Corporate Actions and Adjusted Prices: Why Price History Must Be Restated

Corporate actions are company decisions that change the number of shares outstanding or pay value out to holders, and they break raw price history because the quoted price moves for reasons that have nothing to do with the value of the business. A share that splits into five sees its price fall to roughly a fifth overnight. Nobody lost anything. But an unadjusted price series records a large fall, and every calculation built on that series is wrong from that day backwards.

Adjusted prices fix this by restating older prices onto today’s share basis, so that a return computed across the event reflects what an investor actually experienced rather than an accounting artifact.

What counts as a corporate action

The category is broad, but for price history purposes it splits into actions that change the share count and actions that pay value out.

Stock split. Each existing share is divided into a larger number of shares, with the face value divided in the same proportion. Total ownership is unchanged. The quoted price falls proportionally on the ex-date.

Bonus issue. New shares are created from the company’s reserves and distributed to existing holders in proportion to their holdings. No money comes in. Again the share count rises and the price falls proportionally.

Reverse split or consolidation. The opposite of a split. Shares are combined into fewer shares of higher face value, and the price rises proportionally.

Rights issue. Existing holders are offered new shares, usually at a price below the prevailing market price, in proportion to what they hold. Because new money does come in, but at a discount, the effect on price is a blend of dilution and capital injection. This is the fiddliest of the common actions to adjust for correctly.

Dividend. Cash paid out of the company. On the ex-dividend date the share generally opens lower by roughly the dividend amount, because a holder buying from that date does not receive it.

Buyback, merger, demerger and scheme of arrangement. These change the entity itself rather than just the share count. A demerger in particular hands holders shares in a newly listed entity, so a single price series stops describing a single set of assets.

In India, the mechanics of these events run through the exchanges and the depositories. The company’s board approves the action, the company announces it to the exchanges, and a record date is fixed to determine who is entitled. The ex-date is the first day on which the share trades without the entitlement, and it is the ex-date, not the record date or the announcement date, that the price series must pivot on.

How adjustment actually works

The mechanic is an adjustment factor applied to every price before the ex-date.

For a split or bonus, the factor is simply the ratio of old shares to new. If one share becomes five, every price before the ex-date is divided by five, and every historical volume or share-count figure is multiplied by five. The series is then continuous: a price of 500 before the event and 100 after it now reads as 100 before and 100 after, which correctly says nothing happened to the investor’s wealth.

For a dividend, the same idea applies but the factor is derived from the dividend relative to the price just before the ex-date. Prices before the event are scaled down slightly, so that the series treats the cash payment as if it stayed inside the share.

For a rights issue, the adjustment uses a theoretical price that blends the pre-event market price with the discounted subscription price, weighted by the number of old and new shares. It is theoretical because it assumes the holder participates, which not every holder does.

Two consequences follow, and both are frequently missed.

Adjusted prices are not the prices anyone traded at. After a five-for-one split, the historical adjusted series says the share traded at 100 three years ago. It did not. It traded at 500. The adjusted number is a restatement, useful for measuring returns and useless for reconstructing what a screen looked like on that day.

Every per-share figure must be adjusted with the price, or the ratios break. If you divide an adjusted price by an unadjusted earnings per share, the resulting valuation multiple is wrong by exactly the split factor. This is one of the most common silent errors in home-built datasets, and it is the reason careful desks treat corporate actions as a data-integrity problem rather than a display problem.

How to read adjusted data

Know which series you are holding. A price-only series adjusted for splits and bonuses answers “what happened to the share price”. A total return series that also adds back dividends answers “what happened to an investor’s money”. Comparing a fund’s total return against a price-only index understates the index, which is the point of total return index versus price index.

Check for a seam where two data sources join. If part of your history is adjusted and part is not, the join creates a step that looks like a real move. Any long price series assembled from more than one source needs an explicit check for this, because the artificial gap will show up as a spectacular one-day return in a backtest.

Suspect large single-day moves before investigating them. A very large gap in a price series is more often an unhandled corporate action than a real event. Checking the exchange announcement for that date is the first step, not the last.

Understand what a demerger does to the series. When a business is carved out into a separately listed company, the parent’s price adjusts down and holders receive shares in the new entity. A single continuous series for the parent, adjusted or not, no longer describes one consistent set of assets. This is the price-side twin of the reporting problem covered in why restatements break models.

Match the price basis to the fundamentals basis. If your fundamental history is stated as originally filed, and your price history is fully adjusted, then every ratio spanning a corporate action is mixing two bases. Consistency across the two sides matters more than which convention you pick.

Adjustment is not a cosmetic step applied at the end. It is a statement about what the series is measuring, and getting it wrong quietly corrupts every return, ratio and backtest built on top.

What adjusted prices do not tell you

They do not tell you what you would have paid. Adjusted history is a return-measurement device. Any question about actual traded levels, historical liquidity in rupee terms, or what a price screen showed on a past date needs the unadjusted series.

They do not capture taxes or costs. A total return series that reinvests dividends assumes reinvestment happens instantly, in full, at the closing price, with no tax and no transaction cost. Real investors face all three. The gap is small over short periods and compounds over long ones.

They assume participation in rights issues. The theoretical adjustment for a rights issue assumes the holder subscribed. A holder who did not participate was diluted and experienced a different outcome than the adjusted series shows.

They do not fix survivorship. Adjustment cleans the price history of shares that still trade. It does nothing about companies that were delisted, merged away or suspended, whose histories often disappear from a dataset entirely. That is a separate and larger problem, covered in survivorship bias in backtests.

They do not say anything about the business. A split does not create value, a bonus issue does not create value, and neither is evidence of anything about future performance. They change the unit of account. Reading a corporate action as a signal about the company confuses arithmetic with information.

They can be wrong. Adjustment factors are applied by data providers from exchange announcements, and mistakes happen: a missed action, a factor applied on the wrong date, an action applied twice. The consequence is a permanent distortion in the series that is invisible unless you look for it. Spot-checking large historical gaps against announcements is one of the least glamorous and most valuable checks in a research dataset.

Corporate action handling is plumbing. Nobody notices it when it is right, and when it is wrong it does not throw an error. It just produces returns, ratios and backtests that are quietly, confidently incorrect.

Related reading

• Portfolio metrics explained: the hub for how portfolio and market data is measured and read.
• Total return index versus price index: the dividend question at index level, and why fund comparisons need the total return version.
• Why point-in-time data matters: the same discipline applied to reported fundamentals rather than prices.
• Why restatements break models: what happens to a series when the underlying entity or basis changes.
• Survivorship bias in backtests: the companies missing from your price history altogether.

This article is educational. Altys Labs is not a registered research analyst or investment adviser, and nothing here is investment advice or a recommendation to buy, sell, or hold any security.