Across 30 years of Nifty 50 data, an investor waiting for even a modest 10% correction before investing didn't see one arrive within five years 34% of the time. Wait for 15%, the miss rate is 50%. Wait for 20%, it is 59%. And in those waiting windows, the index averaged a 133% gain — all of it caught by the investor who deployed, all of it missed by the one who waited. This paper measures the cost of holding cash in hope of a better entry. On the data, it is the single most expensive habit in Indian investing — bigger than fund-selection mistakes, bigger than SIP-day mistakes, bigger than nearly anything else a retail investor does wrong.
Every investor has had the thought. "The market feels expensive. I'll wait for a correction. Then I'll invest." It sounds disciplined. It feels safe. It quietly costs more than almost any other mistake in personal finance.
Cash earns roughly 6.5% in a liquid fund. Indian equity has compounded at roughly 12% over three decades. The gap is about 5.5 percentage points a year. Over one year, you barely notice. Over five, it becomes most of the wealth you were trying to build.
The academic literature has said this clearly for decades. Vanguard (2012) found that lump-sum investing beat dollar-cost averaging in two out of three historical windows across US, UK, and Australian markets. Charles Schwab (2021) found that even an investor who deployed at the worst day of every year still beat one who held cash. The result holds across geographies, time periods, and methods. And yet retail behaviour hasn't changed.
The cleanest test is to just run the strategy. Pick a starting day. You have ₹1 Cr. You decide to wait for a 10%, 15%, or 20% drop from today's level before investing. Until then, your money earns 6.5% in a liquid fund.
If the drop arrives, you invest. If five years pass without it arriving, you stay in cash. Then compare your final wealth to a simpler strategy: just invest on day one and do nothing.
Repeat this for every monthly starting point from November 1995 to April 2021 — that's 302 separate experiments. The results:
| Threshold the Investor Was Waiting For | Correction Arrived | Correction Never Came | Median Wait (when it came) | Mean Opportunity Cost (vs immediate) |
|---|---|---|---|---|
| 10% drawdown | 65.9% | 34.1% | 0.40 yr | ~11% of capital |
| 15% drawdown | 50.0% | 50.0% | 0.63 yr | ~24% of capital |
| 20% drawdown | 40.7% | 59.3% | 0.75 yr | ~27% of capital |
Source: NLE backtest engine on 7,587 daily Nifty 50 closes (Nov 1995 → May 2026). 302 starting points sampled at monthly intervals. Cash return = 6.5% nominal (liquid fund yield). Mean opportunity cost = average of (immediate-deployment 5-year wealth − wait-strategy 5-year wealth), expressed as % of starting capital.
The pattern is simple: the bigger the correction you wait for, the less likely it is to arrive — and the more it costs to keep waiting. The patience is real. The reward isn't.
The most painful part of the data is the 34% of starts where a 10% correction simply never arrived in five years. The waiter, having refused to invest, watched from the sidelines.
What did they watch? In those 103 windows, the index averaged a 133% gain over the five years. The median was +100%. The smallest gain was +25%. The biggest was +526%.
The cash did its job — it earned 6.5% and grew the waiter's capital to 1.37x. But the investor who deployed grew theirs to 2.33x. Almost a doubling of capital separated the two — the price of waiting for a correction the market never had to deliver.
The usual defence of the wait-for-correction strategy: when the correction arrives, the cheaper entry pays for the wait. The data says otherwise.
Even when we look only at the starts where the 20% correction did arrive within five years, the "invest immediately" strategy still won 54% of the time. The correction came — but the market had usually run higher first, the invested cohort had already compounded ahead, and even after the drop, they were still further along than the cohort that waited.
| Holding Horizon | 1-Year | 3-Year | 5-Year | 10-Year |
|---|---|---|---|---|
| % of windows where deployed equity beat cash @ 6.5% | 60.8% | 72.6% | 76.4% | 98.1% |
| % where cash beat deployed equity | 39.2% | 27.4% | 23.6% | 1.9% |
Across all overlapping rolling windows of each length on Nifty 50 daily closes, Nov 1995 → May 2026. Cash assumed at 6.5% liquid-fund yield, compounded.
The single phrase that traps more investors than any other: "Why would I invest at an all-time high? It can only fall from here."
The data on Indian equity since 1995 says the opposite.
The reason is structural. A healthy market that compounds upward will, by definition, spend most of its life near or at all-time highs. The Indian index has hit new highs on 7% of all trading days across 30 years. Waiting for "below the ATH" often means waiting for a regime that never arrives — while the chart you keep watching keeps compounding without you.
The strongest version of this test: imagine the unluckiest investor possible — one who put their entire capital in on the absolute peak day before a major Indian crash. Then run the clock forward to today.
The point isn't that timing the bottom is impossible (it is). The point is that being maximally wrong about the entry still costs less than refusing to enter. The gap isn't small. It's the central feature of long-run equity compounding — and it's the simple empirical answer to the wait-for-correction instinct.
A more recent, smaller-scale example: an investor who held ₹1 Cr in cash from 14-Jan-2020 — the pre-COVID peak, weeks before the index crashed 38% — through to today. Their cash at 6.5% became ₹1.49 Cr. The investor who deployed that day, despite the immediate 38% drop, ended at ₹1.96 Cr. The cost of patience: ₹47 lakh on a single crore, in just over six years.
The cost of waiting is settled by the data. The harder question is what to do instead — especially for an investor genuinely worried about a bad entry. Five mechanisms beat "wait."
What ties all five together: they replace prediction with process. Waiting for a correction is a forecast dressed up as caution. The forecast usually fails. The price of the forecast is the cost of cash.
"Cash held for safety is one decision. Cash held in pursuit of a better entry is another. The first is architecture. The second is a forecast. Forecasts have a price. The price is approximately 5.5% per year on the entire capital being held back — compounded against the investor for as long as the wait persists. Across thirty years of Indian equity, the correction the waiter is waiting for did not arrive in a third to two-thirds of starting points. Even when it arrived, it usually arrived too late to redeem the wait. Even the worst-day investor — deployed at a peak, crashed thirty percent the next month — ended with nearly twice what the patient cohort had. The market is not waiting. The capital that waits with it does not catch up."
Vanguard Research (2012, updated 2023). Dollar-Cost Averaging Just Means Taking Risk Later. Empirical study across US, UK, and Australian markets showing lump-sum deployment beat dollar-cost averaging in approximately two-thirds of historical windows.
Charles Schwab (2021). Does Market Timing Work? Multi-decade study comparing perfect-timing, immediate-deployment, dollar-cost-averaging, and "stay-in-cash" investors. The cash investor finished last by a wide margin in every scenario tested.
Dimensional Fund Advisors. Multiple papers on time-in-market vs timing-the-market. Recurring finding: missing the best 10 days in any decade reduces terminal wealth by 30–50%, and those days disproportionately fall during drawdowns the cash-holder is sitting through.
Lynch, P. Quoted: "Far more money has been lost by investors trying to anticipate corrections than has been lost in the corrections themselves." Originally attributed in remarks circa 1994; cited in Bogle, J. C. (2007), The Little Book of Common Sense Investing.
NLE backtest engine. All Indian-market figures in this paper are computed from daily Nifty 50 close prices, 06-Nov-1995 to 08-May-2026 (7,587 trading days). Cash return assumed at 6.5% nominal (representative liquid-fund yield, refreshed against current AMFI category averages). Simulation harness: 302 monthly starting points; 5-year forward horizons.
S&P Dow Jones Indices — SPIVA Scorecards. Persistence and active-vs-passive evidence informing the broader thesis that uninvested capital's headwind is large enough to dominate fund-selection considerations.