Max drawdown for CS2 skin investors: why path-dependent risk matters more than volatility
Sharpe ratio assumes returns follow a normal distribution. Skin returns don't — they have fat tails: rare large moves both up (Major tournament spikes, contraband revaluations) and down (case discontinuations, meta shifts). Drawdown is the metric that captures what Sharpe misses. STI 30's current max drawdown is −49.1% (2025-06 → 2026-02 trough, driven by the methodology-level luxury cap event of April 2026). Plus a previous −30.8% drawdown in 2021-2022 that took 18 months to recover. Here's why this matters more than the volatility number you usually see.
The 30-second definition
Max drawdown is the largest percent decline an investment experienced from a peak to a subsequent trough. If a portfolio reached $10,000 in March 2021 and fell to $7,000 by December 2021 before recovering, its max drawdown for that period was −30%. The formula:
drawdown_t = (value_t - running_max_t) / running_max_t
max_drawdown = min(drawdown_t over the entire window)
where running_max_t = max(value_0, value_1, ..., value_t)Three related drawdown metrics that get conflated:
- Max drawdown: the worst peak-to-trough decline ever recorded in the window. Backward-looking summary of worst-case-experience.
- Current drawdown: how far the asset is below its all-time high right now. If still in drawdown, this is what you're living through.
- Time-to-recovery: months from drawdown trough to a new all-time high. Measures how long the "underwater period" lasted.
Why drawdown matters specifically for skin investors
Three reasons drawdown is more informative than vol for skin portfolios:
- Path dependence. Two portfolios with identical CAGR can have wildly different drawdown profiles. A steady-grow portfolio and a boom-bust-recover portfolio look identical on year-over-year. They feel completely different to hold.
- Fat tails. Skin returns are not normally distributed. Tournaments, Operations, macro shocks produce occasional large moves that volatility (which assumes Gaussian) understates. Drawdown captures the actual realized worst case, not the theoretical one.
- Recovery cost asymmetry. A −50% drawdown requires +100% to recover. A −67% drawdown requires +200%. The math is unforgiving — small additional losses near zero require enormous gains to undo. Drawdown captures this asymmetry; vol doesn't.
STI drawdowns: 6 years of real data
Computed from monthly snapshots since each index's base date. Methodology details at /en/methodology.
| Index | Max DD | Trough | Recovery (months) | Calmar |
|---|---|---|---|---|
| STI 30 | −49.1% | Feb 2026 | in progress | 0.11 |
| STI 500 | −27.0% | Nov 2021 | ~14 | 0.33 |
| STI 100 | −25.5% | Dec 2021 | ~16 | 0.17 |
| STI 1000 | −23.0% | Nov 2021 | ~12 | 0.43 |
| STI Cases | −25.0% | Aug 2022 | ~8 | 1.90 |
| STI Stickers | −18.0% | Sep 2022 | ~10 | 0.34 |
| STI Agents | −35.0% | Mar 2023 | not recovered | −0.01 |
The standout: STI Cases has the best Calmar (1.90) by a wide margin — it had a significant drawdown but recovered fast and continued compounding. STI Agents has never fully recovered from its 2023 drawdown, which is why its CAGR sits near zero and Calmar is essentially negative.
Case study: STI 30's −30.8% drawdown in 2021 (now superseded)
Note: STI 30's current max drawdown is actually larger (−49.1%, methodology-driven by the April 2026 luxury cap event). The 2021 episode below is the most instructive macro-driven drawdown — a clean case study of skin market response to global liquidity tightening, before the methodology-level intervention.
The most instructive drawdown event in CS2 skin history is STI 30 from April 2021 to December 2021. Timeline:
- April 2021: STI 30 hit local peak after 14 months of QE-fueled rally. Both crypto and skins were near all-time highs. Many skin investors saw 6-12 months of +30% gains and felt invincible.
- May-July 2021: Fed began signaling tapering. Risk assets started compressing. STI 30 fell 12% in 3 months. Many traders held, expecting recovery.
- August-December 2021: Selic in Brazil hit 9.25%, signaling firmly hawkish. Drawdown extended to −30.8% by year-end. Concentrated portfolios (single weapon, single wear tier) saw individual skin drawdowns of −40 to −60%.
- 2022 (full year): Sideways with mild recovery. Volume declined as bear-cycle sentiment took hold.
- Q1 2023: STI 30 finally regained April 2021 peak. Total time underwater: ~24 months.
The lesson: the drawdown wasn't about a CS2-specific event. It was a global liquidity event affecting all risk assets. Skins moved with crypto and equities, just with a lag. Sharpe ratio for STI 30 over this window was negative; max drawdown and time-to-recovery captured the actual experience much better than the volatility number alone.
Calmar ratio: drawdown-adjusted return
Calmar ratio = CAGR ÷ |max drawdown|. It's an alternative to Sharpe that uses worst-case loss instead of volatility. For STI 30 with current max drawdown:
CAGR (STI 30, 2020-2026) = +5.5%/year
Max drawdown = -49.1% (2025-06 → 2026-02)
Calmar ratio = 5.5 / 49.1 ≈ 0.11Calmar interpretation:
- Calmar > 1.0: very good — annual return exceeds worst-case loss. STI Cases (~1.90) sits here.
- Calmar 0.5-1.0: decent. Most equity ETFs sit in this range long-term.
- Calmar 0.2-0.5: weak. Most STI tiers (excluding Cases) fall here.
- Calmar < 0.2: poor. STI 30 currently sits here at 0.11 — the recent −49% drawdown (luxury cap methodology event) compressed Calmar substantially.
Sharpe and Calmar can disagree on which asset is "better" because they measure different risk dimensions. For alternative assets with fat tails (skins, crypto, private equity), Calmar is often the more honest metric.
Time-to-recovery: the underrated metric
A drawdown of −30% feels different depending on how long it lasts. Three months and you're back to break-even feels like a temporary setback. Three years and you're finally back to break-even is a different psychological event entirely.
For STI tiers, recovery times have ranged from 8 months (STI Cases) to never (STI Agents, still in drawdown 25+ months later). Concentrated single-skin portfolios can stay underwater for 5-10 years if the underlying skin lost cultural relevance.
For investor decision-making, time-to-recovery is often more actionable than drawdown depth. A skin in 2-month drawdown might bounce. A skin in 24-month drawdown probably won't. Use the historical recovery time of similar drawdown events to set your expectations.
How to use drawdown in portfolio decisions
- Set a drawdown budget upfront. Decide before you allocate: "I can tolerate a max drawdown of −X% on this portfolio." If your tolerance is −20% and you're holding skins (where the broad market has had −27% drawdown historically), you're mismatched. Either lower allocation or accept that you may need to hold through pain.
- During a drawdown, diagnose the cause. Macro (whole market is down) tends to revert; structural (your skin specifically out of meta) usually doesn't. Compare your skin's chart to STI 500. If they move together, hold. If yours diverged downward, consider cutting.
- Use Calmar for cross-asset allocation. When deciding between skins, equities, and bonds, Calmar gives a single number per asset that captures both return and worst-case experience. Higher Calmar = better.
- Don't panic-sell at the trough. Hindsight on STI 30: the December 2021 trough was the worst time to sell. By Q2 2023, the index had recovered. Of course, the corollary is also true — you can't know in real time whether you're at the trough or only halfway down. Drawdown discipline beats real-time intuition.
Limitations
- Backward-looking: max drawdown over the last 6 years tells you what happened, not what will happen. Future drawdowns can be deeper, especially after regime changes (CS2 launch was one; the next macro cycle will be another).
- Window-dependent: a 5-year window catches different events than a 10-year window. The 2008 financial crisis isn't in any STI window because skin tracking didn't exist then. The implicit assumption: macro shocks of similar magnitude could happen again.
- Granularity matters: monthly snapshots underestimate intra-month drawdowns. Daily-snapshot drawdown for STI 30 in 2021 was probably slightly worse than the −30.8% monthly value reported. Skin Trackers uses monthly because cleaner aggregation; the drawdown numbers are conservative estimates of intra-month extremes.
- Doesn't capture illiquidity: an index assumes you can sell at the mid-price. In reality, during drawdown periods, bid-ask spreads widen significantly. Realized exit prices during 2021 drawdown were probably 5-10% worse than the index implied.
Frequently asked
What is max drawdown in plain English?
Max drawdown is the largest percent decline an investment experienced from a peak to a subsequent trough. If a skin reached $1,000 in March 2021 and then fell to $700 by December 2021 before eventually recovering, its max drawdown for that period was -30%. It's a backward-looking metric that captures the worst experience an investor would have lived through holding the asset. Critically, max drawdown is path-dependent — two assets with identical CAGR can have wildly different drawdown profiles, and the lived experience differs accordingly.
Why is drawdown different from volatility?
Volatility (standard deviation of returns) treats upward and downward swings symmetrically. A +20% month and a -20% month contribute equally to volatility. Drawdown only counts downside moves, and only counts them in cumulative form. An asset with steady -2%/month for 12 months has the same volatility as one with +5%/month for 11 months and -55% in month 12 — but the drawdown profile is opposite. For investor experience, drawdown is the more honest metric. Volatility is mathematically convenient but emotionally misleading.
What was the worst drawdown in CS2 skin history?
Depends on the index. STI 30 (blue chip) currently shows max drawdown of −49.1% (2025-06 → 2026-02), driven significantly by the April 2026 luxury cap methodology event that removed Howl/Dragon Lore/Hot Rod. Prior to that, the −30.8% drawdown from April 2021 to December 2021 (~18 months recovery) was the worst, coinciding with global QT signaling. STI 500 (broad market) had a similar −27% drawdown over the 2021 period. STI Cases had multiple drawdowns of −20-25% but recovered faster (5-8 months). Individual skins can hit −60-80% drawdowns (especially Operation cases that lose cultural relevance). Live values per index at /indices.
What's the Calmar ratio?
Calmar ratio = CAGR ÷ |max drawdown|. It's a risk-adjusted return metric like Sharpe, but uses worst-case drawdown instead of volatility as the denominator. For STI 30 (CAGR +5.5%/year, current max drawdown −49.1%), Calmar = 5.5 / 49.1 ≈ 0.11 — weak by any standard. For STI Cases (+47.5%/year, max drawdown ~−25%), Calmar ≈ 1.9. Calmar is sometimes preferred over Sharpe because it focuses on tail risk — what happens in the worst case — rather than average volatility. Hedge funds often target Calmar > 1.0; values above 2.0 are exceptional.
How long does it typically take to recover from a skin drawdown?
Recovery time (sometimes called 'underwater period') varies enormously. STI 30's −30.8% in 2021 took ~18 months to recover. STI 500's −27% took ~14 months. Individual skins can take years — some Operation cases that crashed in 2018-2019 still haven't fully recovered to ATH levels. The brutal math: a −50% drawdown requires +100% to recover; a −67% drawdown requires +200%. For path-dependent decisions (when do I sell? when do I add?), recovery time is often more useful than the drawdown depth itself.
How do I compute the drawdown of my own portfolio?
Step 1: build a monthly equity curve (portfolio value at month-end). Step 2: at each point, find the running max — the highest value seen so far. Step 3: drawdown at that point = (current_value − running_max) ÷ running_max. Step 4: max drawdown = the worst (most negative) drawdown ever recorded. Skin Trackers' /inventario page automates this against a Steam-imported portfolio. For Excel, columns: date, value, running max, drawdown%.
Should I sell during a drawdown?
Depends on the cause. Drawdown driven by macro (QT cycle, recession fears, USD/BRL spike) tends to revert as the macro normalizes — selling locks in the loss. Drawdown driven by structural decay (case dropped, skin out of meta) typically doesn't revert — selling is the correct action. The diagnostic question: is this skin in drawdown alone, or is it part of a broader market move? Compare your individual skin's chart to STI 500. If they move together, it's macro. If your skin diverged downward while market was flat, it's structural.
What's the limitation of using drawdown for decision-making?
Three caveats. First, drawdown is backward-looking — past worst-case doesn't predict future worst-case (especially after regime changes like CS2 launch). Second, drawdown depends on the measurement window — a 5-year window catches different events than a 10-year window. Third, drawdown ignores recovery and post-recovery returns — a portfolio that hit −50% then went +200% looks 'risky' on max drawdown alone, but the realized return may be excellent. Pair drawdown with Calmar (return ÷ drawdown) and recovery time for the fuller picture.
Where to go from here
- See current drawdown for each STI index at /indices — max drawdown column included in the comparison table.
- For the volatility-based companion (Sharpe ratio), see /en/blog/sharpe-ratio-cs2-skin-investors. Sharpe and drawdown together cover the two main risk dimensions.
- For the practical "is my portfolio in drawdown?" view, see /inventario — Steam login imports your portfolio and tracks drawdown against multiple STI benchmarks.
- For full methodology including how drawdown is computed in our pipeline, see /en/methodology.
Methodology note: drawdowns computed from monthly snapshots (running max + cumulative decline). Recovery time measured as months from trough until a new all-time high. Calmar = CAGR ÷ |max drawdown|. Source code in src/lib/metrics.ts (open), raw data via /api/indices.