In February 2007, John Paulson’s investors received a monthly performance statement showing a 66% gain and assumed it was a clerical error — a misplaced decimal on what should have read 6.6%. It was not. By year-end, Paulson & Co.’s flagship credit fund had returned 590%, and the firm had booked $15 billion in profit from a concentrated short against subprime mortgage credit — a trade later documented in Gregory Zuckerman’s The Greatest Trade Ever and now regarded as one of the largest single-year profits in hedge fund history.
What separates Paulson’s trade from a simple bearish macro call is its mechanical structure: rather than shorting housing stocks or equities outright, Paulson built the position through credit default swaps — instruments that behave like insurance contracts, capping annual loss at a fixed premium while offering convex, uncapped payout on default. That structural choice, rooted in decades of merger-arbitrage and risk-arbitrage training rather than macro forecasting, is the throughline of his entire career — including the firm’s difficult retrenchment after 2011 and its eventual conversion into a family office in 2020.
Background & Evolution: From Risk-Arb Specialist to Convexity Architect
Paulson’s path into markets ran through Boston Consulting Group and a stint as a managing director in mergers and acquisitions at Bear Stearns, followed by a general partnership at Gruss Partners — training grounds not in macro speculation but in deal-spread analysis: pricing the probability-weighted outcome of corporate transactions. In 1994, he founded Paulson & Co. with $2 million in capital and a single employee, building a merger-arbitrage and event-driven track record that operated, in his own description, in the top quartile of its peer group for years before subprime brought public notoriety.
That risk-arbitrage discipline is the origin of the philosophy that later defined the subprime trade: identify situations where the market misprices a discrete, analyzable event, then structure exposure so the downside is contractually bounded while the upside is convex. Paulson has been explicit that the 2006–07 trade was not an improvised macro bet but the product of a much longer search for the right setup within an asset class he had studied for decades.
The execution phase tested that discipline directly. Through late 2006 and into early 2007, as housing prices continued to climb, Paulson’s CDS positions showed mark-to-market losses, and investors questioned whether the size he had taken was reckless. The position was not vindicated until subprime delinquencies accelerated in spring 2007 — a multi-quarter stretch of unrealized pain that forged the conviction-holding, negative-carry-tolerant approach that defines his risk management style.
Inside the John Paulson Trading Strategy: Core Rules
Paulson’s framework rests on a single organizing idea: structure exposure so that maximum loss is contractually fixed while payout is convex, then size the position aggressively only once that asymmetry is confirmed at the right entry price. Six mechanical rules, distilled from the subprime trade and the broader Paulson & Co. playbook, describe how that idea is implemented.
Career Performance Timeline
| Year | Return | Benchmark | Outperformance | Key Event | Data Source |
|---|---|---|---|---|---|
| 2007 | +590% (flagship) / +353% (secondary fund) | Not sourced in research notes | Not sourced in research notes | Peak subprime CDS short; firm profit $15B; ~$25B peak notional exposure | Tier 2 — Consensus |
| 2008 | Profitable (exact % not disclosed) | Not sourced in research notes | Not sourced in research notes | One of few funds profitable amid broad market collapse | Tier 3 — Lore |
| 2009 | Profitable (exact % not disclosed) | Not sourced in research notes | Not sourced in research notes | Early rotation into distressed/recovery credit; bullish recovery thesis proved premature | Tier 3 — Lore |
| 2010 | $4.9B personal earnings (gold-driven) | Not sourced in research notes | Not sourced in research notes | Gold Fund thesis (currency debasement) becomes primary alpha source | Tier 2 — Consensus |
| 2011 | Advantage −32.57% to −47% / Advantage Plus −45.35% to −52% (source-conflicted) | Not sourced in research notes | Not sourced in research notes | “Worst year in firm’s 17-year history”; Sino-Forest fraud exposure; net equity exposure too great for macro conditions | Tier 2 — Consensus |
| 2012 | Continued underperformance (exact % not disclosed) | Not sourced in research notes | Not sourced in research notes | Bloomberg BusinessWeek cover: “The Big Loser” | Tier 3 — Lore |
| 2019 | +30% (merger-oriented fund) | Not sourced in research notes | Not sourced in research notes | Final full year with outside capital before family-office conversion | Tier 2 — Consensus |
| 2020 | −10% | Not sourced in research notes | Not sourced in research notes | Full conversion to family office; returns all outside capital | Tier 2 — Consensus |
Core Rules Table
| Core Principle | Market Logic | Replicability |
|---|---|---|
| Structure as insurance purchase, not naked directional bet | CDS premium (bps of notional) caps annual max loss; payout on default is par-minus-recovery, often 100x+ premium — a fixed-cost, convex-payout structure | Low |
| Target the most mispriced tranche in the capital structure | BBB tranches carried thinner subordination (~4–7%) than AAA, offering more convexity per dollar of premium for a given pool-wide default rate | Low |
| Concentrate size once thesis and entry price both clear the bar | Scaled to ~$25B notional with a handful of dealer counterparties rather than diversifying away a high-conviction asymmetric setup | Medium |
| Budget negative carry as a controlled cost of holding optionality | ~$300M/year premium bleed at peak size required capital reserved explicitly for multi-year survivability, not just eventual payoff magnitude | Medium |
| Rotate strategy allocation across the credit cycle | Sequenced subprime short (2006–07) → distressed/recovery credit long (2008–09) → equity/gold reflation (2009–2011) → merger-arb reversion (post-2012) | Medium |
| Merger arbitrage as structural base strategy | Deal-spread capture (target price minus current price, adjusted for break probability and time-to-close) generates steady base carry underneath concentrated convex overlays | Medium |
Career Performance: Verified Metrics
- Asymmetric structuring discipline: the 2006–07 trade capped annual downside at CDS premium cost while retaining uncapped upside on default — a risk-arb mentality applied at macro scale.
- Conviction under mark-to-market pain: held the subprime short through a multi-quarter drawdown period in late 2006–early 2007 before vindication, despite investor pressure to reduce risk.
- Multi-strategy rotation: successfully redeployed capital from subprime credit into distressed credit (2008–09) and then gold (2010), each phase generating outsized personal and firm-level profit.
- Durable base strategy: merger arbitrage has remained a consistent top-quartile performer across the firm’s 25-plus-year history, independent of the headline macro trades.
- Structural adaptability: converted the firm to a family office in 2020 rather than persist under redemption pressure with a diminished capital base, preserving flexibility to hold positions longer without investor liquidity constraints.
| Metric | Value | Period |
|---|---|---|
| Flagship Fund Return | +590% ◎ | 2007 |
| Secondary Fund Return | +353% ◎ | 2007 |
| Firm-Level Profit | $15B ◎ | 2007 |
| Personal Earnings | ~$4B ◎ | 2007 |
| Personal Earnings (Gold) | ~$4.9B ◎ | 2010 |
| Peak Drawdown (Advantage Plus) | −45.35% to −52% ⚠ | 2011 |
| Peak AUM | $36B–$38B ⚠ | 2011 |
Strategy Limitations: Where the John Paulson Edge Decays
The mechanics that produced the 2007 result depended on specific structural conditions in credit markets that have since been substantially reformed or arbitraged away, and the firm’s own subsequent decade illustrates several distinct failure modes of the underlying rule set.
- Regime dependency of the mispricing edge: the trade required both securitization/rating-agency blind spots and an immature CDS market permitting deep pricing dislocation on junior tranches. Post-2008 rating-agency reform and tighter CDS ISDA definitions (accelerated after the Hovnanian “manufactured default” controversy) have substantially closed this specific mispricing channel — the exact trade structure is not repeatable in the same asset class today.
- Macro-timing rules break down outside crisis dislocation windows: the 2011 drawdown shows that concentrated conviction-sizing has no built-in mechanism to distinguish “temporarily wrong, thesis intact” from “thesis broken” — net equity exposure was sized for a growth-and-orderly-European-resolution scenario that did not materialize.
- Negative-carry tolerance fails in prolonged low-volatility regimes: the same carry-budgeting rule that worked across a finite subprime timeline bled capital for years in the grinding post-crisis, low-volatility environment, contributing directly to the AUM decline from $36–38B to $19B between 2011 and 2015.
- Concentration risk has no embedded fraud safeguard: the Sino-Forest Corporation collapse in 2011 shows that concentrating on a high-conviction asymmetric setup assumes systemic/structural mispricing, not single-issuer misrepresentation — the rule set does not distinguish between the two risk types.
- Scale itself decays the edge: at $36–38B peak AUM, the pool of sufficiently large, sufficiently mispriced, sufficiently liquid opportunities shrinks, pushing the firm toward broader, higher-beta macro bets further from the original structured-credit informational advantage.
- Modern market structure raises replication costs: post-Dodd-Frank CDS central clearing and margin requirements increase negative-carry funding costs relative to the thinly-margined, bilateral dealer environment of 2006–07, reducing achievable notional-to-capital leverage for a similar trade today.
Verdict
For a modern proprietary trading desk or retail practitioner, the directly transferable lessons sit in position-sizing discipline and cost-of-carry budgeting rather than in the specific subprime CDS trade itself — the structural mispricing that made that trade possible has been substantially closed by post-2008 regulatory reform. The merger-arbitrage base strategy, by contrast, remains a durable and more accessible template: deal-spread capture as steady carry, with concentrated convex overlays reserved for rare, high-conviction dislocations.
Paulson’s 2020 conversion to a family office — trading external capital and its redemption pressure for full control over holding periods — is itself a final risk-management data point: even the architect of history’s most celebrated convex trade ultimately concluded that outside capital constraints were incompatible with his own risk tolerance for drawdown and time horizon.
