Crypto-Adjacent Short-Timeframe Trade Setup Analysis

Mar 5, 2026 Quant Researcher

Crypto-Adjacent Short-Timeframe Trade Setup Analysis

Date: 2026-03-05 Tickers: BTC-USD, ETH-USD, BMNR, MSTR Type: Swing Trade Setup Research

1. Hypothesis

Primary hypothesis: Crypto assets are attempting to form a bottom after a significant drawdown, and there exists a short-timeframe swing trade opportunity with favorable risk/reward.

Sub-hypotheses tested:

  1. The recent price action shows buying strength consistent with accumulation rather than a dead-cat bounce.
  2. Historical analogs of similar technical setups suggest positive forward returns over 1-4 weeks.
  3. MSTR and BMNR provide leveraged exposure to BTC price movements, making them viable equity proxies for a crypto swing trade.
  4. The current bounce is not yet overbought to the point where chasing carries poor expected value.

2. Data

Ticker Source Date Range Trading Days Notes
BTC-USD data-samples/ohlcv/BTC-USD.csv 2018-01-01 to 2026-03-05 2,985 Crypto trades 7d/week
ETH-USD data-samples/ohlcv/ETH-USD.csv 2018-01-01 to 2026-03-05 2,985 Crypto trades 7d/week
BMNR data-samples/ohlcv/BMNR.csv 2025-06-05 to 2026-03-04 187 Severe data limitation
MSTR data-samples/ohlcv/MSTR.csv 2018-01-02 to 2026-03-04 2,053 Equity market hours only

Data cleaning: None required. OHLCV data is clean with no gaps.

Critical limitation: BMNR has only 187 trading days of history. All statistical conclusions for BMNR should be treated as unreliable and directional only. This is flagged throughout the report.

3. Methodology

  1. Technical indicator computation: RSI(14), MACD(12,26,9), Stochastic(14,3), Bollinger Bands(20,2), ATR(14), OBV, A/D Line, multiple SMAs (20/50/200), historical volatility (20-day).

  2. Buying strength assessment: OBV trend (linear regression), A/D line trend, up/down volume ratio, close position within daily range, higher lows count.

  3. Historical analog matching: Identified periods with similar technical conditions (RSI range, BB %B range, trend context) and computed forward returns at 5, 10, and 20 trading day horizons.

  4. Bear bounce episode analysis: Isolated all distinct episodes where BTC was >30% below ATH and bounced >5% in 10 days, then tracked forward returns and max drawdowns.

  5. Cross-asset dynamics: Correlation analysis (full-period and rolling 30-day), beta regression (OLS), ETH/BTC ratio trend decomposition.

  6. Bottom formation scoring: Five-signal composite: higher lows, declining sell volume, volatility compression, RSI bullish divergence, OBV leading price.

  7. Overbought timing risk: Historical forward returns when RSI>60, Stochastic>80, and price below SMA-50 simultaneously.

All computations performed in Python. No numbers were estimated manually.

4. Results

4.1 Current Technical Snapshot

Metric BTC-USD ETH-USD MSTR BMNR
Price $72,658 $2,122 $146.44 $21.12
Latest daily return +6.4% +7.0% +10.4% +7.9%
RSI(14) 64.3 60.4 62.6 56.6
Stochastic %K 95.2 93.2 90.0 70.2
MACD Histogram +1,137 +44.2 +2.92 +0.45
MACD Cross Bullish Bullish Bullish Bullish
BB %B 1.17 1.02 1.01 0.79
vs SMA-20 +7.4% +7.0% +12.5% +5.5%
vs SMA-50 -4.8% -10.8% -0.9% -17.4%
vs SMA-200 -24.6% -37.0% -47.8% N/A
ATR % 4.3% 5.7% 6.5% 7.2%
Volume ratio 1.81x 1.55x 1.44x 1.39x
Regime Bear Trend Bear Trend Bear Trend Unknown

Key observation: All four tickers had strong up days on the latest bar with above-average volume (1.4x-1.8x). All four have bullish MACD crosses. However, all three with sufficient history classify as Bear Trend (below both SMA-50 and SMA-200, with MAs in bearish alignment). The Stochastic is in overbought territory (>80) for BTC, ETH, and MSTR.

4.2 Buying Strength Assessment

Metric BTC ETH MSTR BMNR
OBV 20-day trend Falling (R2=0.55) Falling (R2=0.75) Flat (R2=0.00) Rising (R2=0.65)
A/D Line 20-day Accumulation Accumulation Accumulation Accumulation
Up/Down volume ratio (20d) 1.02x 0.85x 0.96x 1.19x
Higher lows (10 bars) 6/9 6/9 4/9 4/9
Close position in range (5d) 56.8% 51.8% 61.8% 39.3%

Assessment: The buying strength evidence is mixed. The A/D line shows accumulation across the board, but OBV -- which weights volume by price direction more heavily -- is falling for BTC and ETH. The up/down volume ratios are essentially balanced. This is consistent with a counter-trend bounce rather than the start of a new uptrend. The bounce is real, but the underlying flow picture is not yet convincingly bullish.

4.3 Bottom Formation Evidence

Signal BTC ETH MSTR BMNR
Higher lows (20d vs prior 20d) YES YES NO NO
Declining sell volume YES YES NO NO
Volatility compression YES NO NO NO
RSI bullish divergence (40d) YES YES YES NO
OBV leading price NO NO NO NO
Score 4/5 (80%) 3/5 (60%) 1/5 (20%) 0/5 (0%)

Drawdown context:

Ticker From ATH 52wk Range Position
BTC -41.8% 19.0%
ETH -56.1% 11.7%
MSTR -69.1% 12.0%
BMNR -84.4% (from $161 peak) 11.0%

Assessment: BTC shows the strongest bottom formation evidence (4/5), driven primarily by a textbook RSI bullish divergence -- price made lower lows over 40 days while RSI made significantly higher lows (+18.2 points). ETH has moderate evidence (3/5). MSTR and BMNR show no meaningful bottoming evidence beyond RSI divergence in MSTR's case. All four sit near the bottom of their respective ranges, which is a necessary but not sufficient condition for a bottom.

The critical missing ingredient is OBV confirmation. In none of the four assets is OBV leading price higher, which is historically a prerequisite for a sustainable bottom.

4.4 Historical Analog Forward Returns

BTC-USD: Bear Bounce Analogs (below SMA-50, RSI 50-75, 10d ROC >5%)

Horizon n Mean Median Win Rate 25th-75th
1 Week 88 -1.55% -2.11% 37.5% [-6.1%, +3.2%]
2 Weeks 88 -0.30% -0.27% 48.9% [-5.9%, +7.0%]
4 Weeks 88 -0.42% -0.89% 46.6% [-8.3%, +9.2%]

BTC-USD: Deep Drawdown Bounce Episodes (>30% from ATH, 10d ROC >5%)

28 distinct episodes since 2018.

Horizon n Mean Median Win Rate Worst Best
1 Week 28 +2.1% +0.1% 53.6% -14.3% +19.7%
2 Weeks 28 +3.9% +5.7% 60.7% -15.4% +30.0%
4 Weeks 28 +3.5% +6.3% 57.1% -35.2% +30.9%
8 Weeks 28 +8.4% +0.9% 53.6% -31.9% +65.9%

ETH-USD: Bear Bounce Analogs (>40% from ATH, RSI 50-70, 10d ROC >5%)

Horizon n Mean Median Win Rate 25th-75th
1 Week 331 +1.7% +0.1% 50.8% [-4.6%, +7.2%]
2 Weeks 331 +3.5% +1.2% 55.0% [-7.5%, +12.0%]
4 Weeks 331 +6.7% +2.2% 54.7% [-8.0%, +18.5%]

MSTR: Overbought-in-Downtrend (RSI>60, Stoch>80, below SMA-50)

Horizon n Mean Median Win Rate
5 Days 23 -1.56% -2.32% 39.1%
10 Days 23 +0.24% +0.10% 52.2%

Assessment of historical analogs:

The data tells a nuanced story. The broad analog (any time BTC bounces 5%+ while below SMA-50) has negative expected value at 1 week (-1.55% mean, 37.5% win rate). This is a strong caution signal. These bounces tend to give back gains quickly.

However, the deep drawdown episode analysis is more encouraging. When BTC is >30% below ATH and bounces sharply, the 2-week and 4-week medians are positive (+5.7% and +6.3% respectively) with ~60% win rates. The current -41.8% drawdown fits squarely in this category.

The difference in outcomes likely reflects that deep drawdown bounces capture genuine capitulation-and-recovery sequences, while shallow downtrend bounces include many false starts in trending bear markets.

The critical caveat: Even in deep drawdown episodes, the worst outcomes are severe (-35% over 4 weeks, -32% over 8 weeks). The distribution is wide and right-skewed -- a few massive recoveries pull up the mean while the median tells a more conservative story.

4.5 Immediate Timing Risk: Overbought Assessment

BTC and MSTR are currently overbought on shorter timeframes (RSI>60 AND Stochastic>80) while remaining in a bear trend (below SMA-50). Historical data for this specific condition:

Ticker 5-day fwd mean 5-day win rate 10-day fwd mean 10-day win rate
BTC -1.21% 40.0% +0.51% 38.3%
ETH +1.87% 52.5% +3.66% 52.5%
MSTR -1.56% 39.1% +0.24% 52.2%

This is the most actionable finding: Entering immediately at current levels when BTC and MSTR are overbought in a downtrend has historically produced negative 5-day forward returns with win rates below 40%. ETH shows modestly better behavior, though not compelling enough to justify aggressive entry.

The implication is clear: waiting for a pullback improves expected value.

4.6 Cross-Asset Dynamics

Pair Full-Period Corr 30-Day Corr Beta (90d) R-squared
BTC/ETH 0.83 0.96 -- --
BTC/MSTR 0.51 0.91 1.45 0.71
BTC/BMNR 0.05 0.91 0.91 0.002

Key findings:

  1. MSTR is delivering leveraged BTC exposure as expected. The 90-day beta of 1.45 means MSTR amplifies BTC moves by ~45%. The R-squared of 0.71 is strong, confirming MSTR as a high-fidelity BTC proxy. Today's +10.4% move vs. BTC's +6.4% is consistent with this beta.

  2. BMNR has virtually zero statistical relationship to BTC over its full history (R2=0.002, correlation=0.05). However, the 30-day rolling correlation has surged to 0.91. This suggests BMNR only correlates with BTC during crypto momentum regimes and trades idiosyncratically otherwise. This is important: BMNR is not a reliable BTC proxy.

  3. ETH/BTC ratio is at multi-month lows (0.0292 vs. 90-day avg of 0.0323, 52-week range 0.0226-0.0421). ETH is weakening relative to BTC, with a 30-day trend slope that is negative. ETH is underperforming BTC in this bounce.

  4. Correlations are highly elevated in the last 30 days (0.91-0.96), suggesting crypto assets are moving in lockstep. This is typical of stress/recovery regimes and reduces diversification benefit.

4.7 Pullback Targets (If Waiting for Entry)

Level BTC-USD ETH-USD MSTR
SMA-20 $67,663 (-6.9%) $1,983 (-6.6%) $130.17 (-11.1%)
SMA-50 $76,360 (+5.1%) $2,379 (+12.1%) $147.81 (+0.9%)
BB Lower Band $63,908 (-12.0%) $1,849 (-12.9%) $114.35 (-21.9%)
30-day Low $60,074 (-17.3%) $1,749 (-17.6%) $104.17 (-28.9%)
60-day Low $60,074 (-17.3%) $1,749 (-17.6%) $104.17 (-28.9%)

Note that BTC and ETH are both currently above their SMA-50 only in the case of... actually, no. BTC is at $72,658 vs. SMA-50 at $76,360 -- still below by 4.8%. The SMA-20 at $67,663 is the nearest pullback target.

4.8 Recent Performance Context

Period BTC ETH MSTR BMNR
1 Day +6.4% +7.0% +10.4% +7.9%
5 Days +10.3% +9.9% +8.0% -1.8%
10 Days +7.4% +8.4% +17.0% +6.6%
20 Days +9.7% +9.0% +9.9% -5.5%
60 Days -19.8% -32.1% -21.3% -41.8%
90 Days -21.2% -32.3% -47.9% -56.0%

The bounce is sharp -- BTC and ETH are up ~10% in 5 days -- but comes after devastating 60-90 day drawdowns. MSTR has amplified both the downside (-47.9% over 90 days) and the recent upside (+17.0% over 10 days), as expected from its 1.45 beta. BMNR's 5-day return is actually negative despite the latest +7.9% up day, highlighting its idiosyncratic volatility.

5. Signal Quality

5.1 Bottom Formation Signal

  • False positive risk: MODERATE to HIGH. The 80% bottom formation score for BTC is encouraging, but OBV non-confirmation is historically the signal that separates real bottoms from bear market rallies. In the 2018-2019 bear market, BTC scored similarly on bottom formation metrics multiple times before the actual bottom at ~$3,200.

  • Lead time: POOR. Even when a genuine bottom is forming, the process typically takes weeks to months. Wyckoff accumulation phases include multiple tests of support. The data does not show clear evidence that we are past the "spring" phase of accumulation.

  • Robustness: MODERATE for BTC, WEAK for others. BTC's bullish RSI divergence is the strongest signal (price -18.5% over 40 days while RSI rose +18.2 points). ETH shows a similar pattern. MSTR and BMNR do not show convincing bottom formation.

5.2 Swing Trade Entry Signal

  • False positive risk: HIGH at current levels. Overbought-in-downtrend conditions historically produce 5-day forward returns of -1.2% to -1.6% for BTC and MSTR with sub-40% win rates. Entering now is chasing.

  • Quality improves on pullback. The deep drawdown bounce analog data suggests that if the bounce can establish support and pull back to test it, the second leg up has better expected value. The optimal entry zone would be a pullback to the SMA-20 area.

  • Signal-to-noise ratio: LOW for BMNR. With 834% annualized volatility, 52% daily standard deviation, and a kurtosis of 164, BMNR's price behavior is dominated by noise. Any technical signal is unreliable.

6. Recommended Architecture: Swing Trade Strategy Sketch

Based on the evidence, I recommend against chasing the current move. However, the data supports a pullback entry strategy if the bottom formation thesis continues to develop. Here are two strategy sketches:

Strategy Sketch A: "Bear Bounce Pullback Entry" (BTC-USD / ETH-USD)

Thesis: After a sharp bounce from deep drawdown, wait for a pullback to a support level before entering. Historical analogs suggest 2-4 week forward returns are modestly positive from deep drawdown bounces.

Entry Conditions (ALL must be true):
  1. Price is below SMA-200 (confirms we are in a bear market)
  2. RSI(14) pulled back to 40-55 range (from current overbought levels)
  3. Price is within 3% of SMA-20 (mean reversion pullback target)
  4. MACD histogram remains positive or has turned positive
  5. Volume on pullback day is < 0.8x 20-day average (low-volume pullback)

Exit Conditions (ANY triggers exit):
  1. Take profit: +15% from entry (2-3 ATR move)
  2. Stop loss: Close below 60-day low (BTC: ~$60,074, ETH: ~$1,749)
  3. Time stop: 20 trading days if neither target nor stop hit
  4. RSI reaches 75+ (overbought in downtrend, take profit)

Position Sizing:
  Max risk per trade: 1-2% of account
  ATR-based sizing: Position = Risk$ / (2 * ATR)

Parameter Ranges to Test:
  RSI entry range: [35-55] by 5
  SMA proximity: [2%-5%] by 1%
  Profit target: [10%-20%] by 2.5%
  Lookback for SMA: [15-25] days

Strategy Sketch B: "MSTR Leveraged BTC Swing" (MSTR)

Thesis: MSTR provides ~1.45x leveraged BTC exposure with R-squared of 0.71. Use MSTR as an equity-market proxy for BTC swing trades, gaining the benefit of equity market hours and infrastructure.

Entry Conditions (ALL must be true):
  1. BTC RSI(14) in 40-55 range (pullback from overbought)
  2. MSTR within 5% of its SMA-20
  3. MSTR daily volume > 1.2x 20-day average (institutional interest)
  4. BTC MACD histogram positive (momentum backdrop)

Exit Conditions:
  1. Take profit: +20% from entry (accounts for 1.45x beta amplification)
  2. Stop loss: -12% from entry (tight stop given leverage)
  3. Time stop: 15 trading days
  4. BTC RSI reaches 70+ (upstream overbought)

Risk Management Note:
  MSTR carries company-specific risk (balance sheet, BTC holdings,
  dilution). The 1.45 beta cuts both ways. Current annualized
  volatility of 132.7% means daily swings of 8%+ are normal.

Parameter Ranges to Test:
  BTC RSI entry range: [35-55] by 5
  MSTR SMA proximity: [3%-7%] by 1%
  Profit target: [15%-25%] by 2.5%
  Stop loss: [8%-15%] by 1%

Strategy Sketch C: "BMNR Opportunistic" -- NOT RECOMMENDED

BMNR fails the minimum requirements for a systematic strategy: - R-squared to BTC of 0.002 (no reliable relationship) - 834% annualized volatility (noise dominates signal) - 187 days of history (insufficient for any statistical inference) - 0/5 bottom formation score - -84.4% from ATH with no higher lows

If a position is desired despite these limitations, it should be treated as a high-risk speculative bet sized accordingly (0.5% of account or less), not as a swing trade with definable edge.

7. Limitations

Data Limitations

  1. BMNR history (187 days) is grossly insufficient for any meaningful statistical analysis. The daily return standard deviation of 52.55%, skew of 12.50, and kurtosis of 164 indicate extreme non-normality. All BMNR conclusions should be treated as speculation, not analysis.

  2. Analog sample sizes are modest. The deep drawdown bounce analysis has 28 distinct episodes -- enough for directional insight but not for high-confidence statistical inference. The p-values on forward return means being different from zero would not clear standard significance thresholds.

  3. Crypto data includes weekend trading while equity data (MSTR, BMNR) does not. Cross-asset comparisons on common trading days lose information about weekend crypto price movements.

Methodological Limitations

  1. Analog matching is imprecise. The technical conditions used for matching (RSI range, BB %B range, trend context) are somewhat arbitrary. Tightening criteria reduces sample size; loosening criteria reduces relevance.

  2. No volume profile analysis. The support/resistance identification uses simple price clustering via KDE rather than volume-at-price analysis. Volume profile data (which would show where the most trading occurred) would improve level identification.

  3. No order flow or on-chain data. This analysis uses only price and volume data. On-chain metrics (exchange balances, whale movements, miner revenue), futures funding rates, and options skew would provide additional color on the bottom formation thesis.

Market Structure Limitations

  1. Regime uncertainty. The current macro backdrop (tariff policy, interest rates, regulatory environment) may create conditions that have no historical analog. Past bear bounce episodes may not be predictive if the current drawdown is driven by fundamentally different forces.

  2. MSTR company-specific risk. MSTR's beta to BTC depends on its balance sheet, convertible note structure, and BTC holdings. These factors can cause MSTR to diverge from BTC in ways the beta model does not capture.

  3. Liquidity risk. BMNR's average daily volume of ~42M shares at ~$21 implies ~$880M daily dollar volume, which seems large but includes significant HFT/market maker activity. Actual available liquidity for medium-sized positions may be lower.

Charts

All charts saved to analysis/quant-research/charts/:

Chart File Content
Price Structure crypto_swing_price_structure.png 6-month price with MAs and BBs for all 4 tickers
Momentum crypto_swing_momentum.png RSI, MACD, Stochastic for BTC and ETH (90 days)
Volume crypto_swing_volume.png Volume bars, 20d MA, OBV for all 4 tickers (90 days)
Cross-Asset crypto_swing_cross_asset.png Normalized returns, ETH/BTC ratio, rolling correlations
Bottom Formation crypto_swing_bottom_formation.png RSI divergence, drawdown history, volatility
BTC Proxies crypto_swing_proxies.png MSTR and BMNR price structure, return scatter vs BTC

Summary of Findings

Is the user's intuition correct?

Partially. The data confirms that crypto (BTC and ETH) is bouncing sharply from deep drawdown levels with some accumulation signals. BTC scores 80% on the bottom formation checklist, driven primarily by a strong RSI bullish divergence and declining sell volume. This is consistent with the early stages of a bottoming process.

However, the data does not confirm this is a good entry point right now. The current bounce has pushed BTC and MSTR into overbought territory within an intact bear trend. Historical analogs of similar overbought-in-downtrend conditions show negative 5-day forward returns with sub-40% win rates for BTC and MSTR. The risk of a pullback is elevated.

Probabilistic assessment of outcomes

BTC-USD forward return probabilities (based on deep drawdown bounce episodes, n=28):

Horizon Bull Case (75th pctile) Base Case (median) Bear Case (25th pctile)
1 Week +4.4% (~$75,800) +0.1% (~$72,700) -6.1% (~$68,200)
2 Weeks +8.6% (~$78,900) +5.7% (~$76,800) -5.8% (~$68,400)
4 Weeks +11.0% (~$80,600) +6.3% (~$77,200) -10.5% (~$65,000)

Most likely near-term path: A pullback in the next 5-7 trading days toward the SMA-20 ($67,663 for BTC), followed by either a successful retest of recent lows (bullish) or a break below them (bearish). The 60.7% win rate at 2 weeks and 57.1% at 4 weeks is modestly bullish, but the fat left tail (-35% worst case over 4 weeks) demands respect.

Bottom line

The bounce is real but may be getting ahead of itself. The strategy with the best expected value is to wait for a pullback to the SMA-20 / Fibonacci 61.8% level and enter on a confirmed hold of that support. Chasing the current extended move has negative short-term expected value based on historical analogs.

For MSTR: Same logic applies with amplified magnitude. MSTR is the best proxy for levered BTC exposure (beta 1.45, R2 0.71) but carries additional company-specific risk.

For BMNR: The data is insufficient to make any evidence-based swing trade recommendation. The asset is extremely volatile (834% annualized), has no reliable relationship to BTC over its full history, and shows zero bottom formation signals. Any position should be sized as high-risk speculation.

For ETH: ETH shows slightly better forward return analogs than BTC (55% win rate at 4 weeks vs. 50% for BTC in broad analogs), but is underperforming BTC in this bounce (ETH/BTC ratio at multi-month lows and declining). If taking a crypto position, BTC appears to be the stronger horse in this cycle.

Analysis Script

Full analysis code saved to: analysis/quant-research/scripts/crypto-swing-setup-2026-03-05.py