How to navigate extreme market patterns

Academic studies confirm Indian index returns follow power-law, fat-tailed distributions, similar to global markets. Large daily moves are much more common

The frantic calls from friends and clients are familiar during times of market stress. "Where will the market open?" "Should I switch everything to gold?" These reactions are typical when investors are overwhelmed by noise, seeking certainty where none exists.

While we often comfort ourselves with historical averages and statistical models, these tools frequently fail to capture emotional swings and, more critically, the true nature of market risk. The standard approach of viewing returns and risk through mean and standard deviation assumes an orderly world. It forgets that extremes define the experience.

A fundamental assumption in finance is that market returns follow a normal distribution (the bell curve). This model suggests that extreme daily moves are extraordinarily rare, but this ‘normal’ market is an illusion. For instance, a move beyond 3 standard deviations (σ) should have a probability of just 0.13 per cent.

In reality, financial markets, including India's, systematically violate this assumption. They exhibit "fat tails”, meaning extreme events occur far more frequently than a Gaussian model predicts.

Indian market history is a testament to these fat tails:

During Covid-19, the Sensex saw single-day swings exceeding ±9 per cent.

February 2025 witnessed one of the year’s largest intraday drops.

April 2025 included sessions with tariff-led declines greater than 3 per cent.

These are not statistical impossibilities; they are expected outcomes in a fat-tailed world. Translating annual volatility to daily terms shows that a ±3–4 per cent move is already a large multi-sigma event under a normal model. Yet, during stress, moves of ±5 per cent to ±10 per cent happen regularly. Howard Marks said, “There’s a big difference between probability and outcome. Probable things fail to happen – and improbable things happen – all the time.”

The fallacy of normal-distribution thinking becomes starkly clear with events like the U.S. stock market’s +9.5 per cent rally on April 9, 2025. Under a standard bell curve, this was labeled an "eight-sigma event," implying a probability of 1 in 1.6 quadrillion—a practical impossibility. Yet it happened. Indian markets tell a similar story: what theory deems "unthinkable" occurs with sobering regularity because the underlying distribution is not normal but possesses heavy tails.

Academic studies confirm Indian index returns follow power-law, fat-tailed distributions,similar to global markets. Large daily moves are much more common. A3 per cent day, like Sensex experienced on April 7, 2025, might be a –3σ "rare" event in a naive model, but reality shows such moves cluster during volatile periods, shattering the eight-sigma mirage.

This fat-tailed reality reshapes the investor experience. Over decades, Indian markets trend upward. However, the journey is dominated by volatility and drawdowns, not steady gains.

For example, from 2007–2017, the Nifty delivered a modest ~5.5 per cent CAGR, barely above inflation, with the deep scar of the 2008 crash defining the period.Absolute long-term gains are less informative than the volatility endured to achieve them.

History doesn't promise specific future returns, but it offers crucial patterns:

Crashes are painful but not unprecedented. The Sensex has experienced multiple declines of ~40–60 per cent, followed by recoveries. Recoveries do come, though their speed depends on economic conditions and sentiment. They too are fat-tails and hence staying in the market is essential.

Daily volatility often overestimates long-term risk while underestimating the frequency of extreme shocks. This understanding leads to a pivotal shift in mindset, very crucial for investors. If you believe returns are normal - you will overestimate safety, underestimate potential drawdowns, and likely panic when inevitable extremes occur i.e., selling at the worst time

.If you accept fat tails - you acknowledge that crashes are part of the system, not aberrations. This focuses strategy on survival rather than prediction. Position sizing, asset allocation aligned with personal risk appetite and timelines, and emotional discipline become far more important than forecasting tomorrow's move.

Practical implications for investors today

Expect averages, but don't rely on them: The long-term mean return is a guide, not a guarantee. Actual paths are highly skewed and volatile.

Prepare for more frequent extremes: Large daily moves (±3σ or more) are structural features of markets, not once-in-a-lifetime anomalies. Portfolios must be built to withstand them without triggering impulsive actions.

Focus on process over prognostication: When friends ask for directional guesses, "IDK” (I don't know) is a rational answer. The productive response is to revisit one's financial plan: Is the portfolio's exposure proportionate to goals? Is the asset allocation consistent with risk capacity? This is the work done to "put the lid off" during panics.

Recognise that markets are not broken: The Sensex didn't die in 2008 or 2020. Its survival through repeated extreme events confirms that markets are fat-tailed systems, not Gaussian ones. The "sigma-based thinking" used to label events as impossible is itself the flawed model.

In conclusion, the madness of crowds and the frenzy of headlines are distractions from a more stable truth: markets are inherently prone to extremes. By accepting this statistical reality—that "one-in-a-quadrillion" events happen with observable frequency—investors can move away from futile speculation and towards a robust portfolio construction. The goal shifts from trying to outguess every swing to ensuring you remain positioned to participate in the long-term trend, regardless of the inevitable jagged, unpredictable, and extreme deviations along the way.

(The author is a partner with “Wealocity Analytics”, a SEBI registered Research Analyst firm and could be reached at [email protected])