Buying last year's losers: the NIFTY 500's worst performers kept losing
In short Buying the NIFTY 500’s 20 biggest losers of the past year and holding them for a year lost money: about −1% a year over 12½ years, against about +13% for the index, with an 80% fall along the way. Last year’s losers mostly kept losing.
The idea
Buy what has fallen the most and wait for it to bounce back. It’s one of the most natural instincts in investing, and it has a respectable academic history: a famous 1985 study found that US stocks with the worst returns over the past three to five years went on to beat the market. Prices overshoot, the argument goes, so the most hated stocks become too cheap.
It is also the mirror image of momentum, which buys last year’s winners. We wanted to see what buying last year’s losers does on Indian stocks, tested honestly: on the stocks that were actually in the NIFTY 500 at each point in time, including the ones that later failed, with costs.
The rules
- Universe: NIFTY 500 stocks, point-in-time: on each date, only stocks in the index on that date. A holding is sold if its stock leaves the index; the cash waits for the next rebalance.
- Each year, on the last trading day of December: rank the stocks by their total return over the past year (252 trading days) and hold the 20 worst, in equal weights. Stocks leaving the list are sold, new ones bought, and every holding is reset to an equal share.
- Fills: at the next trading day’s open. Cost: 0.15% per side, on the amount traded.
- No stop-loss, no target: always fully invested.
The backtest
Data: NIFTY 500 stocks, 1 January 2014 to 3 July 2026 (12½ years), about 20 new holdings a year.
| CAGR | Worst drawdown | Calmar | |
|---|---|---|---|
| Last year’s 20 biggest losers | −0.9% | −80.4% | — |
| NIFTY 500 buy & hold (same window) | 13.3% | −38.3% | 0.35 |
(Calmar is CAGR divided by the worst peak-to-trough drawdown; it isn’t meaningful for a losing strategy. The NIFTY 500 figure is the price index; the backtest’s stock prices exclude dividends too, so the comparison is like for like.)
₹100 invested at the start was worth about ₹88 at the end, against about ₹480 in the index.
Year by year, against the index:
| Year | Losers | NIFTY 500 | Difference |
|---|---|---|---|
| 2014 | +7% | +38% | −31 pts |
| 2015 | −22% | −1% | −21 pts |
| 2016 | −15% | +4% | −18 pts |
| 2017 | +51% | +36% | +15 pts |
| 2018 | −27% | −3% | −24 pts |
| 2019 | −30% | +8% | −38 pts |
| 2020 | +17% | +17% | +0 pts |
| 2021 | +17% | +30% | −13 pts |
| 2022 | −17% | +3% | −20 pts |
| 2023 | +35% | +26% | +9 pts |
| 2024 | −11% | +15% | −26 pts |
| 2025 | +7% | +7% | +1 pt |
| 2026* | +8% | −2% | +11 pts |
*2026 to 3 July. It made money in only 7 of the 13 calendar years and beat the index in 5.
The rebounds do come, in speculative years like 2017 and 2023, but they never made up for the years when the losers simply kept falling.
Robustness tests
All on the same engine; the first row is the headline result above.
| Test | CAGR | Worst drawdown |
|---|---|---|
| As tested, 0.15% per side | −0.9% | −80.4% |
| Double cost, 0.30% per side | −1.2% | −80.8% |
| First half, Jan 2014 – Dec 2019 | −9.2% | −64.3% |
| Second half, Jan 2020 – Jul 2026 | 7.5% | −53.0% |
| Hold stocks that leave the index until the next rebalance | −0.1% | −88.5% |
| Setting changed | CAGR | Worst drawdown |
|---|---|---|
| 10 stocks instead of 20 | −8.2% | −88.0% |
| 30 stocks | 1.6% | −78.5% |
| Losers over 6 months instead of a year | −1.5% | −82.5% |
| Losers over 2 years | 2.6% | −73.6% |
| Losers over 3 years | 6.3% | −77.7% |
| Rebalance every six months | −2.7% | −89.9% |
- Nothing rescues it. No setting came close to the index’s 13% a year, and every one had a drawdown of −74% or worse. Starting in a later year (July 2014, 2015 or 2016) gave −3.1% to +0.4% a year.
- The losers really were in trouble. The worst holdings were companies in genuine collapse, such as an airline that stopped flying and an infrastructure company caught in a group default. Over 12 years, 92 of the 246 holdings were dropped from the NIFTY 500 during the year they were held.
- It isn’t our index rule. We sell a stock when it leaves the index. Holding those stocks until the next rebalance instead made the result no better (−0.1% a year) and the drawdown deeper (−88%).
- A hint of the longer-term effect. Buying 3-year losers, closer to the original study, made 6.3% a year. That is still less than half the index, with a −78% drawdown along the way.
Risks
- Catastrophic drawdowns. −80% from June 2014 to March 2020: almost six years of losses.
- Distressed companies. The list fills up with businesses facing default, fraud or collapse. Some never recover.
- Index exits. Many holdings get dropped from the index, which often means they keep falling or become hard to trade.
- Waiting for a rebound that doesn’t come. The occasional strong year (2017, 2023) makes the idea look like it’s about to work.
Pros & cons
Pros
- Simple, once-a-year routine.
- Cheap to run: costs barely affect the result.
Cons
- Lost money over 12½ years, against about 13% a year for the index.
- An 80% drawdown lasting almost six years.
- Made money in only 7 of 13 years.
- Concentrated in the market’s most troubled companies.
Our verdict
Rejected. It failed seven of our eight checks; the only one it passed is having enough data. Buying the NIFTY 500’s 20 biggest losers each year lost about 1% a year over 12½ years, against a gain of about 13% a year for the index, and fell 80% along the way. No setting we tried, and no start date, changed the picture. In this market over this period, last year’s losers mostly kept losing.
Was this useful?
More to say, or spotted a mistake? Write to [email protected]. We read everything.
Hypothetical, backtested results on historical data at least three months old. They include modelled costs but can't capture every real-world effect, and past performance does not predict future results. This is educational research, not a recommendation. Full disclaimer.