From first principles to institutional-grade fluency. What hedge funds actually are, how they generate returns, what they cost — and when they belong in your portfolio.
Nine chapters that build from foundational concepts to expert-level due diligence.
A hedge fund is a privately organised investment vehicle that pools capital from sophisticated investors and deploys it using a broader range of strategies — including short selling, leverage, and derivatives — than a conventional mutual fund, with the goal of generating positive returns regardless of market direction.
The name "hedge" is somewhat misleading today. Alfred Winslow Jones, who invented the structure in 1949, used simultaneous long and short positions to hedge out market risk, isolating pure stock-picking skill. Modern hedge funds often take on considerable market exposure — many are not truly "hedged" at all.
| Feature | Mutual Fund | Hedge Fund |
|---|---|---|
| Investors | Open to general public | Accredited / qualified investors only |
| Regulation | Heavily regulated (SEC, UCITS) | Lightly regulated; registered offshore |
| Short selling | Rarely permitted | Core tool |
| Leverage | Tightly restricted | Extensively used |
| Derivatives | Limited | Unrestricted |
| Liquidity | Daily redemption | Monthly / quarterly / annual lock-ups |
| Fees | ~0.1–1.0% management fee | ~1.5–2% management + 20% performance |
| Transparency | Full disclosure required | Minimal public disclosure |
| Return target | Beat a benchmark (relative) | Positive return always (absolute) |
Hedge funds are restricted to accredited investors in most jurisdictions — individuals with net worth exceeding $1M (excluding primary residence) or annual income above $200K in the US. Institutional investors (pension funds, endowments, family offices) are the dominant capital source today. Minimum investments typically range from $500K to $5M+.
Family offices sit at a structural advantage: they can negotiate lower minimums, better fee terms, and access capacity-constrained funds that are closed to institutional money. However, the same access advantage applies to the downside — you are expected to perform your own due diligence without intermediary protection.
Understanding the industry's evolution is essential to understanding its current character — and its recurring failure modes.
Hedge funds are not one asset class — they are a legal structure that houses dozens of fundamentally different investment approaches. The most important step in evaluating any hedge fund is identifying which strategy family it belongs to.
The original Jones model. Buy undervalued stocks, short overvalued ones. Net market exposure can range from 0% (market-neutral) to 100%+ (net long, leveraged). The spread between longs and shorts is the primary alpha source. This is the largest strategy category by AUM.
Key variants: Market-neutral (beta ≈ 0), long-biased (retains significant equity upside/downside), sector-specialist (healthcare, technology, financials).
Top-down bets on macroeconomic themes: interest rates, currencies, commodity supercycles, central bank policy. Soros's Black Wednesday trade is the archetype. Instruments include forwards, options, futures on currencies, bonds, equities, and commodities.
The best macro managers (Bridgewater, Brevan Howard, Caxton) combine deep macro research with disciplined risk management. The worst are simply directional bets with expensive fees.
After a merger is announced, the target company typically trades at a discount to the deal price — compensating for deal-failure risk. The arb buys the target, shorts the acquirer (in stock deals), and collects the spread at closing.
Returns are normally uncorrelated to markets but spike during crises (when deal financing dries up and correlations spike). Annualised returns: 4–8% in normal conditions. Key risk: regulatory blocking, deal withdrawal, financing failure.
Investing in the debt or equity of companies in or near bankruptcy. Distressed investors buy when forced sellers (pension funds prohibited from holding sub-investment-grade bonds, banks cleaning balance sheets) create mispricings far below fundamental value.
The investor must then navigate complex restructuring negotiations, often taking an active role as a creditor or becoming an equity holder post-reorganisation. Requires deep legal expertise.
Exploits pricing anomalies between related fixed income instruments — on-the-run vs. off-the-run Treasuries, cash bonds vs. futures, or bonds of similar credit quality at different yields.
This is the LTCM strategy. The spread between instruments is tiny (basis points), so extreme leverage (10–30×) is required to generate meaningful returns. This makes it acutely vulnerable to liquidity crises — when everyone tries to exit simultaneously, spreads widen rather than converging.
Uses mathematical models and algorithms to identify and exploit statistical relationships between securities. Holding periods range from milliseconds (HFT) to weeks (medium-frequency stat arb). Renaissance Technologies' Medallion Fund is the canonical example — generating 66% gross returns annually for decades.
Modern quant funds include machine learning, alternative data (satellite imagery, credit card transactions, NLP on earnings calls), and factor-based approaches.
Internally diversified pods, each running different strategy books under one umbrella. Capital is dynamically allocated between pods based on risk-adjusted opportunity. Citadel, Millennium, Point72, and Balyasny are the dominant players.
The advantage: genuine diversification within a single fund. The risk: complexity and opacity; investors trust the chief risk officer to manage cross-strategy correlation, especially in stress.
Commodity Trading Advisors (CTAs) use momentum and trend-following algorithms across futures markets: commodities, currencies, rates, and equity indices. They are systematically long what has risen and short what has fallen.
Uniquely, trend-following tends to perform well during sustained crises (e.g., the 2008 crash) because it rides extended directional moves. This makes it a genuinely diversifying allocation — a rare quality.
Three tools define what hedge funds can do that conventional funds cannot: leverage, short selling, and derivatives. Understanding each is non-negotiable before committing capital.
Leverage means investing more capital than you actually have by borrowing the remainder. A fund with $1 billion in equity capital might borrow $2 billion more, giving it $3 billion in buying power (3× leverage). This amplifies both returns and losses proportionally.
At 25× leverage, a 4% adverse move in your positions wipes out your entire equity capital. LTCM discovered this when Russian default caused bond spreads to widen rather than converge. Their models showed this as a 10-sigma event — meaning it "should not happen" in the lifetime of the universe. It happened within months. Leverage transforms survivable drawdowns into existential ones.
A short sale is the sale of a security you do not own, borrowed from a broker for a fee. The mechanics: borrow shares → sell them at current price → repurchase them later (ideally at a lower price) → return them to the lender, pocketing the difference minus the borrowing cost.
Company X trades at $100. You believe it's worth $60. You borrow 1,000 shares and sell them: +$100,000 cash received. Stock falls to $60. You buy 1,000 shares to return: −$60,000 cost. Gross profit: $40,000 (minus borrow cost and borrowing fees).
A long position can only lose 100% (if the stock goes to zero). A short position has theoretically unlimited loss — there is no ceiling on how high a stock can rise. GameStop in 2021 saw shorts lose 500%+ as a coordinated retail squeeze forced them to buy at any price. This is why position sizing on shorts is critical.
Derivatives derive their value from an underlying asset. Hedge funds use them to gain leveraged exposure, hedge existing positions, or express complex views that cannot be captured through outright long/short positions.
Rights (not obligations) to buy (call) or sell (put) at a fixed price. Used for hedging downside (buying puts), leveraged upside (buying calls), or income generation (selling covered calls). The key feature: limited downside for the buyer (maximum loss = premium paid), while the seller has potentially unlimited exposure.
Obligations to buy or sell at a predetermined future price. Futures are exchange-traded and standardised (hence liquid); forwards are OTC and customisable. Used by CTAs for systematic trend-following and by macro funds for currency/rate positioning. The margin requirement is a fraction of the notional, creating natural leverage.
Insurance against a borrower's default. The protection buyer pays a periodic premium; the protection seller pays face value if default occurs. John Paulson's $15B+ profit in 2007–08 came from buying CDS on mortgage-backed securities — essentially buying insurance on assets he predicted would fail. The CDS market was $62T at its 2007 peak.
Exchange of fixed for floating interest payments on a notional principal. Used by fixed income relative value funds to express views on the yield curve's shape without outright bond purchases. Notional values are enormous relative to the economic exposure, making swaps a core tool for leveraged rates strategies.
Most hedge funds use a limited partnership structure: the fund manager is the General Partner (GP), investors are Limited Partners (LPs). The LP's liability is limited to their investment; the GP bears unlimited liability and makes all investment decisions.
Lock-up period: Minimum period capital must remain invested (3–12 months typically). Gate provision: Fund can limit redemptions to a percentage of NAV per period to prevent forced liquidations. Side pocket: Illiquid investments segregated from main fund — redeeming investors cannot exit until the side pocket is liquidated. These provisions can trap investors in declining funds.
Fees are the most underappreciated destroyer of hedge fund returns. The 2&20 structure is mathematically brutal when compounded over time.
The manager cannot charge a performance fee until losses from prior periods are fully recovered. If the fund drops 20% and then recovers 20%, the manager earns no performance fee on the recovery — the investor must be made "whole" first.
The limitation: After large drawdowns, talented managers often close the fund and launch a new one with a fresh HWM, walking away from their responsibility to restore investor capital. This is legal but deeply problematic.
A minimum return threshold the manager must exceed before performance fees apply. Typically set at LIBOR/SOFR or a fixed rate (e.g., 5%). If the fund returns 4% and the hurdle is 5%, no performance fee is charged.
This protects investors from paying performance fees for returns that could be earned risk-free — a basic fairness mechanism. Yet many funds negotiate hurdle rates as low as 0%, meaning any positive return triggers a performance fee.
Allows investors to reclaim previously paid performance fees if subsequent losses occur. Rare in practice. Most fund documents provide for a "clawback period" (e.g., 3 years) but collecting it is legally complex and commercially difficult.
Swensen's analysis is damning: even a long/short equity manager who generates top-quartile gross returns needs to sustain that level consistently for investors to materially outperform a passive strategy after fees. A manager producing 8% gross becomes 5.4% net after 2&20, below what a 60/40 index might return. The fee structure is not trivially overcome.
Evaluating hedge fund performance requires more sophistication than evaluating a stock or mutual fund. Raw returns tell you almost nothing — the quality, consistency, and source of those returns matters far more.
| Metric | Formula / Definition | What It Tells You | Watch For |
|---|---|---|---|
| Sharpe Ratio | (R_p − R_f) / σ_p Excess return per unit of total risk |
Overall risk-adjusted return efficiency. A Sharpe above 1.0 is good; above 2.0 is exceptional (and should invite scrutiny). | Assumes normally distributed returns. Strategies that sell options can show artificially high Sharpe ratios — until the tail risk materialises. |
| Sortino Ratio | (R_p − R_f) / σ_downside Excess return per unit of downside risk only |
Penalises only harmful volatility (downside). More appropriate for asymmetric return strategies where upside volatility is desirable. | Can flatter strategies with infrequent but severe drawdowns — because downside volatility looks low until the crash occurs. |
| Maximum Drawdown | Peak-to-trough decline in NAV |
The worst loss an investor would have experienced. Essential for understanding real psychological and financial risk. A 40% drawdown requires a 67% subsequent gain to recover. | Historical max drawdown is a floor, not a ceiling — future drawdowns can always be worse. |
| Calmar Ratio | Annualised Return / Max Drawdown |
How much return you got per unit of maximum pain. A Calmar of 0.5+ over 3+ years suggests reasonable risk management. | Short track records make this meaningless — a fund with 2 years of great returns and no drawdown yet is not validated. |
| Alpha (Jensen's) | R_p − [R_f + β(R_m − R_f)] Return above what market beta explains |
The true "skill premium." A fund with beta of 0.8 to the S&P 500 is expected to return 0.8× the market premium. Alpha is what the manager adds beyond this. | Many hedge funds have more market beta than they claim. In bull markets, high-beta funds look like alpha generators. |
| Correlation to S&P 500 | Pearson correlation coefficient (−1 to +1) |
How much of the fund's movement is explained by general market moves. A true diversifier should show low or negative correlation (<0.3). | Correlations increase in crisis periods — this is called "correlation breakdown." A fund that looks uncorrelated in normal times may move with the market exactly when you need diversification most. |
| Volatility (Std Dev) | σ of monthly / annual returns |
Overall return variability. A rough proxy for risk — but not sufficient alone. Two funds can have identical volatility with very different drawdown profiles. | Illiquid hedge funds can show artificially low volatility because positions are infrequently marked to market. "Stale pricing" disguises risk. |
Academic research (Ibbotson, Malkiel) shows that hedge fund databases overstate historical returns by 7–12% per annum due to two biases: Survivorship bias — failed funds are removed from databases, leaving only survivors; Backfill bias — new funds added to databases submit historical records selectively (only funds with good histories bother to register). The implication: the "average" hedge fund return you see in industry databases is a fiction. The true net-of-fees, net-of-bias return is substantially lower — often below what a 60/40 portfolio delivers.
The barriers to effective hedge fund due diligence are real: opacity, complexity, and information asymmetry favour the manager. A rigorous process is your only defence.
These are the patterns that have preceded fund failures, frauds, and blow-ups. No single flag is necessarily disqualifying, but multiple flags in combination are.
The decision to allocate to hedge funds — and how much — depends critically on your goals, your ability to conduct due diligence, and your access to genuinely skilled managers.
Swensen classifies hedge funds under "Absolute Return" — a target allocation of approximately 20–25% for institutional investors. The key criterion: the strategy must genuinely exhibit low correlation to equities and bonds. Funds that simply provide leveraged equity beta at high fees add nothing that a leveraged index ETF could not.
Curated by depth and focus. Start with Tier 1, which provides both breadth and narrative. Move to Tier 2 for rigour. Tier 3 is specialist reading for those who wish to move from informed allocator to practitioner-level understanding.