Adjusting Price History for Corporate Actions, Explained

Look at a raw, unadjusted price chart of a stock that split 4-for-1 last year and you'll see a price that appears to fall 75% overnight for no reason — nothing happened to the company, only the number of shares changed. A raw chart also shows a small drop on every ex-dividend date. Neither drop reflects a real loss to a holder who kept the shares (and reinvested any cash). Adjusting for corporate actions rewrites the historical price series so it reflects what an investor actually experienced, not accounting artifacts.

The mechanism is an adjustment factor applied to every price before the action date, working backward from today. For a split of ratio $r$ (e.g. 4-for-1 means $r=4$ ):

P_{\text{adjusted}} = \frac{P_{\text{raw}}}{r}

For a cash dividend $D$ paid when the pre-dividend price was $P_{\text{cum}}$ , the standard approach scales every earlier price by the factor $(1 - D/P_{\text{cum}})$ , compounding one factor per historical dividend so that a long-run adjusted series reflects dividends reinvested at the ex-date price.

Adjustment always works backward from the present: today's price is never touched, and every dividend or split further in the past multiplies one more factor into everything before it. That is why a stock's "adjusted price from ten years ago" keeps quietly changing every time a new dividend is paid today — the historical series is being rescaled, not the past itself.

The raw series has a fake cliff on the split date; the adjusted series is smooth, because it reflects a holder's actual continuous return.

A worked example

A stock's raw closing prices are $100 (Jan), $105 (before a 2-for-1 split in Feb, becoming $52.50 raw after), $55 (Jun). To adjust the January price for the split, divide by the split ratio: $100 / 2 = 50$ , i.e. $50 adjusted. Now every subsequent price is untouched — $52.50 in Feb, $55 in Jun — while the adjusted January price of $50 makes the whole series consistent: a holder's return from January ($50 adjusted) to June ($55 raw, unaffected by the earlier split) is a clean $10\%$ gain, matching what actually happened to their wealth, rather than the raw series's misleading $-45\%$ apparent move around the split.

Backtests that pull "close" prices without checking whether they are split-and-dividend adjusted will show phantom crashes and phantom rallies around every corporate action — and worse, a strategy that happens to buy right before a split can show a fake, enormous return. Always confirm a data vendor's price field is adjusted, and know whether dividends are included (a "total return" adjusted series) or excluded (a "price return" adjusted series).

Adjusting Price History for Corporate Actions

A worked example

Related concepts

Practice in interviews

Further reading