Thursday, October 30, 2025

QQQ vs. NVDA — A Sharpe-Only Look (1-Year & 3-Year)

 

  • 1-year: QQQ has the better risk-for-reward trade-off (higher Sharpe).

  • 3-year: NVDA wins on efficiency per unit of risk (higher Sharpe), despite a bumpier ride.

  • Use it: Keep QQQ as core, add NVDA as a smaller satellite, and rebalance on bands.


The Data (assumes a 4.0% risk-free rate)

AssetHorizonReturn (annualized)Risk-free rateStd-dev (annualized)Sharpe
QQQ1 year23.7%4.0%15.5%1.3
QQQ3 years32.1%4.0%16.8%1.7
NVDA1 year46.6%4.0%37.5%1.1
NVDA3 years146.5%4.0%65.7%2.2

Sharpe = (Return − Risk-free) / Volatility, all annualized and measured over the stated windows.


What the numbers say (plain English)

  • Last 12 months:
    QQQ is more efficient (Sharpe 1.3 vs 1.1). NVDA returned more in absolute terms, but its swings were ~2.4× larger, so you were paid less per unit of bumpiness.

  • Past 3 years:
    NVDA dominates (Sharpe 2.2 vs 1.7). Its extraordinary compounding more than outweighed higher volatility, so each unit of risk paid more than QQQ’s.

  • Regime signal:
    Both assets show higher Sharpe at 3Y than 1Y, hinting the broader 2023–2025 stretch was stronger than the most recent year alone.

  • Risk reality check:
    NVDA’s volatility (37.5%/65.7%) is far higher than QQQ’s (15.5%/16.8%). Expect bigger drawdowns and wider day-to-day moves with NVDA.


Projection (Sharpe-only)

  • Next 12 months: QQQ is more likely to post the higher Sharpe (steadier path). NVDA can beat QQQ only if returns stay exceptional and volatility cools.

  • Next ~3 years: NVDA can retain the Sharpe lead if AI-driven growth persists; QQQ remains stable and consistent, with a smoother Sharpe through cycles.


How to use this (simple playbook)

  1. Core–satellite: Make QQQ your core equity sleeve; use NVDA as a smaller satellite sized to your drawdown tolerance.

  2. Rebalance on bands: Example: target 80% QQQ / 20% NVDA; rebalance when either drifts ±20% of target (i.e., NVDA >24% or <16%).

  3. Stay consistent: Recompute Sharpe on the same windows, with the same risk-free and return frequency so the comparison remains apples-to-apples.

No comments:

Post a Comment

test