Splits, bonuses and dividends: how corporate actions break your chart

A stock splits 1:5 and a ₹2,500 share becomes a ₹500 share overnight. Economically nothing has happened — every holder owns five times as many shares worth a fifth as much. But if your data vendor has not restated history, your chart now shows an 80% collapse that never occurred, and every indicator reading that chart is computing on a fiction.

The four that matter

Stock split. One share becomes several, price divides accordingly. Companies do it to bring a high nominal price into a range retail investors find approachable. Market capitalisation is unchanged.

Bonus issue. Additional shares issued free to existing holders — a 1:1 bonus doubles your holding and halves the price. Mechanically similar to a split, legally different, and it hits your chart the same way.

Dividend. Cash paid out. On the ex-dividend date the price typically drops by roughly the dividend amount, because the company is now worth exactly that much less cash. A small, regular, easily-ignored discontinuity.

Rights issue. Existing holders get the right to buy new shares, usually below market. The dilution changes the value of what was already held.

Mergers, demergers and buybacks add further complications, but these four cover the overwhelming majority of what will affect a chart you are looking at.

What adjustment actually does

Adjustment restates history so the series stays comparable across the event. For a 1:5 split, every price before the split date is divided by five, and every volume figure multiplied by five.

The result is a continuous series. A 10% rally three years ago still reads as a 10% rally. Your moving average does not lurch. Your percentage returns remain meaningful.

Note what this means: adjusted historical prices are not the prices that traded. If you look up what the stock actually printed on a date before the split, it will not match your adjusted chart. That is correct and intended — but it surprises people who are reconciling against an old contract note, and it is worth understanding rather than reporting as a bug.

Where vendors get this wrong

Almost every vendor adjusts the daily series. It is the visible one, it is what people check, and the corporate action data is published for it.

The failure is narrower and more damaging: many adjust the daily series and not the intraday one.

So your daily chart of a recently-split stock looks perfect. Your five-minute chart of the same stock has a cliff in it. And every intraday indicator — profile, VWAP, moving averages, any strategy running on that series — is reading a discontinuity as a real price move.

Three symptoms, in order of how often people report them:

• A backtest with an inexplicable loss or gain on one date. The strategy did not fail; it traded a phantom gap.
• An indicator that goes haywire for a period after the split, as its lookback window straddles the discontinuity.
• Support and resistance levels that look wrong, because levels drawn before the event are on an unadjusted scale.

The volume half, which gets forgotten

Price adjustment gets attention. Volume adjustment gets skipped, and it matters just as much for anything volume-based.

After a 1:5 split, the same rupee value of trading represents five times as many shares. If historical volumes are not scaled, every pre-split day appears to have a fifth of the activity it really had. Relative-volume comparisons across the split date become meaningless, and any “unusual volume” screen will systematically misfire on recently-split names.

Worth asking a vendor explicitly, because it is a separate process from price adjustment and is genuinely sometimes missed.

Corporate actions and derivatives

Futures and options contracts are adjusted too, and the mechanics differ: strike prices and lot sizes are revised so contract holders are economically unaffected.

The consequence for data is that an option chain around a corporate action can contain strikes at unusual values — adjusted rather than round numbers. Systems that assume strikes fall on tidy intervals will simply fail to display them, which looks like missing data and is not.

What to ask, and what to do

Four questions worth putting to a vendor in writing:

1. Are corporate actions applied to intraday history, or only the daily series? The single most useful question here.
2. Is volume adjusted as well as price? A separate process, and separately skippable.
3. How far back does adjustment reach? Some vendors adjust recent events and leave older history untouched, which produces a series that is correct at one end.
4. Can I get unadjusted data if I need it? Sometimes you do — reconciling against contract notes, or studying what actually printed.

And on your own side: keep a list of corporate actions in the instruments you trade. When something looks strange in a backtest, the first question worth asking is whether a corporate action fell in that window. It is astonishing how often the answer is yes, and how much time is lost debugging strategy logic before anyone checks.

None of this is exotic. It is simply a category of problem that is invisible until it costs you something, and trivially checkable once you know to look.