Golden cross and death cross on NSE stocks

A moving average is a summary of where price has been. The 50-day speaks for the last quarter, the 200-day for the last ten months. So when the faster line overtakes the slower one:

where sma(50) crosses above sma(200)

what has been recorded is that the recent regime beat the long-term regime for long enough to drag a quarter's average up through a year's. Months of net strength. A two-week rally cannot produce it.

That is the golden cross, and its funeral-suited twin the death cross is the same lines crossing the other way. The names are theatrical and the financial press loves them for it. The mechanics are more interesting than the names, because the same property that makes the signal real is the property that makes it disappointing.

Late by design

By the time a golden cross prints, the stock has usually already risen a long way off its low. The cross confirms a trend change that happened earlier. It does not catch one, and anybody using it to time an entry gets disappointed on a predictable schedule. I spent a year treating it as a buy trigger and the year taught me nothing except that the signal is not what I thought it was.

What it is good for is deciding which list a stock belongs to.

Where each side earns its keep

The golden cross works as a universe definition. Stocks that have printed one and hold above both averages are the pond where pullback-buying and breakout entries behave the way the textbooks claim. The developed form of that regime is price above the 20, above the 50, above the 200, in that order, which is the Stacked moving averages scan. The cross itself, on the day it fires, is Golden cross.

The death cross is a worse shorting signal than its reputation suggests. Bear moves often end violently right around where it prints, and I would not trade against a stock on the strength of it. As a prompt to ask "why am I still holding this?", it is much better. Reviewing whatever turns up on the death cross scan costs a few minutes a week and occasionally saves a great deal more.

The whipsaw tax

The failure mode is well known. A sideways market walks the two averages into each other over and over, printing crosses that reverse within weeks. Every crossover system pays this tax. The only real questions are how often it pays and whether the trending stretches cover the bill.

Two mitigations are standard. Require price itself to confirm on the cross day with close > sma(200). Or use the earlier and related signal of price reclaiming the 200-day directly, which is the Bounce off the 200-day scan.

"Does it work?" is a measurable question

For NSE stocks in this particular year, the honest answer is not in any article, mine included. Open the golden cross and read the hit-rate panel: every session in the last 250 where the scan matched, and what followed over the next 1, 5 and 20 days. It is close-to-close and it ignores costs, so treat it as a filter.

What I expect you will see is what the mechanics predict. Crosses fire rarely, they cluster at regime turns, and they look brilliant in a trending year while paying the whipsaw tax in a flat one. Read the panel to find out which kind of year this has been, then do the actual stock-picking one level down, inside the momentum family, where the scans assume the regime and go looking for entries.