Tim's Wiki

Portfolio Risk Management

Properties2
descriptionSystematic approaches to managing portfolio risk through position sizing, hedging, and volatility control rather than purely maximizing returns.
tagstrading, risk, portfolio, quantitative

5 min read

Portfolio Risk Management

Definition

Portfolio risk management is the discipline of controlling downside exposure and stabilizing returns through systematic techniques — rather than simply picking securities that maximize expected return. It treats risk as a manageable variable, not just an unavoidable cost of investing.

Core Methodologies

Dollar Neutral

Balancing long and short positions so the net dollar exposure is zero or near-zero. Isolates alpha from market direction.

Beta Neutral

More sophisticated than dollar neutral — adjusts positions so net portfolio beta is zero. Reduces exposure to broad market moves while allowing individual security selection.

Beta neutral reduces drawdown more effectively than dollar neutral because it accounts for how sensitive each position is to the market.

Volatility Targeting

Scales position size inversely to recent volatility. When volatility rises, reduce exposure; when it falls, increase it. Stabilizes the portfolio’s realized risk over time.

Method (Harvey/Rattray/van Hemert, Ch. 3): pick a target annualized vol (σ_target, e.g. 10–15%); estimate rolling realized vol (σ_est) over a ~20–60 day window as √252 × stdev(returns); scale each position by σ_target / σ_est. Clamp the factor (e.g. 0.5× to 2×) to avoid extreme leverage, smooth the estimate with an EMA, and only rebalance when the factor moves >10–15% (transaction-cost bands). Works because volatility clusters — scaling inversely pre-emptively cuts risk before crises deepen. Variants: asset-level, portfolio-level, risk-parity integration, and adaptive targeting (lower the target when drawdown/correlation risk rises). It is defensive, not predictive — “controlling how much risk you take when markets are unpredictable.”

Drawdown Control

Sets maximum acceptable drawdown thresholds and de-risks the portfolio when approaching limits. Protects capital during extended downturns.

Framework (Ch. 5): the authors argue volatility is a weak risk measure (treats upside/downside equally) — investors feel drawdown ((Peak − Current)/Peak), which drives panic and forced liquidation. Track a live “underwater curve” and scale risk down as drawdown deepens — a “drawdown governor”:

DrawdownRisk scaling
0–5%100% (full target)
5–10%75%
10–15%50%
>15%30% or pause

Re-risk only after stabilization (volatility normalizing, trend turning positive) to avoid re-risking too early. Apply at both strategy level (pause/cut a manager breaching its historical drawdown norm) and portfolio level (scale total risk, shift budget from high- to low-drawdown strategies). Backtests limited losses in 2008 and the 2020 COVID crash without materially reducing long-term returns.

Strategic Rebalancing

Rebalance on risk, not the calendar. Naive fixed-interval rebalancing (e.g. quarterly reset to 60/40) is mechanical and can force buying into falling assets, amplifying losses in stress events (Ch. 4). Instead:

  • Keep risk contributions stable, not dollar weights (aim for “50% of portfolio vol from equities” rather than “60% equity weight”).
  • Add trend filters — only rebalance into an asset once it stabilizes (e.g. price > 200-day MA) to avoid catching a falling knife.
  • Link to drawdown thresholds — freeze/slow rebalancing above ~10–15% drawdown, resume gradually on recovery.
  • Use tolerance bands (±5–10%) to avoid over-trading. “Rebalancing should be a strategic decision, not a mechanical act.”

Tail Risk Management

Specifically protects against extreme (non-normal) market events. Uses options, trend-following overlays, or dynamic hedging to limit losses in tail scenarios.

Relationships Between Methods

graph TD
    A[Portfolio Risk Management]
    A --> B[Dollar Neutral]
    A --> C[Beta Neutral]
    A --> D[Volatility Targeting]
    A --> E[Drawdown Control]
    A --> F[Tail Risk Management]
    C --reduces--> E
    D --stabilizes--> E
    F --protects against--> E
  • Beta neutral and volatility targeting both reduce drawdown, but through different mechanisms
  • Tail risk management is distinct from volatility targeting — vol targeting handles normal regime changes, tail risk handles extreme events
  • Dollar neutral and beta neutral are independent approaches (beta neutral is generally superior)

Key Concepts

Kelly Formula

Optimal position sizing formula that maximizes long-term growth rate given known edge and variance. In practice, fractional Kelly (e.g., half-Kelly) is used to account for estimation error. See Kelly Criterion.

Sharpe Ratio as Design Target

The book “Strategic Risk Management” (Man Group) focuses on achieving better risk-adjusted returns rather than maximum absolute returns. Core insight: reducing risk cheaply is more reliable than finding alpha. See Sharpe Ratio.

Portfolio Construction Ideas

From personal portfolio research:

AllocationComponentsRationale
BRK.B 20% / MAGS 40% / SPMO 40%Wide industry + momentumBetter SR and CAGR than equal-weight
SHLD 33% / MAGS 33% / SPMO 34%Defense + tech + momentumSimilar performance, narrower
Long bonds + ETFs + cash for optionsMargin-efficient layeringUses bond margin to fund equity + options

Backtesting Tools

Existing Libraries

  • PyPortfolioOpt — mean-variance optimization, Black-Litterman, HRP, shrinkage. Extensive and extensible.
  • pyfolio — performance and risk analysis (Quantopian, discontinued)

TiPortfolio (Personal Project Idea)

A backtester focused on portfolio and risk management with:

  • Strategy as a function producing signals (0, 1, -1)
  • Portfolio layer using ABC with trigger and allocate methods
  • Metrics, live trading, and broker connection layers
  • Architecture: strategy receives prices → emits signals → portfolio allocates → metrics evaluate

Source References

  • Portfolio risk management focused backtester

Graph View

Backlinks