Stop loss for long term investors: a trailing stop tested on the Nifty 50 since 2012

What a stop loss is meant to do

A stop loss is an order to sell if the price falls to a level you set in advance. A fixed stop sits at one price, say 10% below what you paid. A trailing stop moves up with the price: it sits a set percentage below the highest price since you bought, so it locks in part of a gain.

The appeal is obvious. It promises to cut a loss before it grows, without needing a decision in the moment. The risk management module covers stops alongside the other tools, and the note stops set by feel, not by math shows how most stop levels are chosen.

But a stop answers only half the question. It says when to get out. It does not say when to get back in, and for long term money that second decision is where most of the cost hides.

The test: a trailing stop on the Nifty 50

To see what a stop does to a diversified holding, I ran the simplest version on NSE's daily closes of the Nifty 50, from 31 Dec 2012 to 12 Oct 2026:

1. Hold the indexStart fully invested at the close of 31 Dec 2012.
2. Track the highNote the highest close since the last purchase.
3. Sell on the stopSell at the close when the index is X% below that high.
4. Buy back on the reboundBuy at the close when the index is X% above its lowest close since the sale.

Cash earns nothing while out of the market, and there are no costs or tax unless stated. Both make the stop look slightly better than real life on costs and tax, and slightly worse on interest. It is the Nifty 50 price index, without dividends.

The stops did what they promised on the downside: every one cut the deepest fall, the 5% stop to about half. But they paid for it in return, and not by a little. The 10% stop ended with ₹2.23 lakh against ₹3.79 lakh for doing nothing. Adding a cost of 0.2% on each trade, the 10% stop drops further to 5.58% a year and the 5% stop to 5.95%.

Why the stops cost so much: whipsaw

The 10% stop worked well once. It sold on 6 Mar 2020 at 10,989.45, before the worst of the pandemic crash, and bought back on 26 Mar 2020 at 8,641.45. That single round trip saved about 21%.

Then the market did what markets do after a crash: it swung. The same rule sold again on 18 May 2020 at 8,823.25 and bought back on 1 Jun 2020 at 9,826.15, 11% higher. In 2022 it was worse:

1. 20 Dec 2021Sell at 16,614.20

   10% below the high of 18,477.05.

2. 17 Jan 2022Buy back at 18,308.10

   10% above the low since the sale, and 10% above the sale price.

3. 24 Feb 2022Sell at 16,247.95

   The day the war in Ukraine began.

4. 30 Mar 2022Buy back at 17,498.25

   7.7% above the sale price.

5. 10 May 2022Sell at 16,240.05

   The third sale in five months.

6. 28 Jul 2022Buy back at 16,929.60

   The index ended 2022 at 18,105.30.

Three round trips in seven months, each selling lower and buying higher. A broad index rarely falls in a straight line; it drops, bounces, drops again. A stop narrow enough to catch the start of a fall is also narrow enough to be triggered by the ordinary noise inside a rising market. The blog note on a stop that did not stop adds the other problem: in a gap down, a stop order fills at the next price, not at the level you set.

The tax that the table leaves out

Each sale in your own demat is a taxable event. Ten of the 10% stop's fourteen sales came within a year of the purchase before them, so any gain was short term, taxed at 20% under the rules for transfers from 23 Jul 2024, and the clock for the 12 month long term holding restarts at every purchase. A buy and hold investor in an index fund pays tax once, on sale, at 12.5% above ₹1.25 lakh of long term gains a year. The capital gains guide works through the arithmetic. On a taxable portfolio, the gap in the table widens.

Where a stop loss does make sense

None of this means stops are useless. They do a different job in different places:

• A single stock. A company can fall 80% and never come back; an index of 50 companies has always recovered so far. A stop on a concentrated position protects against the outcome a diversified index does not have.
• Short term trading. A trader with a defined edge and a defined risk per trade uses a stop to keep each loss small. That is position sizing, covered in the position sizing module.
• Leverage. Borrowed money or F&O positions can lose more than the cash put in. A hard exit is part of surviving.

For a diversified long term holding, the cheaper ways to limit a fall are the ones that do not require timing the exit and the re-entry: hold an amount of equity whose worst fall you can sit through, keep money needed soon out of equity, and rebalance on a calendar.

The module on the disposition effect explains the opposite failure: investors who refuse to sell a falling stock at all. A rule protects you from both errors only if it is written before the fall.

How this shows up in RupeeCase

RupeeCase rules do not use price stops on the portfolio. They control risk through the universe, the number of stocks, equal weights and a rebalance calendar: a stock that falls out of favour is replaced at the next rebalance rather than sold on a trigger in between. Midcap holds 20 stocks from the Nifty Midcap 150, rebalanced every 2 weeks: model CAGR 38.75%, deepest fall −24.87%, as of 27 Apr 2026. The deepest fall is shown on its page with its dates, so the risk is known before you invest, and the drawdown recovery calculator shows how long a fall of that size takes to recover at a given return.

Questions investors ask