The Dow Jones Industrial Average just closed its third consecutive year of double-digit gains. By any gut-feel metric, this should be a screaming sell signal. Yet the data says otherwise — and the same statistical framework that challenges the 'too high, must crash' narrative on Wall Street also exposes a dangerous blind spot for crypto traders who treat every rally as a prelude to a rug.
I have spent the last decade dissecting market structure across both traditional and crypto markets. The empirical bias is the same: narratives are noise, order flow is signal. Mark Hulbert's recent analysis in MarketWatch caught my attention not because it is new, but because it is a rare example of someone applying rigorous statistical thinking to a question that is usually answered with fear or greed.
Hulbert's core finding: since 1896, the Dow has experienced 14 three-year winning streaks of double-digit annual returns. In the year following such streaks, the index still posted a double-digit gain 49% of the time. There is no statistical basis for expecting a crash. The probability of a 40%+ drawdown over the next two years, according to a model from Harvard and HKU, sits at 19% — actually below the 5-year average of 26%.
This is not a bullish call. It is a call against the gambler's fallacy — the belief that a long streak of wins must be followed by a loss. In crypto, this fallacy is amplified by the 24/7 nature of the market, the addiction to leverage, and the constant stream of 'top is in' tweets. Every time Bitcoin rallies past its previous all-time high, the chorus of 'this time it's different' and 'we are due for a crash' grows louder. Hulbert's framework suggests both sides are statistically illiterate.
But here is where the crypto context matters. The 129-year Dow dataset includes multiple monetary regimes, fiscal shocks, and technological revolutions. The 49% unconditional probability masks a critical conditional probability: what happens when valuations are at the 99th percentile? The Shiller CAPE ratio for the S&P 500 is currently around 37, a level exceeded only in 1929 and 2000. In crypto, the equivalent metric is the Mayer Multiple (Bitcoin price vs. 200-day moving average) or the MVRV Z-score (market value vs. realized value). As of late 2025, Bitcoin's MVRV Z-score was around 3.0, a level that historically preceded drawdowns of 50%+ within 12 months.
Volatility is just noise waiting to be priced.
Hulbert himself acknowledges the limitation: his model does not incorporate valuation. This is not a flaw in the statistical method — it is a deliberate choice to isolate the question 'does the length of a winning streak alone predict crashes?' The answer is no. But the moment you ask 'does a winning streak combined with extreme valuations predict crashes?' the answer changes. The Harvard/HKU model partially addresses this by conditioning on the prior two years of returns, but it still does not include absolute valuation levels.
From my experience auditing smart contracts and trading options, I have seen the same mistake repeated across cycles. In late 2020, when Bitcoin broke $20,000 and then $30,000, the same 'too far, too fast' narrative was used to justify short positions. Those shorts were liquidated as Bitcoin marched to $64,000. In late 2021, when Bitcoin was at $68,000, the 'this time is different' narrative was used to justify long positions. Those longs were destroyed by the 2022 bear market. The common thread is not the direction of the move — it is the failure to distinguish between unconditional and conditional probabilities.
What crypto traders need to internalize: the 49% probability of double-digit gains does not mean you should be yoloing into levered longs. It means the market is not statistically 'due' for a crash. But the risk of a crash is still present — and the asymmetric payoff of options makes them the only rational tool for navigating this environment.
Options give you the right to walk away.
Consider the current state of Bitcoin options implied volatility. The term structure is in backwardation — short-dated IV is higher than long-dated IV — which is typical during periods of uncertainty around regulatory events or ETF flows. But the skew is flat, meaning puts and calls are priced similarly. This is unusual for a market that has just experienced a triple-digit rally from the 2022 lows. In traditional finance, a flat skew after a long rally signals that the market is 'pricing in no crash risk.' That is precisely the kind of environment where a tail hedge (buying out-of-the-money puts) is cheap relative to the potential payoff.
The floor is a suggestion, not a law.
The State Street model that Hulbert references shows a 19% probability of a 40%+ drawdown over two years. To put that in perspective: the probability of a credit event in a BB-rated bond is typically around 5% over one year. A 19% probability means that roughly one in five investors will experience a catastrophic loss if they remain fully exposed. That is not a tiny risk. That is a risk that demands active management through position sizing, portfolio hedging, or both.
Where does this leave crypto? The AI narrative that drove the 2024-2025 stock rally has its parallel in crypto: the AI-agent token frenzy, the GPU-backed tokenization, and the 'inference economy' narratives. The same concentration risk is present. Just as the top 10 stocks in the S&P 500 now account for 38% of the index, the top 10 tokens by market cap (excluding stablecoins) account for over 70% of crypto total value. If the AI narrative in equities falters, the spillover into crypto AI tokens could be severe. And if the broader macro environment shifts — a resurgence of inflation, a Fed rate hike, a liquidity crisis — the 19% model may prove optimistic.
Chaos is just data with no label yet.
My own experience with the Terra/Luna collapse taught me that the most dangerous moments are when the crowd is convinced the system is safe. Before the depeg, the UST-LUNA pair was paying 20% yields. The implied volatility in options was near zero. The market was pricing in a perfect world. That was the signal to short. Today, the Bitcoin options market is pricing in a world where the only risk is a gradual correction. The skew is flat, the term structure is backwardated, and the VIX is low. This is the same pattern I saw in early 2022 before the cascade.
To be clear: I am not predicting a crash. I am predicting that the market's current pricing of tail risk is too low. Hulbert's statistical framework is a useful corrective to the 'too high, must crash' crowd, but it should not be misinterpreted as a green light for unhedged longs. The correct response is to acknowledge the 49% probability of double-digit gains and the 19% probability of a 40% drawdown, and then position accordingly. That means carrying a portfolio that can survive both outcomes — a barbell of Bitcoin spot exposure hedged with out-of-the-money puts, or a short volatility position that benefits from the 49% scenario while capping downside.
Actionable Levels:
- If Bitcoin holds above $95,000, the 49% scenario is alive. Look for continuation to $115,000-$120,000.
- If Bitcoin loses $85,000, the 19% scenario becomes active. A drop to $60,000 is possible within 6 months.
- The key level to watch is the 200-day moving average (currently around $80,000). A break below that would confirm that the 19% tail risk is materializing.
In the end, the market does not care about your narrative. It only cares about the order flow. Hulbert's data tells us that the order flow is not statistically biased toward a crash. But the flows are increasingly concentrated in a few names, and the valuation backdrop is extreme. That is the recipe for a volatility event — not a crash, but a violent repricing that will punish those who are not prepared.
Liquidity vanishes the moment you need it most.
The next time you hear someone say 'the market is due for a correction,' ask them to show you the conditional probability. The answer is 49% — same as any other year. The market does not owe you a crash just because it has been good. But it also does not owe you a continuation. The only reliable strategy is to price the volatility and let the market prove you wrong.
