exactly how far it will go.
Source

Howard Marks, The Most Important Thing (2011), Chapter 9 "Awareness of the Pendulum." The concept first appeared in the 1991 memo "First Quarter Performance."

"The mood swings of the securities market resemble the movement of a pendulum. Although the midpoint best describes the location of the pendulum 'on average,' it actually spends very little of its time there. Instead, it is almost always swinging toward or away from the extremes of its arc." — Howard Marks, The Most Important Thing (2011)
Deeper Reading

The pendulum has three pairs of extremes: (1) greed vs fear; (2) optimism vs pessimism; (3) risk tolerance vs risk aversion. They are not independent — they usually swing together. Marks's key insight: the pendulum moves fastest through the middle. That is, the moments that "look normal" are often when the market is racing toward an extreme. A gentle rally is often the prelude to a euphoric one. The investor's job is to recognize the extremes, not the midpoint. Signs of an extreme: brainless issuance (IPOs many times oversubscribed), brainless product (leveraged ETFs at new highs), brainless participants (taxi drivers discussing stocks), unanchored valuation ("this time is different" stories replacing DCF). At the extreme, capital is forced into worse and worse projects and future returns must be poor. The reverse extreme is the source of opportunity.

— Investor Psychology Pendulum —
Greed Fear · Pessimism "It will never go up again" Midpoint — rarely lingers Greed · Optimism "This time is different" ← Extremely cheap Extremely expensive →
Bottom signalsindiscriminate selling, gloomy headlines, fundraising drought
Top signalshot IPOs, rising leverage, "this time is different"
Classic Case

Nasdaq 1999 vs 2002: in 1999 the pendulum was pinned to the right — day traders quit their jobs to trade full time, unprofitable Pets.com IPO'd at $290M, Cisco briefly became the world's most valuable company, median P/E hit 60x. In January 2000 Marks wrote his famous memo "bubble.com" — those who saw it found it obvious; those who didn't, never would. Two years later the pendulum hit the left extreme: Nasdaq fell from 5,048 to 1,114 (-78%), Amazon -95%, Apple -75%, many analysts swore "never tech again." An investor who bought Amazon at the 2002 left extreme made roughly 200x over the next 22 years. Same asset, two different points in time, wildly different returns — the difference was not the company but the pendulum's position.

Counterexamples and Limits

Common misuses: (1) treating "pendulum in the middle" as "no information," and consequently holding too much risk in the middle; (2) using the pendulum alone for large contrarian trades — pendulums can sit at an extreme for years (Nasdaq 1998–2000 stayed at the right extreme and still doubled); (3) lumping all assets into one pendulum — US equities, China A-shares, credit, and commodities each have their own. Better practice: use pendulum position as a sizing signal, not a timing signal. At the extremes, shift positioning by a moderate 10–20%, not all in or all out. Marks's own example: he started reducing credit risk in 2005–2006, missed the last year of the rally, but preserved the ammunition that bought 2008–09.

Decision Checklist
  • On a 1–10 scale (1 = extreme fear, 10 = extreme greed), where is market sentiment right now?
  • How has that score moved over the past 12 months, and in which direction?
  • Which "brainless" signals are flashing — leveraged ETFs, IPOs, retail margin, headlines?
  • Is my own emotional baseline aligned with the pendulum? (Alignment is the danger zone.)
  • If the pendulum returned to the midpoint over the next 12 months, would my portfolio gain or lose?
English Insight

"In the world of investing, nothing is as dependable as cycles. Fundamentals, psychology, prices and returns will rise and fall, presenting opportunities to make mistakes or to profit from the mistakes of others." — Howard Marks

Reflection This Week
Cast your mind back to March 2020, October 2022, and Q4 2024. Recall your emotional state and positioning at each. When did you swing with the pendulum? When did you swing against it? Is your "counter-pendulum muscle" getting stronger or weaker?
PRINCIPLE 03

Understanding & Controlling RiskUnderstanding & Controlling Risk

Risk management
Statement of the Principle
Risk is not volatility; it is the probability of permanent loss of capital. It is invisible beforehand and reveals itself only once, after the fact. Superior investing is not about high return — it is about return per unit of risk.
Source

Howard Marks, The Most Important Thing (2011), Chapters 5–7: "Understanding Risk / Recognizing Risk / Controlling Risk."

"Risk means more things can happen than will happen. Investing consists of exactly one thing: dealing with the future. And because none of us can know the future with certainty, risk is inescapable. Thus, dealing with risk is an essential — I think the essential — element in investing." — Howard Marks, The Most Important Thing (2011)
Deeper Reading

Marks rejects the academic equation of risk with standard deviation. Volatility is just price jitter; what actually kills compounding is permanent loss of capital. A stock that fell 90% and one that fell 30% and rebounded may have similar historical volatility, yet the damage to long-term compounding is orders of magnitude apart. He frames risk in three layers: (1) recognizing risk — "when valuations are high, credit is loose, and sentiment is hot, risk is quietly rising even as prices keep climbing"; (2) measuring risk — never infer it backwards from returns; assess the probability and depth of loss scenarios ex ante; (3) controlling risk — through entry price, diversification, and structural protection (seniority, covenants). His deepest line: "High returns come from taking risk, but taking risk does not necessarily produce high returns."

— Loss-recovery asymmetry · the price of permanent loss —
Lose 10%
need +11%
Lose 25%
need +33%
Lose 50%
need +100%
Lose 75%
need +300%
Lose 90%
need +900%
Lose 95%
need +1900%
Controlling the downside costs far less than chasing the upside. Compounding hates deep holes.
Classic Case

The collapse of Silicon Valley Bank in March 2023: on paper, SVB's assets (long-duration Treasuries and MBS) carried almost no credit risk, and traditional risk models showed very low VaR. But under the Marks framework, the real risks were: (1) deposit concentration (a homogeneous base of tech startups); (2) duration mismatch (ten-year bonds funded by short-dated callable deposits); (3) psychological risk (a bank run accelerated by social media). Once all three lit at once, hundreds of billions in deposits evaporated in 48 hours. It confirms Marks's point — risk is invisible beforehand because prices look calm, and only shows up afterwards, but it only shows up once. Real risk management means examining positions that look "calm in price but fragile in structure," not staring at historical volatility.

Counterexamples and Limits

Common misuses: (1) confusing "low volatility" with "low risk" — long Treasuries 2020–2022 had modest historical sigma but duration risk wiped out 30%+ once it triggered; (2) using VaR-style models for false reassurance — LTCM's 1998 VaR said daily losses were capped near $50M; actual single-day losses ran to $550M; (3) "diversifying" into assets that turn out to be highly correlated (almost everything tied to real estate fell together in 2008); (4) becoming so risk-averse you never bet — missing the compounding entirely. Marks's resolution: risk is not to be eliminated, it is to be priced — only take it when you are amply paid for it.

Decision Checklist
  • In the worst 10% scenario, what is the permanent loss on this position? Can I bear it?
  • What is on my "invisible-risk" list — liquidity, correlation, leverage, single-point dependence?
  • Is historical volatility under-stating the real risk? (Especially for "quiet for a long time" assets.)
  • If a lower-risk alternative offers the same return, why am I not switching?
  • Am I taking risk because the market is paying for it — or because I am being driven by my own optimism?
English Insight

"There are old investors, and there are bold investors, but there are no old bold investors." — Howard Marks

Reflection This Week
For each holding in your portfolio, write a single line describing the "permanent loss scenario" — not a drawdown, but the road of no return. If you cannot write that line, you may not yet truly understand the risk of that investment.
PRINCIPLE 04

Second-Level ThinkingSecond-Level Thinking

Cognitive edge
Statement of the Principle
To beat the market your thinking must be one layer deeper than the market's. First-level thinking asks "is this company good or bad?" Second-level thinking asks "what does the market already expect, how much is in the price, and is my view still right outside the consensus?"
Source

Howard Marks, The Most Important Thing (2011), Chapter 1 "Second-Level Thinking."

"First-level thinking says, 'It's a good company; let's buy the stock.' Second-level thinking says, 'It's a good company, but everyone thinks it's a great company, and it's not. So the stock's overrated and overpriced; let's sell.' First-level thinkers think the same way other first-level thinkers do about the same things, and they generally reach the same conclusions. By definition, this can't be the route to above-average results." — Howard Marks, The Most Important Thing (2011)
Deeper Reading

Market prices are a snapshot of consensus expectations. To earn excess returns, two conditions must hold at once: (1) your view of the future differs from the consensus; and (2) you are right and they are wrong — variant perception that turns out to be correct. Marks calls this finding mispricings. First-level thinking is recognizable by: (1) judging by intuition and surfaces ("AI is the trend, buy"); (2) drawing on the same information as everyone else; (3) reaching easily reproducible conclusions. Second-level thinking is recognizable by: (1) constantly asking "who else knows this? how much is already priced?"; (2) separating fundamental expectations from sentiment expectations; (3) actively constructing the counter-thesis. Marks warns: second-level thinking is not "contrarianism" — mechanically opposing the crowd is not edge. Second-level is being right in a differentiated way. Undifferentiated agreement (buying what everyone likes) and undifferentiated disagreement (selling what everyone likes) are both edgeless.

First-Level
Second-Level
It's a good company — buy.
The market already prices it as flawless — there is no margin for imperfection. How does my view differ from consensus?
The economy is bad — sell.
"Bad economy" is the consensus — how deeply is that already in the price? Selling means I am forecasting "even worse" more accurately than the market.
The company missed earnings — it will fall.
Was the miss already pre-adjusted by sell-side estimates? Is the short-term reaction overdone, giving me a better price?
Rising rates are bad for equities.
"Rates are rising" has been the most-discussed story for 18 months — how much is in the price? The real marginal information is the speed and the terminal level.
A high P/E is expensive.
The P/E embeds a market path for the denominator. Where does my own 5-year EPS and terminal-value view sit in the distribution of consensus?
Classic Case

From late 2008 to March 2009, at the bottom of the credit market, Oaktree deployed roughly $10 billion into distressed debt. First-level thinking: "Lehman has failed, the global financial system is collapsing, stay away from debt." Second-level thinking: (1) high-yield was yielding 22% — historically a level from which you do not lose money over five years; (2) the sellers were forced sellers (fund redemptions, margin calls), so prices were detached from fundamentals; (3) even at default rates as bad as the Great Depression, expected returns would beat equities. Those investments delivered roughly 50%+ returns over the following 3–5 years. The edge was not "calling the bounce" but seeing clearly that prices were extremely pessimistic, sellers were forced, and the risk was being over-paid for — a clear-eyed view outside the consensus.

Counterexamples and Limits

Common misuses: (1) treating "different from consensus" as second-level thinking — you might simply be wrong; (2) becoming so attached to "independent views" that you stand against consensus every time — consensus is right most of the time; (3) turning second-level thinking into endless internal debate, missing the execution; (4) no feedback loop — you need to verify ex post: when your view differs from consensus, how often are you right? If under 50%, you are not doing second-level thinking, you are being stubborn. Marks's standard: "To be a second-level thinker, you must be able to articulate specific, falsifiable reasons the consensus could be wrong — not just 'I think they're wrong.'"

Decision Checklist
  • What is the market consensus on this investment? Can I restate it accurately in one sentence?
  • Exactly where does my view differ — and is that difference falsifiable?
  • If I am right, who loses to me? If I am wrong, who wins?
  • Is my information source the same as everyone else's? What angle is nobody else looking at?
  • Over my last ten "different from consensus" calls, what is my hit rate?
English Insight

"You can't do the same things others do and expect to outperform." — Howard Marks

Reflection This Week
For your largest holding, write two short paragraphs: (A) the current market consensus on this stock; (B) precisely where my view differs from that consensus. If (B) has no identifiable difference from (A), you may own it only by following — not by edge.