Liquidity & Technicals

Liquidity & Technicals

Tesla has compounded at 41.7% per year since 2010 IPO with 57% annualized vol — a 0.82 lifetime Sharpe achieved by enduring 34 separate drawdowns including a 73.6% peak-to-trough (Nov-2021 → Jan-2023). Today the stock sits 20.2% below the Dec-2025 all-time high of $489.88, just below its 200-day SMA, and the algo classifies the regime as Stage 1: Basing with vol in the 31st percentile of own history. The question this page settles: does the quant signature support buying the base, or is this just the calm before the next leg down?

---

1. Statistical character — does TA even apply to this stock?

H = 0.57 (mildly trending) but the supporting tests reject it on the daily timeframe — VR(5) z = −0.09, every lag from 1 to 63 days has zero significant autocorrelation. The trending signature lives in multi-month price regimes (the long bull legs the Hurst rescaled-range test detects with lags up to 665 days), not in day-to-day pattern-trading. Treat short-horizon chart signals here as noise; size on regime + factor + earnings, not on RSI flips.

---

2. Multi-factor decomposition — what actually drives the price?

The market explains only 24% of TSLA's daily variance — and once you add DXY, brent, gold and treasuries the multi-factor R² barely moves to 24.6%. 76% of TSLA's daily move is idiosyncratic — company-specific news, deliveries, Musk, AI/robotics narrative shifts. Factor models will never give the analyst a meaningful edge here. Worse, after stripping market beta the stock has bled −13.3% annualized alpha over the regression window — i.e., a market-neutral pair (long TSLA / short 1.57× SPY) would have lost ~13% per year. Rolling market beta has fallen from a 2021 peak near 2.5× to ~2.0× today, but TSLA remains a leveraged play on the index with a thick layer of company-specific noise on top.

---

3. Conditional return tables — base rate for today's regime

Current regime: Stage 1: Basing, 30-day realized vol = 41.5% (below-average bucket, 31st percentile of own history).

From this exact regime (Stage 1 basing + below-avg vol), across 118 historical occurrences, the next 30 days have been mildly negative (median −3.3%, only 39% positive) and the next 90 days have been flat (median −1%, hit rate 49%). The picture only turns constructive at the 180-day horizon where median forward return is +7.9% with a 58.5% hit rate. Translation: the base does eventually resolve up more often than down, but the next quarter is a coin flip with a slight downward tilt — don't expect quick relief.

---

4. Forward distribution — explicit price probabilities

The vol-conditional Monte Carlo carries an expected 90-day return of −1.7% with only a 42.6% probability of being positive — modestly bearish drift. Two crisp numbers a PM can act on: 63% chance the stock revisits $352 (−10%) within the next quarter, and only a 30% chance it breaks above the prior ATH at ~$489 (+20%). Said differently, the bootstrap thinks a re-test of the early-April low ($343) is more likely than a fresh ATH over the next three months.

---

5. Position sizing — what an actual PM would do with this

Half-Kelly comes back at +0.44 — a strikingly aggressive number that says "lifetime risk-adjusted return has been so far above the 6% rf hurdle that history justifies a >40% portfolio weight." For a $1B fund running 1% daily-σ risk per stock, the math allows up to ~710k shares (~$277M, 27.7% of AUM). No PM should actually run that: the +0.44 figure is dominated by the 2010-2021 super-compounding regime that may not repeat. CVaR(95%) of 8.2% means a typical bad-day loss against a $277M position is $22.7M — and that 8.2% is itself an average of 1-in-20 days; the 99% CVaR is 13.7%. Use Kelly as a ceiling, not a target — a more defensible bound is half-Kelly × half-confidence = 5–10% portfolio weight on a high-conviction view.

---

6. Cover — the setup at a glance

---

7. The story since IPO

The IPO-to-today log chart shows a stock that traded in a flat sub-$5 range for three years (2010–2013), one tight band $10–25 for five years (2014–2019), then a parabolic 25× compression into late-2021 followed by a 73.6% wipeout into Jan-2023 — the deepest drawdown in mega-cap-tech history. The regime classifier counts 34 bull and 35 bear segments on the 20% rule. The dominant feature today: price is at the same level as it first hit four years ago (Sep-2021), having round-tripped the entire 2024 rally.

---

8. Drawdown profile — what bad looks like for THIS stock

The current 20% drawdown is the 8th-deepest of 34 — uncomfortable but unremarkable for this stock. Median lifetime drawdown depth is 16% with a 9-day median recovery, but that headline obscures a heavy tail: four drawdowns of 50%+ including one that took 779 days peak-to-recovery (Nov-2021 → Dec-2024). The current open drawdown started 90 days ago and has 73 days of decline plus 24 of basing — the trajectory most closely resembles the early-2022 setup that ultimately bottomed 12 months later, but vol is far lower today (41% vs 78% in early-2022) suggesting a less violent path.

---

9. Volatility cone — calm or stressed?

Today's 30-day realized vol is 41.5% — the 31st percentile of TSLA's lifetime distribution. That is a calm regime by TSLA standards (the stock has spent more than two-thirds of its life with realized vol above today's level). Calm vol on TSLA historically resolves either into a sharp upside breakout (April 2024, October 2025) or a sleepy continuation lower (early 2025). Conditional on Stage 1 + this vol bucket, the next 90 days have median return −1% — see §3.

---

10. Where we are now — multi-MA + setup card

The setup is conflicted at every layer. Price has reclaimed the 50d (just) but a fresh death cross fired on April 10 — the fourth in three years (Feb-2024, Apr-2025, Aug-2025, Apr-2026). The Weinstein stage flipped to Stage 1 only 10 days ago after a long Stage 4 phase, which historically marks the beginning of base-building, not the end — meaning premature long entries get chopped. The 200d at $401 is the immediate ceiling; reclaiming and holding it would invalidate the bear case.

---

11. Patterns the algorithm flagged

The detector flags 44 lifetime breakouts and 175 double bottoms but the only currently-live structure of consequence is the double bottom at $413/$419 — and that has already been violated downward (April 8 low at $343.25), so the buy-the-base trade off that pattern has failed. The HH/HL ladder shows lower highs across the last four swings and a fresh lower low at $343 — the algo's "mixed" verdict is generous. Read it as a soft downtrend that's currently consolidating. The Sep-2025 breakout from $369 — the engine of the rally to $490 — has been fully reversed.

---

12. Earnings reaction footprint

Drift asymmetry: when TSLA gaps up on earnings, the next month's median return is −1.7% (initial pop fades). When it gaps down, the next month's median return is +2.8% (mean reversion). This is a textbook fade-the-reaction pattern — the immediate post-print move is, more often than not, wrong. The 90-day stat is the mirror: median +5.6% with 59% win rate suggests holding through the noise pays. Do not chase the earnings gap — wait 1–2 weeks for the fade.

---

13. Risk metrics — institutional benchmarks

The Sortino-to-Sharpe ratio of 1.52 confirms what every investor already feels: TSLA's distribution is right-skewed — losses are clustered around modest negatives while the gains include rare massive outliers (2013, 2020). That's why holding through drawdowns has paid lifetime, but it's also why Calmar is unimpressive — those rare upside outliers don't compensate intra-period for the depth of the drawdowns. Up/down capture isn't computed (rolling beta hit insufficient history) but the 1.57× SPY beta at full-period level + 24% R² is the load-bearing number: TSLA is 1.6× index beta plus a fat tail of name-specific noise.

---

14. Year-by-year + holding-period grid

The hold-from-2021 column is the single most important number on the page: a buyer at the start of 2021 has compounded at 9.4% per year since — barely above SPY — despite a 720% rally and a 130% rally happening during their hold. The 2022-buyer is essentially flat to today (CAGR −0.5%). Translation: the real money was made by holders who entered before the Dec-2019 inflection; everything since has been extremely volatile flat-to-modest. The 2026 YTD CAGR of −30.5% is just the current drawdown from the Dec-2025 ATH.

---

15. Seasonality

Best month: November (+0.47% avg daily, 53.5% win rate). Worst month: March (−0.025% avg daily, 46.8% win rate). The June + November combination — Q2-end and post-election year-end positioning — is the seasonality edge; March is the only month with a negative average daily return across 16 years. Magnitudes are modest (<0.5% avg daily) and not large enough to override regime/factor signals.

---

16. Volume profile + anchored VWAP — where the action sits

The point-of-control sits at $15.71 (the $10.83–$20.60 bucket holds 30% of all lifetime TSLA volume — that's the 2014–2019 sub-$25 grind). At $390.82 today, the stock is trading in a thin volume air-pocket (current bucket is only 0.36% of lifetime volume). There is no meaningful prior-volume support between today's price and the $245 area — implying that if $343 (the April-8 low) breaks, the slide can be fast and gap-filled down to the $235–$245 high-volume node.

Spot is above every anchored VWAP except the ATH-anchored one (−5.1% below). The IPO-anchored VWAP at $110 is irrelevant defensively (price would need to fall 72% to test it), but the ATH-anchored VWAP at $412 is the live battle line — every buyer since Dec-16 is, on average, underwater. Reclaim of $412 = capitulation has ended; failure to reclaim = sellers from the all-time-high zone still in control.

---

17. Liquidity — execution capacity

Daily turnover is $24.5B = 1.78% of market cap per day — TSLA is among the deepest single-name pools on earth. A 1% issuer position is exitable in 6 days at conservative 10% ADV participation. The 60-day median daily range is 1.46% — execution slippage is negligible at any institutional size below ~2% mcap. Liquidity is not a constraint here at any AUM under $500B. Conviction, not capacity, is the binding limit.

---

18. Stance + invalidation

Stance — Neutral with downside skew over 90 days, neutral over 180 days. The base case from this regime is mildly negative drift (median 90d = −1%, expected = −1.7%, P(positive) = 42.6%). The §1 statistical character finding constrains how confident we can be: daily TA has no edge here, so the case rests on the combination of factor exposure (1.6× SPY beta), conditional base rates (49% hit rate at 90d, 58.5% at 180d), and structural patterns (broken HH/HL). The 180-day regime base rate is the only reason to not be outright bearish.

Invalidation levels:

• Bullish confirmation: above $430. That reclaims both the 200-day SMA ($401), the ATH-anchored VWAP ($412), and the prior swing-high cluster ($419/$435), and would confirm the death cross was a fakeout. Above $430, the 90d touch probability of +20% ($469) becomes the working target.
• Bearish confirmation: below $343. Loses the April-8 swing low. Volume profile shows no support until ~$245 (the 4th-densest HVN bucket), so this is a high-velocity zone — the bootstrap implies 32% probability of touching $313 (−20%) within 90 days, but a clean break of $343 makes that figure structurally higher.

Implementation: Liquidity is not the constraint. Action: wait — the regime base rate doesn't reward chasing here; let the stock either reclaim $430 (buy strength) or break $343 (short the breakdown). Patient sizing within the half-Kelly ceiling, not at it.